d0p1/ack
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity
ack/mach/powerpc/ncg/table

2620 lines
68 KiB
Plaintext
Raw Permalink Normal View History

Add some comments before I forget how this stuff works.
2018-01-24 20:17:32 +00:00
/*
* PowerPC table for ncg
*
* David Given created this table.
* George Koehler made many changes in years 2016 to 2018.
*
* This back end provides 4-byte integers, 4-byte floats, and 8-byte
* floats. It should provide enough of EM for the ACK's compilers.
* - It doesn't provide "mon" (monitor call) nor "lor 2", "str 2"
* (heap pointer). Programs should call procedures in libsys to
* make system calls or allocate heap memory.
* - It generates only a few EM traps:
* - EARRAY from aar, lar, sar
* - ERANGE from rck
* - ECASE from csa, csb
* - It uses floating-point registers to move 8-byte values that
* aren't floats. This might cause extra FPU context switches in
* programs that don't use floating point.
*
* The EM stack is less than optimal for PowerPC, and incompatible
* with the calling conventions of other compilers (like gcc).
* - EM and ncg use the stack to pass parameters to procedures. For
* PowerPC, this is probably slower than passing them in registers.
* - This back end misaligns some 8-byte floats, because EM's stack
* has only 4-byte alignment. (This kind of misalignment also
* happened in IBM's AIX and Apple's Mac OS, where data structures
* had 8-byte floats with only 4-byte alignment.)
*/
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
EM_WSIZE = 4
EM_PSIZE = 4
EM_BSIZE = 8 /* two words saved in call frame */
FP_OFFSET = 0 /* Offset of saved FP relative to our FP */
PC_OFFSET = 4 /* Offset of saved PC relative to our FP */
Turn off comments again. I turned them on by accident in c416889.
2017-01-30 20:45:46 +00:00
#define COMMENT(n) /* comment {LABEL, n} */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
#define nicesize(x) ((x)==1 || (x)==2 || (x)==4 || (x)==8)
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
#define smalls(n) sfit(n, 16)
#define smallu(n) ufit(n, 16)
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
/* Finds FRAME_V tokens that overlap myoff, mysize. */
#define fover(myoff, mysize) (%off+%size>(myoff) && %off<((myoff)+(mysize)))
/* Checks if we can use {LXFRAME, x}. */
#define nicelx(x) ((x)>=1 && (x)<=0x8000)
In powerpc table, fix macros los() and his(). Change the operator in his() from a - minus to a + plus. When los(n) becomes negative, then his(n) needs to add 0x10000, not subtract it. Also change los(n) to do the sign extension, because smalls(los(n)) should be true, not false. Also change hi(n) and lo(n) to wrap n in parentheses, as (n), because these are macros and n might still contain operators.
2016-10-15 03:59:26 +00:00
#define lo(n) ((n) & 0xFFFF)
#define hi(n) (((n)>>16) & 0xFFFF)
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Use these for instructions that treat the low half as signed --- his()
* includes a modifier to produce the correct value when the low half gets
* sign extended. Er, do make sure you load the low half second. */
In powerpc table, fix macros los() and his(). Change the operator in his() from a - minus to a + plus. When los(n) becomes negative, then his(n) needs to add 0x10000, not subtract it. Also change los(n) to do the sign extension, because smalls(los(n)) should be true, not false. Also change hi(n) and lo(n) to wrap n in parentheses, as (n), because these are macros and n might still contain operators.
2016-10-15 03:59:26 +00:00
#define los(n) (lo(n) | (((0-(lo(n)>>15)) & ~0xFFFF)))
#define his(n) ((hi(n) + (lo(n)>>15)) & 0xFFFF)
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
PROPERTIES
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
GPR /* general-purpose register */
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
SPFP /* sp or fp */
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
REG /* allocatable GPR */
REG3 /* coercion to r3 */
FPR(8) /* floating-point register */
FREG(8) /* allocatable FPR */
FSREG /* allocatable single-precision FPR */
SPR /* special-purpose register */
CR /* condition register */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
REGISTERS
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
/*
Add some comments before I forget how this stuff works.
2018-01-24 20:17:32 +00:00
* We use r1 as stack pointer and r2 as frame pointer.
* Our assembler has aliases sp -> r1 and fp -> r2.
*
* We preserve r13 to r31 and f14 to f31 across function
* calls to mimic other compilers (like gcc). See
* - http://refspecs.linuxbase.org/elf/elfspec_ppc.pdf
* - https://github.com/ryanarn/powerabi -> chap3-elf32abi.sgml
* - Apple's "32-bit PowerPC Function Calling Conventions"
*
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
* When ncg allocates regvars, it seems to start with the last
* register in the first class. To encourage ncg to allocate
* them from r31 down, we list them in one class as
* r13, r14, ..., r31: GPR, REG regvar(reg_any).
*/
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
r0, r12 : GPR.
sp, fp : GPR, SPFP.
Add splitting coercions for IND_ALL_D. Delete my wrong comment (from commits cfbc537, a8f62f4, 5432bd0) which claimed that such coercions are not possible.
2017-12-19 01:59:04 +00:00
r3 : GPR, REG, REG3.
r4, r5, r6, r7, r8, r9, r10, r11 : GPR, REG.
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
r13, r14, r15, r16, r17, r18, r19, r20, r21, r22, r23, r24,
r25, r26, r27, r28, r29, r30, r31
: GPR, REG regvar(reg_any).
Check in incomplete Linux PowerPC and M68K work. --HG-- branch : dtrg-experimental-powerpc
2013-05-07 23:48:48 +00:00
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
f0 : FPR.
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
f1, f2, f3, f4, f5, f6, f7, f8, f9, f10, f11, f12, f13
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
: FPR, FREG.
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
f14, f15, f16, f17, f18, f19, f20, f21, f22, f23, f24, f25,
f26, f27, f28, f29, f30, f31
: FPR, FREG regvar(reg_float).
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
fs1("f1")=f1, fs2("f2")=f2, fs3("f3")=f3, fs4("f4")=f4,
fs5("f5")=f5, fs6("f6")=f6, fs7("f7")=f7, fs8("f8")=f8,
fs9("f9")=f9, fs10("f10")=f10, fs11("f11")=f11, fs12("f12")=f12,
fs13("f13")=f13
: FSREG.
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
Add reglap to ncg. Add 4-byte reg_float to PowerPC ncg. The new feature "reglap" allows two sizes of floating-point register variables (reg_float), if each register overlaps a single register of the other size. PowerPC ncg uses reglap to define 4-byte instances of f14 to f31 that overlap the 8-byte instances. When ncgg sees the definition of fs14("f14")=f14, it removes the 8-byte f14 from its rvnumbers array, and adds the 4-byte fs14 in its place. Later, when ncg puts a variable in fs14, if it is an 8-byte variable, then ncg switches to the 8-byte f14. The code has /* reglap */ comments in util/ncgg or #ifdef REGLAP in mach/proto/ncg reglap became necessary because my commit a20b87c caused PowerPC ego to allocate reg_float in both 4-byte and 8-byte sizes.
2017-10-14 16:40:04 +00:00
/* reglap: reg_float may have subregister of different size */
fs14("f14")=f14, fs15("f15")=f15, fs16("f16")=f16, fs17("f17")=f17,
fs18("f18")=f18, fs19("f19")=f19, fs20("f20")=f20, fs21("f21")=f21,
fs22("f22")=f22, fs23("f23")=f23, fs24("f24")=f24, fs25("f25")=f25,
fs26("f26")=f26, fs27("f27")=f27, fs28("f28")=f28, fs29("f29")=f29,
fs30("f30")=f30, fs31("f31")=f31
: FSREG regvar(reg_float).
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
lr, ctr : SPR.
Add some comments before I forget how this stuff works.
2018-01-24 20:17:32 +00:00
cr0 : CR. /* We use cr0, ignore cr1 to cr7. */
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
Add tests for C <setjmp.h> and Modula-2 Semaphores. Fix PowerPC ncg so setjmp() returns the correct value. I got unlucky when ncg picked r3 for "uses REG"; this destroyed the return value in r3 and caused the new test to fail.
2018-01-03 19:51:14 +00:00
/* The stacking rules can't allocate registers. We use these
* scratch registers to stack tokens.
Add splitting coercions for IND_ALL_D. Delete my wrong comment (from commits cfbc537, a8f62f4, 5432bd0) which claimed that such coercions are not possible.
2017-12-19 01:59:04 +00:00
*/
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
#define RSCRATCH r0
#define FSCRATCH f0
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
TOKENS
/* Primitives */
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
C /* constant */ = { INT val; } 4 val.
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
LABEL = { ADDR adr; } 4 adr.
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
LABEL_HI = { ADDR adr; } 4 "hi16[" adr "]".
LABEL_HA = { ADDR adr; } 4 "ha16[" adr "]".
LABEL_LO = { ADDR adr; } 4 "lo16[" adr "]".
LOCAL = { INT off; } 4 ">>> BUG IN LOCAL".
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
DLOCAL = { INT off; } 8 ">>> BUG IN DLOCAL".
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Allows us to use regvar() to refer to registers */
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
REG_EXPR = { REG reg; } 4 reg.
FREG_EXPR = { FREG reg; } 8 reg.
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
FSREG_EXPR = { FSREG reg; } 4 reg.
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
/* Constants on the stack */
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
CONST_N8000 = { INT val; } 4 val.
CONST_N7FFF_N0001 = { INT val; } 4 val.
CONST_0000_7FFF = { INT val; } 4 val.
CONST_8000 = { INT val; } 4 val.
CONST_8001_FFFF = { INT val; } 4 val.
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
CONST_HI_ZR = { INT val; } 4 val.
CONST_HI_LO = { INT val; } 4 val.
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Expression partial results */
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
SEX_B = { GPR reg; } 4. /* sign extension */
SEX_H = { GPR reg; } 4.
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
SUM_RIS = { GPR reg; INT offhi; } 4. /* reg + (offhi << 16) */
SUM_RC = { GPR reg; INT off; } 4. /* reg + off */
SUM_RL = { GPR reg; ADDR adr; } 4. /* reg + lo16[adr] */
SUM_RR = { GPR reg1; GPR reg2; } 4. /* reg1 + reg2 */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
SUB_CR = { INT val; GPR reg; } 4. /* val - reg */
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
SUB_RR = { GPR reg1; GPR reg2; } 4. /* reg1 - reg2 */
NEG_R = { GPR reg; } 4. /* -reg */
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
MUL_RC = { GPR reg; INT val; } 4. /* reg * val */
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
MUL_RR = { GPR reg1; GPR reg2; } 4. /* reg1 * reg2 */
DIV_RR = { GPR reg1; GPR reg2; } 4. /* reg1 / reg2 signed */
DIV_RR_U = { GPR reg1; GPR reg2; } 4. /* reg1 / reg2 unsigned */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
/* Indirect loads and stores */
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
IND_RC_B = { GPR reg; INT off; } 4 off "(" reg ")".
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
IND_RL_B = { GPR reg; ADDR adr; } 4 "lo16[" adr "](" reg ")".
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
IND_RR_B = { GPR reg1; GPR reg2; } 4.
IND_RC_H = { GPR reg; INT off; } 4 off "(" reg ")".
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
IND_RL_H = { GPR reg; ADDR adr; } 4 "lo16[" adr "](" reg ")".
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
IND_RR_H = { GPR reg1; GPR reg2; } 4.
IND_RC_H_S = { GPR reg; INT off; } 4 off "(" reg ")".
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
IND_RL_H_S = { GPR reg; ADDR adr; } 4 "lo16[" adr "](" reg ")".
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
IND_RR_H_S = { GPR reg1; GPR reg2; } 4.
IND_RC_W = { GPR reg; INT off; } 4 off "(" reg ")".
IND_RL_W = { GPR reg; ADDR adr; } 4 "lo16[" adr "](" reg ")".
IND_RR_W = { GPR reg1; GPR reg2; } 4.
IND_RC_D = { GPR reg; INT off; } 8 off "(" reg ")".
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
IND_RL_D = { GPR reg; ADDR adr; } 8 "lo16[" adr "](" reg ")".
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
IND_RR_D = { GPR reg1; GPR reg2; } 8.
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
/* Local variables in frame */
FRAME_B = { INT level; GPR reg; INT off; INT size; }
4 off "(" reg ")".
FRAME_H = { INT level; GPR reg; INT off; INT size; }
4 off "(" reg ")".
FRAME_H_S = { INT level; GPR reg; INT off; INT size; }
4 off "(" reg ")".
FRAME_W = { INT level; GPR reg; INT off; INT size; }
4 off "(" reg ")".
FRAME_D = { INT level; GPR reg; INT off; INT size; }
8 off "(" reg ")".
LXFRAME = { INT level; } 4.
/* Bitwise logic */
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
NOT_R = { GPR reg; } 4. /* ~reg */
AND_RIS = { GPR reg; INT valhi; } 4.
AND_RC = { GPR reg; INT val; } 4.
AND_RR = { GPR reg1; GPR reg2; } 4.
ANDC_RR = { GPR reg1; GPR reg2; } 4. /* reg1 & ~reg2 */
OR_RIS = { GPR reg; INT valhi; } 4.
OR_RC = { GPR reg; INT val; } 4.
OR_RR = { GPR reg1; GPR reg2; } 4.
ORC_RR = { GPR reg1; GPR reg2; } 4. /* reg1 | ~reg2 */
XOR_RIS = { GPR reg; INT valhi; } 4.
XOR_RC = { GPR reg; INT val; } 4.
XOR_RR = { GPR reg1; GPR reg2; } 4.
NAND_RR = { GPR reg1; GPR reg2; } 4. /* ~(reg1 & reg2) */
NOR_RR = { GPR reg1; GPR reg2; } 4. /* ~(reg1 | reg2) */
EQV_RR = { GPR reg1; GPR reg2; } 4. /* ~(reg1 ^ reg2) */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
/* Comparisons */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
COND_RC = { GPR reg; INT val; } 4.
COND_RR = { GPR reg1; GPR reg2; } 4.
CONDL_RC = { GPR reg; INT val; } 4.
CONDL_RR = { GPR reg1; GPR reg2; } 4.
COND_FS = { FSREG reg1; FSREG reg2; } 4.
COND_FD = { FREG reg1; FREG reg2; } 4.
XEQ = { GPR reg; } 4.
XNE = { GPR reg; } 4.
XGT = { GPR reg; } 4.
XGE = { GPR reg; } 4.
XLT = { GPR reg; } 4.
XLE = { GPR reg; } 4.
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
SETS
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
/* signed 16-bit integer */
CONST2 = CONST_N8000 + CONST_N7FFF_N0001 + CONST_0000_7FFF.
/* integer that, when negated, fits signed 16-bit */
CONST2_WHEN_NEG = CONST_N7FFF_N0001 + CONST_0000_7FFF + CONST_8000.
/* unsigned 16-bit integer */
UCONST2 = CONST_0000_7FFF + CONST_8000 + CONST_8001_FFFF.
/* any constant on stack */
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
CONST_STACK = CONST_N8000 + CONST_N7FFF_N0001 + CONST_0000_7FFF +
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
CONST_8000 + CONST_8001_FFFF +
CONST_HI_ZR + CONST_HI_LO.
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
CONST = C + CONST_STACK.
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
SET_RC_B = IND_RC_B + IND_RL_B + FRAME_B.
SET_RC_H = IND_RC_H + IND_RL_H + FRAME_H.
SET_RC_H_S = IND_RC_H_S + IND_RL_H_S + FRAME_H_S.
SET_RC_W = IND_RC_W + IND_RL_W + FRAME_W.
SET_RC_D = IND_RC_D + IND_RL_D + FRAME_D.
IND_ALL_B = IND_RC_B + IND_RL_B + IND_RR_B.
IND_ALL_H = IND_RC_H + IND_RL_H + IND_RR_H +
IND_RC_H_S + IND_RL_H_S + IND_RR_H_S.
IND_ALL_W = IND_RC_W + IND_RL_W + IND_RR_W.
IND_ALL_D = IND_RC_D + IND_RL_D + IND_RR_D.
IND_V = IND_ALL_B + IND_ALL_H + IND_ALL_W + IND_ALL_D.
FRAME_V = FRAME_B + FRAME_H + FRAME_H_S + FRAME_W + FRAME_D.
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
/* anything killed by sti (store indirect) */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
MEMORY = IND_V + FRAME_V.
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
/* any integer from stack that we can easily move to GPR */
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
INT_W = SPFP + REG + CONST_STACK + SEX_B + SEX_H +
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
SUM_RIS + SUM_RC + SUM_RL + SUM_RR +
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
SUB_CR + SUB_RR + NEG_R +
MUL_RC + MUL_RR + DIV_RR + DIV_RR_U +
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
IND_ALL_B + IND_ALL_H + IND_ALL_W +
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
FRAME_B + FRAME_H + FRAME_H_S + FRAME_W +
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
NOT_R + AND_RIS + AND_RC + AND_RR + ANDC_RR +
OR_RIS + OR_RC + OR_RR + ORC_RR +
XOR_RIS + XOR_RC + XOR_RR + NAND_RR + NOR_RR + EQV_RR +
XEQ + XNE + XGT + XGE + XLT + XLE.
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
FLOAT_D = FREG + IND_ALL_D + FRAME_D.
FLOAT_W = FSREG + IND_ALL_W + FRAME_W.
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
INSTRUCTIONS
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
/* We give time as cycles of total latency from Freescale
* Semiconductor, MPC7450 RISC Microprocessor Family Reference
* Manual, Rev. 5, section 6.6.
*
* We have only 4-byte alignment for doubles; 8-byte alignment is
* optimal. We guess the misalignment penalty by adding 1 cycle to
* the cost of loading or storing a double:
* lfd lfdu lfdx: 4 -> 5
* stfd stfdu stfdx: 3 -> 4
*/
cost(4, 1) /* space, time */
Remove most of GPRE from mach/powerpc/ncg/table We only need GPRE in a few places where we write {GPRE, regvar(...)} because ncgg can't parse plain regvar(...). In all other places, a plain GPR works. Also remove gpr_gpr_gpr and a few other unused and fake instructions from the list of instructions.
2016-10-07 02:59:27 +00:00
add GPR:wo, GPR:ro, GPR:ro.
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
addX "add." GPR:wo:cc, GPR:ro, GPR:ro.
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
addi GPR:wo, GPR:ro, CONST+LABEL_LO:ro.
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
li GPR:wo, CONST:ro.
addis GPR:wo, GPR:ro, CONST+LABEL_HI+LABEL_HA:ro.
lis GPR:wo, CONST+LABEL_HI+LABEL_HA:ro.
Remove most of GPRE from mach/powerpc/ncg/table We only need GPRE in a few places where we write {GPRE, regvar(...)} because ncgg can't parse plain regvar(...). In all other places, a plain GPR works. Also remove gpr_gpr_gpr and a few other unused and fake instructions from the list of instructions.
2016-10-07 02:59:27 +00:00
and GPR:wo, GPR:ro, GPR:ro.
andc GPR:wo, GPR:ro, GPR:ro.
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
andiX "andi." GPR:wo:cc, GPR:ro, CONST:ro.
andisX "andis." GPR:wo:cc, GPR:ro, CONST:ro.
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
b LABEL:ro.
bc CONST:ro, CONST:ro, LABEL:ro.
Implement blm and bls using an inline loop. This replaces a call to memmove() in libc. That was working for me, but it can fail because EM programs don't always link to libc. blm and bls only need to copy aligned words. They don't need to copy bytes, and they don't need to copy between overlapping buffers, as memmove() does. So the new loop is simpler than memmove().
2017-02-12 00:30:12 +00:00
bdnz LABEL:ro.
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
beq LABEL:ro.
bne LABEL:ro.
bgt LABEL:ro.
bge LABEL:ro.
blt LABEL:ro.
ble LABEL:ro.
bxx LABEL:ro. /* dummy */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
bcctr CONST:ro, CONST:ro, CONST:ro.
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
bctr.
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
bcctrl CONST:ro, CONST:ro, CONST:ro.
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
bctrl.
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
bclr CONST:ro, CONST:ro, CONST:ro.
Remove .ret from libem and inline the code. This removes a wrong-way dependency of libsys on libem. The C functions in libsys called .ret, but libsys is after libem in the linker arguments, so the linker didn't find .ret unless something else had called .ret. Almost everything called .ret, but I got a linker error when I wrote an assembly program using the EM runtime, because my assembly program didn't call .ret. Add a dummy comment to build.lua, so git checkout touches that file, the build system reconfigures itself, and the *.s glob sees that ret.s has gone.
2017-02-17 02:18:39 +00:00
blr.
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
bl LABEL:ro.
These instructions write to the CR.
2017-12-10 19:01:14 +00:00
cmp CR:wo, CONST:ro, GPR:ro, GPR:ro kills :cc.
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
cmpw GPR:ro, GPR:ro kills :cc.
These instructions write to the CR.
2017-12-10 19:01:14 +00:00
cmpi CR:wo, CONST:ro, GPR:ro, CONST:ro kills :cc.
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
cmpwi GPR:ro, CONST:ro kills :cc.
These instructions write to the CR.
2017-12-10 19:01:14 +00:00
cmpl CR:wo, CONST:ro, GPR:ro, GPR:ro kills :cc.
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
cmplw GPR:ro, GPR:ro kills :cc.
These instructions write to the CR.
2017-12-10 19:01:14 +00:00
cmpli CR:wo, CONST:ro, GPR:ro, CONST:ro kills :cc.
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
cmplwi GPR:ro, CONST:ro kills :cc.
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
divw GPR:wo, GPR:ro, GPR:ro cost(4, 23).
divwu GPR:wo, GPR:ro, GPR:ro cost(4, 23).
Remove most of GPRE from mach/powerpc/ncg/table We only need GPRE in a few places where we write {GPRE, regvar(...)} because ncgg can't parse plain regvar(...). In all other places, a plain GPR works. Also remove gpr_gpr_gpr and a few other unused and fake instructions from the list of instructions.
2016-10-07 02:59:27 +00:00
eqv GPR:wo, GPR:ro, GPR:ro.
extsb GPR:wo, GPR:ro.
extsh GPR:wo, GPR:ro.
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
fadd FREG+DLOCAL:wo, FREG:ro, FREG:ro cost(4, 5).
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
fadds FSREG+LOCAL:wo, FSREG:ro, FSREG:ro cost(4, 5).
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
fcmpo CR:wo, FREG:ro, FREG:ro cost(4, 5).
fcmpo CR:wo, FSREG:ro, FSREG:ro cost(4, 5).
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
fctiwz FREG:wo, FREG:ro cost(4, 5).
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
fdiv FREG+DLOCAL:wo, FREG:ro, FREG:ro cost(4, 35).
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
fdivs FSREG+LOCAL:wo, FSREG:ro, FSREG:ro cost(4, 21).
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
fmr FPR:wo, FPR:ro cost(4, 5).
fmr FSREG:wo, FSREG:ro cost(4, 5).
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
fmul FREG+DLOCAL:wo, FREG:ro, FREG:ro cost(4, 5).
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
fmuls FSREG+LOCAL:wo, FSREG:ro, FSREG:ro cost(4, 5).
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
fneg FREG+DLOCAL:wo, FREG:ro cost(4, 5).
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
fneg FSREG+LOCAL:wo, FSREG:ro cost(4, 5).
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
frsp FSREG+LOCAL:wo, FREG:ro cost(4, 5).
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
fsub FREG+DLOCAL:wo, FREG:ro, FREG:ro cost(4, 5).
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
fsubs FSREG+LOCAL:wo, FSREG:ro, FSREG:ro cost(4, 5).
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
lbz GPR:wo, SET_RC_B:ro cost(4, 3).
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
lbzx GPR:wo, GPR:ro, GPR:ro cost(4, 3).
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
lfd FPR+DLOCAL:wo, SET_RC_D:ro cost(4, 5).
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
lfdu FPR:wo, IND_RC_D:rw cost(4, 5).
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
lfdx FPR:wo, GPR:ro, GPR:ro cost(4, 5).
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
lfs FSREG+LOCAL:wo, SET_RC_W:ro cost(4, 4).
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
lfsu FSREG:wo, IND_RC_W:rw cost(4, 4).
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
lfsx FSREG:wo, GPR:ro, GPR:ro cost(4, 4).
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
lha GPR:wo, SET_RC_H_S:ro cost(4, 3).
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
lhax GPR:wo, GPR:ro, GPR:ro cost(4, 3).
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
lhz GPR:wo, SET_RC_H:ro cost(4, 3).
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
lhzx GPR:wo, GPR:ro, GPR:ro cost(4, 3).
Teach top to merge or delete "addi sp, sp, X". This reduces code size, because ncg emits too many "addi sp, sp, X" instructions when unstacking things. Now top lowers "addi sp, sp, X" by lifting other instructions. This sometimes creates chances to merge or delete _addi_ instructions. If no such chance is found, the _addi_ remains uselessly lowered. Edit ncg/table to remove something that top now does. Edit ncg/mach.c to remove some spaces after commas. This removes a whitespace difference between *.s and *.so files, because top removes the space.
2018-01-05 22:55:50 +00:00
lwz GPR+LOCAL:wo, SET_RC_W:ro cost(4, 3).
lwzu writes to the register in the token.
2017-10-19 16:44:46 +00:00
lwzu GPR:wo, IND_RC_W:rw cost(4, 3).
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
lwzx GPR:wo, GPR:ro, GPR:ro cost(4, 3).
mfcr GPR:wo cost(4,2).
mfspr GPR:wo, SPR:ro cost(4, 3).
mtspr SPR:wo, GPR:ro cost(4, 2).
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
mulli GPR:wo, GPR:ro, CONST:ro cost(4, 3).
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
mullw GPR:wo, GPR:ro, GPR:ro cost(4, 4).
nand GPR:wo, GPR:ro, GPR:ro.
neg GPR:wo, GPR:ro.
nor GPR:wo, GPR:ro, GPR:ro.
Remove most of GPRE from mach/powerpc/ncg/table We only need GPRE in a few places where we write {GPRE, regvar(...)} because ncgg can't parse plain regvar(...). In all other places, a plain GPR works. Also remove gpr_gpr_gpr and a few other unused and fake instructions from the list of instructions.
2016-10-07 02:59:27 +00:00
or GPR:wo, GPR:ro, GPR:ro.
Use "mr" and make a few other tweaks in PowerPC ncg table. Use extended "mr" instead of basic "or" to move registers. Both "mr" and "or" encode the same machine instruction. With "mr", I can more easily search the assembly output for register moves. Fold several stacking rules into a single rule ANY_BHW-REG to STACK. Remove the EM patterns for loc mlu $2==2 and loc slu. The first pattern had the wrong size (should be $2==4, not $2==2). Both patterns were redundant. They rewrote loc mlu as loc mli and loc slu as loc sli, but this table doesn't have patterns for loc mli or loc sli, so it is enough to rewrite mlu as mli and slu as sli.
2017-02-10 16:45:50 +00:00
mr GPR:wo, GPR:ro.
orX "or." GPR:wo:cc, GPR:ro, GPR:ro.
Optimize `mr. X, X` after some instructions. For example, when ncg emits slw r9,r8,r5 mr. r9,r9 then top simplifies the code to slw. r9,r8,r5
2017-12-23 03:32:16 +00:00
mrX_readonly "mr." GPR:ro:cc, GPR:ro.
Remove most of GPRE from mach/powerpc/ncg/table We only need GPRE in a few places where we write {GPRE, regvar(...)} because ncgg can't parse plain regvar(...). In all other places, a plain GPR works. Also remove gpr_gpr_gpr and a few other unused and fake instructions from the list of instructions.
2016-10-07 02:59:27 +00:00
orc GPR:wo, GPR:ro, GPR:ro.
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
ori GPR:wo, GPR:ro, CONST+LABEL_LO:ro.
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
oris GPR:wo, GPR:ro, CONST:ro.
Remove most of GPRE from mach/powerpc/ncg/table We only need GPRE in a few places where we write {GPRE, regvar(...)} because ncgg can't parse plain regvar(...). In all other places, a plain GPR works. Also remove gpr_gpr_gpr and a few other unused and fake instructions from the list of instructions.
2016-10-07 02:59:27 +00:00
rlwinm GPR:wo, GPR:ro, CONST:ro, CONST:ro, CONST:ro.
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
extlwi GPR:wo, GPR:ro, CONST:ro, CONST:ro.
extrwi GPR:wo, GPR:ro, CONST:ro, CONST:ro.
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
rotlwi GPR+LOCAL:wo, GPR:ro, CONST:ro.
rotrwi GPR+LOCAL:wo, GPR:ro, CONST:ro.
slwi GPR+LOCAL:wo, GPR:ro, CONST:ro.
srwi GPR+LOCAL:wo, GPR:ro, CONST:ro.
Add test for EM rol, ror. Fix i80, i86, powerpc. EM instructions _rol_ and _ror_ do rotate an integer left or right. Our compilers and optimizers never emit _rol_ nor _ror_, but I might want to use them in the future. Add _rol_ and _ror_ to powerpc. Fix `rol 4` and `ror 4` in both i80 and i86, where the rules for `rol 4` and `ror 4` seem to have never been tested until now.
2017-12-07 22:16:21 +00:00
rlwnm GPR:wo, GPR:ro, GPR:ro, CONST:ro, CONST:ro.
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
rotlw GPR+LOCAL:wo, GPR:ro, GPR:ro.
slw GPR+LOCAL:wo, GPR:ro, GPR:ro.
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
sraw GPR+LOCAL:wo, GPR:ro, GPR:ro /* kills xer */ cost(4, 2).
srawi GPR+LOCAL:wo, GPR:ro, CONST:ro /* kills xer */ cost(4, 2).
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
srw GPR+LOCAL:wo, GPR:ro, GPR:ro.
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
stb GPR:ro, SET_RC_B:rw cost(4, 3).
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
stbx GPR:ro, GPR:ro, GPR:ro cost(4, 3).
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
stfd FPR:ro, SET_RC_D:rw cost(4, 4).
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
stfdu FPR:ro, IND_RC_D:rw cost(4, 4).
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
stfdx FPR:ro, GPR:ro, GPR:ro cost(4, 4).
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
stfs FSREG:ro, SET_RC_W:rw cost(4, 3).
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
stfsu FSREG:ro, IND_RC_W:rw cost(4, 3).
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
stfsx FSREG:ro, GPR:ro, GPR:ro cost(4, 3).
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
sth GPR:ro, SET_RC_H:rw cost(4, 3).
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
sthx GPR:ro, GPR:ro, GPR:ro cost(4, 3).
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
stw GPR:ro, SET_RC_W:rw cost(4, 3).
Add costs to powerpc instructions. Also show how andi., andis., or., set condition codes.
2016-10-17 18:57:21 +00:00
stwx GPR:ro, GPR:ro, GPR:ro cost(4, 3).
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
stwu GPR:ro, IND_RC_W:rw cost(4, 3).
Teach top to merge or delete "addi sp, sp, X". This reduces code size, because ncg emits too many "addi sp, sp, X" instructions when unstacking things. Now top lowers "addi sp, sp, X" by lifting other instructions. This sometimes creates chances to merge or delete _addi_ instructions. If no such chance is found, the _addi_ remains uselessly lowered. Edit ncg/table to remove something that top now does. Edit ncg/mach.c to remove some spaces after commas. This removes a whitespace difference between *.s and *.so files, because top removes the space.
2018-01-05 22:55:50 +00:00
subf GPR:wo, GPR:ro, GPR:ro.
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
subfic GPR:wo, GPR:ro, CONST:ro /* kills xer */.
Remove most of GPRE from mach/powerpc/ncg/table We only need GPRE in a few places where we write {GPRE, regvar(...)} because ncgg can't parse plain regvar(...). In all other places, a plain GPR works. Also remove gpr_gpr_gpr and a few other unused and fake instructions from the list of instructions.
2016-10-07 02:59:27 +00:00
xor GPR:wo, GPR:ro, GPR:ro.
xori GPR:wo, GPR:ro, CONST:ro.
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
xoris GPR:wo, GPR:ro, CONST:ro.
Remove the "invalid" stacking rule. When ncg fell back on this rule, it did emit the string "invalid" in the assembly code and caused a syntax error in the assembler. Adjust the stacking rules so we can stack LOCAL, CONST, and LABEL without falling back on the "invalid" rule, and so we can stack them when we have no free register except the scratch register.
2016-09-27 20:46:11 +00:00
PowerPC ncg never uses the rules to stack LOCAL or DLOCAL.
2017-10-15 19:22:52 +00:00
bug ">>> BUG" LABEL:ro cost(0, 0).
Add assembler support for fixing up arbitrary oris/addi pairs of instructions; this should allow oris/lwz constant value loads, which will save an opcode.
2017-01-14 23:15:01 +00:00
comment "!" LABEL:ro cost(0, 0).
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
MOVES
from GPR to GPR
Use "mr" and make a few other tweaks in PowerPC ncg table. Use extended "mr" instead of basic "or" to move registers. Both "mr" and "or" encode the same machine instruction. With "mr", I can more easily search the assembly output for register moves. Fold several stacking rules into a single rule ANY_BHW-REG to STACK. Remove the EM patterns for loc mlu $2==2 and loc slu. The first pattern had the wrong size (should be $2==4, not $2==2). Both patterns were redundant. They rewrote loc mlu as loc mli and loc slu as loc sli, but this table doesn't have patterns for loc mli or loc sli, so it is enough to rewrite mlu as mli and slu as sli.
2017-02-10 16:45:50 +00:00
gen mr %2, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add reglap to ncg. Add 4-byte reg_float to PowerPC ncg. The new feature "reglap" allows two sizes of floating-point register variables (reg_float), if each register overlaps a single register of the other size. PowerPC ncg uses reglap to define 4-byte instances of f14 to f31 that overlap the 8-byte instances. When ncgg sees the definition of fs14("f14")=f14, it removes the 8-byte f14 from its rvnumbers array, and adds the 4-byte fs14 in its place. Later, when ncg puts a variable in fs14, if it is an 8-byte variable, then ncg switches to the 8-byte f14. The code has /* reglap */ comments in util/ncgg or #ifdef REGLAP in mach/proto/ncg reglap became necessary because my commit a20b87c caused PowerPC ego to allocate reg_float in both 4-byte and 8-byte sizes.
2017-10-14 16:40:04 +00:00
from FSREG to FSREG
gen fmr %2, %1
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
from FPR to FPR
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
gen fmr %2, %1
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
/* Constants */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
from CONST smalls(%val) to GPR
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
COMMENT("move CONST->GPR smalls")
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
li %2, %1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
from CONST lo(%val)==0 to GPR
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
COMMENT("move CONST->GPR shifted")
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
lis %2, {C, hi(%1.val)}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
from CONST to GPR
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
COMMENT("move CONST->GPR")
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
lis %2, {C, hi(%1.val)}
ori %2, %2, {C, lo(%1.val)}
/* Can't use addi %2, %2, {C, los(%1.val)}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
* because %2 might be R0. */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from LABEL to GPR
gen
COMMENT("move LABEL->GPR")
Tweak some tokens in PowerPC ncg. Remove the GPRINDIRECT token, and use the IND_RC_* tokens as operands to instructions. We no longer need to unpack an IND_RC_* token and repack it as a GPRINDIRECT to use it in an instruction. Allow storing IND_ALL_B and IND_ALL_H in register variables. Create a set ANY_BHW for anything that we can store in a regvar. Push register variables on the stack without using GPRE, by changing stwu to accept LOCAL. Then ncg will replace the string ">>> BUG IN LOCAL" with the register name. (I copied ">>> BUG IN LOCAL" from mach/arm/ncg/table.) Fix the rule for "pat lil inreg($1)>0" to yield a IND_RC_W token, not a register. We might need to kill the token with "kills MEMORY". Rename CONST_ALL to CONST_STACK, because it only includes constants on the stack, and excludes CONST tokens. Instructions still don't allow CONST_STACK operands, so we still need to repack each CONST_STACK as a CONST to use it in an instruction. Rename LABEL_OFFSET_HI to just LABEL_HI, and same for LABEL_HA and LABEL_HO.
2017-02-08 17:12:28 +00:00
lis %2, {LABEL_HI, %1.adr}
ori %2, %2, {LABEL_LO, %1.adr}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
from LABEL_HA to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lis %2, %1
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Sign extension */
from SEX_B to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen extsb %2, %1.reg
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from SEX_H to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen extsh %2, %1.reg
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Register + something */
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
from SUM_RIS to GPR
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen addis %2, %1.reg, {C, %1.offhi}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from SUM_RC to GPR
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen addi %2, %1.reg, {C, %1.off}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
from SUM_RL to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen addi %2, %1.reg, {LABEL_LO, %1.adr}
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from SUM_RR to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen add %2, %1.reg1, %1.reg2
Remove most of GPRE from mach/powerpc/ncg/table We only need GPRE in a few places where we write {GPRE, regvar(...)} because ncgg can't parse plain regvar(...). In all other places, a plain GPR works. Also remove gpr_gpr_gpr and a few other unused and fake instructions from the list of instructions.
2016-10-07 02:59:27 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
/* Other arithmetic */
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
from SUB_CR to GPR
/* val - reg -> subtract reg from val */
gen subfic %2, %1.reg, {C, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
from SUB_RR to GPR
/* reg1 - reg2 -> subtract reg2 from reg1 */
gen subf %2, %1.reg2, %1.reg1
from NEG_R to GPR
gen neg %2, %1.reg
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
from MUL_RC to GPR
gen mulli %2, %1.reg, {C, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
from MUL_RR to GPR
gen mullw %2, %1.reg1, %1.reg2
from DIV_RR to GPR
gen divw %2, %1.reg1, %1.reg2
from DIV_RR_U to GPR
gen divwu %2, %1.reg1, %1.reg2
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
/* Read byte */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from SET_RC_B to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lbz %2, %1
Simplify moves to and from IND_RC_* Now that SUM_RC always has a signed 16-bit constant, it happens that the various IND_RC_* tokens also have a signed 16-bit constant, so we no longer need to touch the scratch register.
2016-10-16 20:02:25 +00:00
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
from IND_RR_B to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lbzx %2, %1.reg1, %1.reg2
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
/* Write byte */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from GPR to SET_RC_B
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen stb %1, %2
Simplify moves to and from IND_RC_* Now that SUM_RC always has a signed 16-bit constant, it happens that the various IND_RC_* tokens also have a signed 16-bit constant, so we no longer need to touch the scratch register.
2016-10-16 20:02:25 +00:00
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
from GPR to IND_RR_B
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen stbx %1, %2.reg1, %2.reg2
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
/* Read halfword (short) */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from SET_RC_H to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lhz %2, %1
Simplify moves to and from IND_RC_* Now that SUM_RC always has a signed 16-bit constant, it happens that the various IND_RC_* tokens also have a signed 16-bit constant, so we no longer need to touch the scratch register.
2016-10-16 20:02:25 +00:00
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
from IND_RR_H to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lhzx %2, %1.reg1, %1.reg2
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from SET_RC_H_S to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lha %2, %1
Simplify moves to and from IND_RC_* Now that SUM_RC always has a signed 16-bit constant, it happens that the various IND_RC_* tokens also have a signed 16-bit constant, so we no longer need to touch the scratch register.
2016-10-16 20:02:25 +00:00
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
from IND_RR_H_S to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lhax %2, %1.reg1, %1.reg2
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
/* Write halfword */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from GPR to SET_RC_H
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen sth %1, %2
Simplify moves to and from IND_RC_* Now that SUM_RC always has a signed 16-bit constant, it happens that the various IND_RC_* tokens also have a signed 16-bit constant, so we no longer need to touch the scratch register.
2016-10-16 20:02:25 +00:00
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
from GPR to IND_RR_H
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen sthx %1, %2.reg1, %2.reg2
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Read word */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from SET_RC_W to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lwz %2, %1
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from IND_RR_W to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lwzx %2, %1.reg1, %1.reg2
Simplify moves to and from IND_RC_* Now that SUM_RC always has a signed 16-bit constant, it happens that the various IND_RC_* tokens also have a signed 16-bit constant, so we no longer need to touch the scratch register.
2016-10-16 20:02:25 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from SET_RC_W to FSREG
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lfs %2, %1
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
from IND_RR_W to FSREG
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lfsx %2, %1.reg1, %1.reg2
Remove IND_LABEL_W and IND_LABEL_D Because li32 always loads a label into a GPR, it is sufficient to coerce LABEL to REG, then use IND_RC_W or IND_RC_D for indirection through the label.
2016-10-16 20:33:24 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Write word */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from GPR to SET_RC_W
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen stw %1, %2
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from GPR to IND_RR_W
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen stwx %1, %2.reg1, %2.reg2
Simplify moves to and from IND_RC_* Now that SUM_RC always has a signed 16-bit constant, it happens that the various IND_RC_* tokens also have a signed 16-bit constant, so we no longer need to touch the scratch register.
2016-10-16 20:02:25 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from FSREG to SET_RC_W
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen stfs %1, %2
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
from FSREG to IND_RR_W
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen stfsx %1, %2.reg1, %2.reg2
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Read double */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from SET_RC_D to FPR
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
gen lfd %2, %1
Simplify moves to and from IND_RC_* Now that SUM_RC always has a signed 16-bit constant, it happens that the various IND_RC_* tokens also have a signed 16-bit constant, so we no longer need to touch the scratch register.
2016-10-16 20:02:25 +00:00
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
from IND_RR_D to FPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen lfdx %2, %1.reg1, %1.reg2
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Write double */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from FPR to SET_RC_D
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
gen stfd %1, %2
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
from FPR to IND_RR_D
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen stfdx %1, %2.reg1, %2.reg2
Remove IND_LABEL_W and IND_LABEL_D Because li32 always loads a label into a GPR, it is sufficient to coerce LABEL to REG, then use IND_RC_W or IND_RC_D for indirection through the label.
2016-10-16 20:33:24 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
/* LXFRAME is a lexical frame from the static chain. We define a move
so "uses REG={LXFRAME, $1}" may find a register with the same
frame, and not repeat the move. This move can't search for a REG
with {LXFRAME, $1-1}, but must always start from fp. The static
chain, if it exists, is the argument at fp + EM_BSIZE. */
from LXFRAME %level==1 to REG
gen lwz %2, {IND_RC_W, fp, EM_BSIZE}
from LXFRAME %level==2 to REG
gen lwz %2, {IND_RC_W, fp, EM_BSIZE}
/* PowerPC can't add r0 + EM_BSIZE,
* so %2 must not be r0. */
lwz %2, {IND_RC_W, %2, EM_BSIZE}
from LXFRAME %level==3 to REG
gen lwz %2, {IND_RC_W, fp, EM_BSIZE}
lwz %2, {IND_RC_W, %2, EM_BSIZE}
lwz %2, {IND_RC_W, %2, EM_BSIZE}
from LXFRAME %level==4 to REG
gen lwz %2, {IND_RC_W, fp, EM_BSIZE}
lwz %2, {IND_RC_W, %2, EM_BSIZE}
lwz %2, {IND_RC_W, %2, EM_BSIZE}
lwz %2, {IND_RC_W, %2, EM_BSIZE}
from LXFRAME to REG /* assuming %level in 2 to 0x8000 */
gen li %2, {C, %1.level-1}
mtspr ctr, %2
lwz %2, {IND_RC_W, fp, EM_BSIZE}
1: lwz %2, {IND_RC_W, %2, EM_BSIZE}
bdnz {LABEL, "1b"}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Logicals */
from NOT_R to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen nor %2, %1.reg, %1.reg
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
from AND_RIS to GPR
gen andisX %2, %1.reg, {C, %1.valhi}
from AND_RC to GPR
gen andiX %2, %1.reg, {C, %1.val}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from AND_RR to GPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen and %2, %1.reg1, %1.reg2
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
from ANDC_RR to GPR
gen andc %2, %1.reg1, %1.reg2
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
from OR_RIS to GPR
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen oris %2, %1.reg, {C, %1.valhi}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from OR_RC to GPR
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen ori %2, %1.reg, {C, %1.val}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
from OR_RR to GPR
gen or %2, %1.reg1, %1.reg2
from ORC_RR to GPR
gen orc %2, %1.reg1, %1.reg2
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
from XOR_RIS to GPR
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen xoris %2, %1.reg, {C, %1.valhi}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from XOR_RC to GPR
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen xori %2, %1.reg, {C, %1.val}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
from XOR_RR to GPR
gen xor %2, %1.reg1, %1.reg2
from NAND_RR to GPR
gen nand %2, %1.reg1, %1.reg2
from NOR_RR to GPR
gen nor %2, %1.reg1, %1.reg2
from EQV_RR to GPR
gen eqv %2, %1.reg1, %1.reg2
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
/* Conditions */
/* Compare values, then copy cr0 to GPR. */
from COND_RC to GPR
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
cmpwi %1.reg, {C, %1.val}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
mfcr %2
from COND_RR to GPR
gen
cmpw %1.reg1, %1.reg2
mfcr %2
from CONDL_RC to GPR
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
cmplwi %1.reg, {C, %1.val}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
mfcr %2
from CONDL_RR to GPR
gen
cmplw %1.reg1, %1.reg2
mfcr %2
from COND_FS to GPR
gen
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
fcmpo cr0, %1.reg1, %1.reg2
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
mfcr %2
from COND_FD to GPR
gen
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
fcmpo cr0, %1.reg1, %1.reg2
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
mfcr %2
/* Given a copy of cr0 in %1.reg, extract a condition bit
* (lt, gt, eq) and perhaps flip it.
*/
from XEQ to GPR
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
extrwi %2, %1.reg, {C, 1}, {C, 2}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
from XNE to GPR
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
extrwi %2, %1.reg, {C, 1}, {C, 2}
xori %2, %2, {C, 1}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
from XGT to GPR
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
extrwi %2, %1.reg, {C, 1}, {C, 1}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
from XGE to GPR
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
extrwi %2, %1.reg, {C, 1}, {C, 0}
xori %2, %2, {C, 1}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
from XLT to GPR
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
extrwi %2, %1.reg, {C, 1}, {C, 0}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
from XLE to GPR
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
extrwi %2, %1.reg, {C, 1}, {C, 1}
xori %2, %2, {C, 1}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
/* REG_EXPR exists solely to allow us to use regvar() (which can only
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
be used in an expression) as a register constant. We can then use
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
our moves to GPR or REG to set register variables. This is easier
than defining moves to LOCAL, and avoids confusion between GPR and
FSREG in LOCAL. */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
from INT_W + LXFRAME to REG_EXPR
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
gen move %1, %2.reg
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
from FLOAT_D to FREG_EXPR
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
gen move %1, %2.reg
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
from FLOAT_W to FSREG_EXPR
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %1, %2.reg
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
TESTS
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Optimize `mr. X, X` after some instructions. For example, when ncg emits slw r9,r8,r5 mr. r9,r9 then top simplifies the code to slw. r9,r8,r5
2017-12-23 03:32:16 +00:00
/* Given "mrX %1, %1", ncgg would say, "Instruction destroys
* %1, not allowed here". We use mrX_readonly to trick ncgg.
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
*/
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
to test GPR
gen
Optimize `mr. X, X` after some instructions. For example, when ncg emits slw r9,r8,r5 mr. r9,r9 then top simplifies the code to slw. r9,r8,r5
2017-12-23 03:32:16 +00:00
mrX_readonly %1, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
STACKINGRULES
Remove the "invalid" stacking rule. When ncg fell back on this rule, it did emit the string "invalid" in the assembly code and caused a syntax error in the assembler. Adjust the stacking rules so we can stack LOCAL, CONST, and LABEL without falling back on the "invalid" rule, and so we can stack them when we have no free register except the scratch register.
2016-09-27 20:46:11 +00:00
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
from SPFP+REG to STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
COMMENT("stack SPFP+REG")
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
stwu %1, {IND_RC_W, sp, 0-4}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
from INT_W-SPFP-REG to STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
COMMENT("stack INT_W-SPFP-REG")
Rename SCRATCH to RSCRATCH. Never stack RSCRATCH nor FSCRATCH. Rename the scratch gpr (currently r11) from SCRATCH to RSCRATCH so I can search for RSCRATCH without finding FSCRATCH. I also want to avoid confusion with the SCRATCH keyword of the old code generator (cg which came before ncg). Change the stacking rules to prevent stacking of RSCRATCH or FSCRATCH or any other GPR or FPR that isn't an allocatable REG or FREG. Then ncgg rejects any rule that tries to stack a GPR or FPR, so change such rules to stack a REG or FREG.
2016-10-07 00:47:42 +00:00
move %1, RSCRATCH
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
stwu RSCRATCH, {IND_RC_W, sp, 0-4}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
from FLOAT_D-FREG to STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
COMMENT("stack FLOAT_D-FREG")
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
move %1, FSCRATCH
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
stfdu FSCRATCH, {IND_RC_D, sp, 0-8}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
PowerPC ncg never uses the rules to stack LOCAL or DLOCAL.
2017-10-15 19:22:52 +00:00
from FREG to STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
COMMENT("stack FREG")
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
stfdu %1, {IND_RC_D, sp, 0-8}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
from FSREG to STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
COMMENT("stack FSREG")
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
stfsu %1, {IND_RC_W, sp, 0-4}
Remove the "invalid" stacking rule. When ncg fell back on this rule, it did emit the string "invalid" in the assembly code and caused a syntax error in the assembler. Adjust the stacking rules so we can stack LOCAL, CONST, and LABEL without falling back on the "invalid" rule, and so we can stack them when we have no free register except the scratch register.
2016-09-27 20:46:11 +00:00
PowerPC ncg never uses the rules to stack LOCAL or DLOCAL.
2017-10-15 19:22:52 +00:00
/*
* We never stack LOCAL or DLOCAL tokens, because we only use
* them for register variables, so ncg pushes the register,
* not the token. These rules only prevent an error in ncgg.
*/
from LOCAL to STACK
gen bug {LABEL, "STACKING LOCAL"}
from DLOCAL to STACK
gen bug {LABEL, "STACKING DLOCAL"}
Remove the "invalid" stacking rule. When ncg fell back on this rule, it did emit the string "invalid" in the assembly code and caused a syntax error in the assembler. Adjust the stacking rules so we can stack LOCAL, CONST, and LABEL without falling back on the "invalid" rule, and so we can stack them when we have no free register except the scratch register.
2016-09-27 20:46:11 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
COERCIONS
Teach top to merge or delete "addi sp, sp, X". This reduces code size, because ncg emits too many "addi sp, sp, X" instructions when unstacking things. Now top lowers "addi sp, sp, X" by lifting other instructions. This sometimes creates chances to merge or delete _addi_ instructions. If no such chance is found, the _addi_ remains uselessly lowered. Edit ncg/table to remove something that top now does. Edit ncg/mach.c to remove some spaces after commas. This removes a whitespace difference between *.s and *.so files, because top removes the space.
2018-01-05 22:55:50 +00:00
/* The unstacking coercions emit many "addi sp, sp, X"
* instructions; the target optimizer (top) will merge them.
*/
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from STACK
uses REG
gen
COMMENT("coerce STACK->REG")
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
lwz %a, {IND_RC_W, sp, 0}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
addi sp, sp, {C, 4}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields %a
In powerpc ncg, add a speed hack for sti 8. ncg is too slow with this many registers. A stack pattern "with GPR GPR GPR" or "with REG REG REG" takes too long to pick registers, causing ncg 8 to take about 2 seconds on each sti 8. I introduce REG_PAIR and there are only 4 such pairs. For programs that use sti 8 (including C programs that copy 8-byte structs), this speed hack improves the ncg run from several seconds to almost instantaneous. Also add a few COMMENT(...) lines in stacking rules.
2016-10-18 00:31:59 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from STACK
uses FREG
gen
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
COMMENT("coerce STACK->FREG")
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
lfd %a, {IND_RC_D, sp, 0}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
addi sp, sp, {C, 8}
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
from STACK
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
uses FSREG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
COMMENT("coerce STACK->FSREG")
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
lfs %a, {IND_RC_W, sp, 0}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
addi sp, sp, {C, 4}
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
In powerpc ncg, add a speed hack for sti 8. ncg is too slow with this many registers. A stack pattern "with GPR GPR GPR" or "with REG REG REG" takes too long to pick registers, causing ncg 8 to take about 2 seconds on each sti 8. I introduce REG_PAIR and there are only 4 such pairs. For programs that use sti 8 (including C programs that copy 8-byte structs), this speed hack improves the ncg run from several seconds to almost instantaneous. Also add a few COMMENT(...) lines in stacking rules.
2016-10-18 00:31:59 +00:00
In coercions, try to reuse a register with the same token. This reduces code size.
2017-12-17 17:45:27 +00:00
/* "uses REG=%1" may find and reuse a register containing the
* same token. For contrast, "uses REG gen move %1, %a" would
* pick a different register before doing the move.
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
*
* "reusing %1" helps when coercing an INT_W token like
* {SUM_RC, r3, 0-4} to REG3, by not stacking the token.
*/
In coercions, try to reuse a register with the same token. This reduces code size.
2017-12-17 17:45:27 +00:00
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
from INT_W
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, REG=%1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
In powerpc ncg, add a speed hack for sti 8. ncg is too slow with this many registers. A stack pattern "with GPR GPR GPR" or "with REG REG REG" takes too long to pick registers, causing ncg 8 to take about 2 seconds on each sti 8. I introduce REG_PAIR and there are only 4 such pairs. For programs that use sti 8 (including C programs that copy 8-byte structs), this speed hack improves the ncg run from several seconds to almost instantaneous. Also add a few COMMENT(...) lines in stacking rules.
2016-10-18 00:31:59 +00:00
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
from FLOAT_D
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, FREG=%1
Do a move when coercing FREG to FREG or FSREG to FSREG.
2017-10-16 16:07:55 +00:00
yields %a
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
from FLOAT_W
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, FSREG=%1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
In powerpc ncg, add a speed hack for sti 8. ncg is too slow with this many registers. A stack pattern "with GPR GPR GPR" or "with REG REG REG" takes too long to pick registers, causing ncg 8 to take about 2 seconds on each sti 8. I introduce REG_PAIR and there are only 4 such pairs. For programs that use sti 8 (including C programs that copy 8-byte structs), this speed hack improves the ncg run from several seconds to almost instantaneous. Also add a few COMMENT(...) lines in stacking rules.
2016-10-18 00:31:59 +00:00
Add splitting coercions for IND_ALL_D. Delete my wrong comment (from commits cfbc537, a8f62f4, 5432bd0) which claimed that such coercions are not possible.
2017-12-19 01:59:04 +00:00
/* Splitting coercions can't allocate registers.
* PowerPC can't add r0 + constant. Use r12.
*/
from IND_RC_D %off<=0x7FFA
yields
{IND_RC_W, %1.reg, %1.off+4}
{IND_RC_W, %1.reg, %1.off}
from IND_RC_D
/* Don't move to %1.reg; it might be a regvar. */
gen move {SUM_RC, %1.reg, %1.off}, r12
yields {IND_RC_W, r12, 4} {IND_RC_W, r12, 0}
from IND_RR_D
gen move {SUM_RR, %1.reg1, %1.reg2}, r12
yields {IND_RC_W, r12, 4} {IND_RC_W, r12, 0}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
from FRAME_D %off<=0x7FFA
yields
{FRAME_W, %1.level, %1.reg, %1.off+4, 4}
{FRAME_W, %1.level, %1.reg, %1.off, 4}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
PATTERNS
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
/* Constants */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
pat loc $1==(0-0x8000) /* Load constant */
yields {CONST_N8000, $1}
pat loc $1>=(0-0x7FFF) && $1<=(0-1)
yields {CONST_N7FFF_N0001, $1}
pat loc $1>=0 && $1<=0x7FFF
yields {CONST_0000_7FFF, $1}
pat loc $1==0x8000
yields {CONST_8000, $1}
pat loc $1>=0x8001 && $1<=0xFFFF
yields {CONST_8001_FFFF, $1}
pat loc lo($1)==0
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
yields {CONST_HI_ZR, $1}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
pat loc
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
yields {CONST_HI_LO, $1}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
/* Stack shuffles */
/* The peephole optimizer does: loc $1 ass 4 -> asp $1
* To optimize multiplication, it uses: dup 8 asp 4
*/
pat asp $1==4 /* Adjust stack by constant */
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
with exact INT_W+FLOAT_W
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
/* drop %1 */
with STACK
gen addi sp, sp, {C, 4}
pat asp smalls($1)
with STACK
gen addi sp, sp, {C, $1}
pat asp lo($1)==0
with STACK
gen addi sp, sp, {C, hi($1)}
pat asp
with STACK
gen
addis sp, sp, {C, his($1)}
addi sp, sp, {C, los($1)}
pat ass $1==4 /* Adjust stack by variable */
with REG STACK
gen add sp, sp, %1
/* To duplicate a token, we coerce the token into a register,
* then duplicate the register. This decreases code size.
*/
pat dup $1==4 /* Duplicate word on top of stack */
with REG+FSREG
Add a rule for sdl ldl $1==$2 to work around a bug. In our powerpc table, sdl fails to kill the old value of the local. This is a bug, because a later ldl can load the old value instead of the newly stored value. By rewriting "sdl 0" "ldl 0" as "dup 8" "sdl 0", the newly added rule works around the bug, but only when the ldl is immediately after the sdl. This rule improves code that uses double-precision floating point. The output of printf("%f", 6.0) in C changes from all zero digits to "6000000" but still doesn't print the decimal point. The result of atof("-123.456") becomes correct. In startrek, I can now move the Enterprise, but I still can't fire phasers without crashing the game. We already have a rule for stl lol $1==$2. We had two copies of the rule, so I am deleting the second copy.
2016-09-30 15:50:50 +00:00
yields %1 %1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat dup $1==8 /* Duplicate double-word */
with REG+FSREG REG+FSREG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields %2 %1 %2 %1
Add a rule for sdl ldl $1==$2 to work around a bug. In our powerpc table, sdl fails to kill the old value of the local. This is a bug, because a later ldl can load the old value instead of the newly stored value. By rewriting "sdl 0" "ldl 0" as "dup 8" "sdl 0", the newly added rule works around the bug, but only when the ldl is immediately after the sdl. This rule improves code that uses double-precision floating point. The output of printf("%f", 6.0) in C changes from all zero digits to "6000000" but still doesn't print the decimal point. The result of atof("-123.456") becomes correct. In startrek, I can now move the Enterprise, but I still can't fire phasers without crashing the game. We already have a rule for stl lol $1==$2. We had two copies of the rule, so I am deleting the second copy.
2016-09-30 15:50:50 +00:00
with FREG
yields %1 %1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat dup /* Duplicate other size */
leaving
loc $1
dus 4
pat dus $1==4 /* Duplicate variable size */
with REG STACK
/* ( a size%1 -- a a ) */
uses REG, REG
gen
srwi %a, %1, {C, 2}
mtspr ctr, %a
add %b, sp, %1
1: lwzu %a, {IND_RC_W, %b, 0-4}
stwu %a, {IND_RC_W, sp, 0-4}
bdnz {LABEL, "1b"}
pat exg $1==4 /* Exchange top two words */
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
with INT_W+FLOAT_W INT_W+FLOAT_W
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields %1 %2
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat exg defined($1) /* Exchange other size */
leaving
loc $1
cal ".exg"
pat exg !defined($1)
leaving
cal ".exg"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Type conversions */
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat loc loc ciu /* signed -> unsigned */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
loc $1
loc $2
cuu
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat loc loc cui /* unsigned -> signed */
leaving
loc $1
loc $2
cuu
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat loc loc cuu $1<=4 && $2<=4 /* unsigned -> unsigned */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* nop */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat loc loc cii $1<=4 && $2<=$1
/* signed -> signed of smaller or same size,
* no sign extension */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat loc loc cii $1==1 && $2<=4 /* sign-extend char */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {SEX_B, %1}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat loc loc cii $1==2 && $2<=4 /* sign-extend short */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {SEX_H, %1}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Local variables */
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
pat lal smalls($1) /* Load address of local */
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
yields {SUM_RC, fp, $1}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
pat lal /* Load address of local */
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
uses REG={SUM_RIS, fp, his($1)}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
yields {SUM_RC, %a, los($1)}
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat lal loi smalls($1) && $2==1 /* Load byte from local */
yields {FRAME_B, 0, fp, $1, 1}
/* Load half-word from local and sign-extend */
pat lal loi loc loc cii smalls($1) && $2==2 && $3==2 && $4==4
yields {FRAME_H_S, 0, fp, $1, 1}
pat lal loi smalls($1) && $2==2 /* Load half-word from local */
yields {FRAME_H, 0, fp, $1, 1}
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
/* Load word from local */
Add reglap to ncg. Add 4-byte reg_float to PowerPC ncg. The new feature "reglap" allows two sizes of floating-point register variables (reg_float), if each register overlaps a single register of the other size. PowerPC ncg uses reglap to define 4-byte instances of f14 to f31 that overlap the 8-byte instances. When ncgg sees the definition of fs14("f14")=f14, it removes the 8-byte f14 from its rvnumbers array, and adds the 4-byte fs14 in its place. Later, when ncg puts a variable in fs14, if it is an 8-byte variable, then ncg switches to the 8-byte f14. The code has /* reglap */ comments in util/ncgg or #ifdef REGLAP in mach/proto/ncg reglap became necessary because my commit a20b87c caused PowerPC ego to allocate reg_float in both 4-byte and 8-byte sizes.
2017-10-14 16:40:04 +00:00
pat lol inreg($1)==reg_any || inreg($1)==reg_float
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {LOCAL, $1}
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat lol smalls($1)
yields {FRAME_W, 0, fp, $1, 4}
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat lol
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lal $1
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
loi 4
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat ldl inreg($1)==reg_float /* Load double-word from local */
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
yields {DLOCAL, $1}
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat ldl smalls($1) && smalls($1+4)
/* smalls($1+4) implies FRAME_D %off<=0xFFFA */
yields {FRAME_D, 0, fp, $1, 8}
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat ldl
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lal $1
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
loi 8
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat lal sti smalls($1) && $2==1 /* Store byte to local */
with REG
kills IND_V, FRAME_V %level==0 && fover($1, 1)
gen move %1, {FRAME_B, 0, fp, $1, 1}
pat lal sti smalls($1) && $2==2 /* Store half-word to local */
with REG
kills IND_V, FRAME_V %level==0 && fover($1, 2)
gen move %1, {FRAME_H, 0, fp, $1, 2}
pat stl inreg($1)==reg_any /* Store word to local */
Rename ANY_BHW to INT_W; add FLOAT_W, FLOAT_D. INT_W, the integer set, continues to exclude FSREG, because we can't easily move FSREG to GPR. ANY4 becomes ISET+FLOAT_W and ANY8 becomes FLOAT_D.
2017-12-14 21:26:19 +00:00
with exact INT_W
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
/* ncg fails to infer that regvar($1) is dead! */
kills regvar($1)
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
gen move %1, {REG_EXPR, regvar($1)}
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
with STACK
gen
lwz {LOCAL, $1}, {IND_RC_W, sp, 0}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
addi sp, sp, {C, 4}
Add reglap to ncg. Add 4-byte reg_float to PowerPC ncg. The new feature "reglap" allows two sizes of floating-point register variables (reg_float), if each register overlaps a single register of the other size. PowerPC ncg uses reglap to define 4-byte instances of f14 to f31 that overlap the 8-byte instances. When ncgg sees the definition of fs14("f14")=f14, it removes the 8-byte f14 from its rvnumbers array, and adds the 4-byte fs14 in its place. Later, when ncg puts a variable in fs14, if it is an 8-byte variable, then ncg switches to the 8-byte f14. The code has /* reglap */ comments in util/ncgg or #ifdef REGLAP in mach/proto/ncg reglap became necessary because my commit a20b87c caused PowerPC ego to allocate reg_float in both 4-byte and 8-byte sizes.
2017-10-14 16:40:04 +00:00
pat stl inreg($1)==reg_float
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
with exact FLOAT_W
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
kills regvar_w($1, reg_float)
gen move %1, {FSREG_EXPR, regvar_w($1, reg_float)}
Add reglap to ncg. Add 4-byte reg_float to PowerPC ncg. The new feature "reglap" allows two sizes of floating-point register variables (reg_float), if each register overlaps a single register of the other size. PowerPC ncg uses reglap to define 4-byte instances of f14 to f31 that overlap the 8-byte instances. When ncgg sees the definition of fs14("f14")=f14, it removes the 8-byte f14 from its rvnumbers array, and adds the 4-byte fs14 in its place. Later, when ncg puts a variable in fs14, if it is an 8-byte variable, then ncg switches to the 8-byte f14. The code has /* reglap */ comments in util/ncgg or #ifdef REGLAP in mach/proto/ncg reglap became necessary because my commit a20b87c caused PowerPC ego to allocate reg_float in both 4-byte and 8-byte sizes.
2017-10-14 16:40:04 +00:00
with STACK
gen
lfs {LOCAL, $1}, {IND_RC_W, sp, 0}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
addi sp, sp, {C, 4}
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat stl smalls($1)
with REG+FSREG
kills IND_V, FRAME_V %level==0 && fover($1, 4)
gen move %1, {FRAME_W, 0, fp, $1, 4}
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat stl
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lal $1
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
sti 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat sdl inreg($1)==reg_float /* Store double-word to local */
with exact FLOAT_D
Use my new regvar_w() and regvar_d() in PowerPC ncg. Rename GPRE to GPR_EXPR, then define FPR_EXPR and FSREG_EXPR. Use them for moves to register variables. Keep "kills regvar($1)", because deleting it and recompiling libc would cause many failures in my test programs. Add comment to warn, /* ncg fails to infer that regvar($1) is dead! */ Remove "kills LOCAL %off==$1" because it seems to have no effect.
2017-10-17 18:15:33 +00:00
kills regvar_d($1, reg_float)
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
gen move %1, {FREG_EXPR, regvar_d($1, reg_float)}
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
with STACK
gen
lfd {DLOCAL, $1}, {IND_RC_D, sp, 0}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
addi sp, sp, {C, 8}
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat sdl smalls($1) && smalls($1+4)
with REG REG
kills IND_V, FRAME_V %level==0 && fover($1, 8)
gen
move %1, {FRAME_W, 0, fp, $1, 4}
move %2, {FRAME_W, 0, fp, $1+4, 4}
with FREG
kills IND_V, FRAME_V %level==0 && fover($1, 4)
gen move %1, {FRAME_D, 0, fp, $1, 8}
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat sdl
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lal $1
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
sti 8
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove two obsolete patterns. These patterns seem to have no effect on the generated code.
2017-12-23 00:57:42 +00:00
pat lil /* Load indirect from local */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lol $1
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
loi 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat sil /* Store indirect to local */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lol $1
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
sti 4
Add a rule for sdl ldl $1==$2 to work around a bug. In our powerpc table, sdl fails to kill the old value of the local. This is a bug, because a later ldl can load the old value instead of the newly stored value. By rewriting "sdl 0" "ldl 0" as "dup 8" "sdl 0", the newly added rule works around the bug, but only when the ldl is immediately after the sdl. This rule improves code that uses double-precision floating point. The output of printf("%f", 6.0) in C changes from all zero digits to "6000000" but still doesn't print the decimal point. The result of atof("-123.456") becomes correct. In startrek, I can now move the Enterprise, but I still can't fire phasers without crashing the game. We already have a rule for stl lol $1==$2. We had two copies of the rule, so I am deleting the second copy.
2016-09-30 15:50:50 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat zrl /* Zero local */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
loc 0
stl $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat inl /* Increment local */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lol $1
loc 1
adi 4
stl $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat del /* Decrement local */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lol $1
loc 1
sbi 4
stl $1
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
/* Local variables of procedures on static chain */
/* lxa (lexical argument base) -> lxl (lexical local base) */
pat lxa adp nicelx($1)
leaving lxl $1 adp $2+EM_BSIZE
pat lxa lof nicelx($1)
leaving lxl $1 lof $2+EM_BSIZE
pat lxa ldf nicelx($1)
leaving lxl $1 ldf $2+EM_BSIZE
pat lxa stf nicelx($1)
leaving lxl $1 stf $2+EM_BSIZE
pat lxa sdf nicelx($1)
leaving lxl $1 stf $2+EM_BSIZE
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
pat lxa nicelx($1)
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
leaving lxl $1 adp EM_BSIZE
/* Load locals in statically enclosing procedures */
pat lxl adp loi nicelx($1) && smalls($2) && $3==1
uses REG={LXFRAME, $1}
yields {FRAME_B, $1, %a, $2, 1}
pat lxl adp loi loc loc cii nicelx($1) && smalls($2) &&
$3==2 && $4==2 && $5==4
uses REG={LXFRAME, $1}
yields {FRAME_H_S, $1, %a, $2, 2}
pat lxl adp loi nicelx($1) && smalls($2) && $3==2
uses REG={LXFRAME, $1}
yields {FRAME_H, $1, %a, $2, 2}
pat lxl lof nicelx($1) && smalls($2)
uses REG={LXFRAME, $1}
yields {FRAME_W, $1, %a, $2, 4}
pat lxl ldf nicelx($1) && smalls($2) && smalls($2+4)
uses REG={LXFRAME, $1}
/* smalls($2+4) implies FRAME_D %off<=0xFFFA */
yields {FRAME_D, $1, %a, $2, 8}
/* Store locals in statically enclosing procedures */
pat lxl adp sti nicelx($1) && smalls($2) && $3==1
with REG
kills IND_V, FRAME_V %level==$1 && fover($2, 1)
uses REG={LXFRAME, $1}
gen move %1, {FRAME_B, $1, %a, $2, 1}
pat lxl adp sti nicelx($1) && smalls($2) && $3==2
with REG
kills IND_V, FRAME_V %level==$1 && fover($2, 2)
uses REG={LXFRAME, $1}
gen move %1, {FRAME_H, $1, %a, $2, 2}
pat lxl stf nicelx($1) && smalls($2)
with REG+FSREG
kills IND_V, FRAME_V %level==$1 && fover($2, 4)
uses REG={LXFRAME, $1}
gen move %1, {FRAME_W, $1, %a, $2, 4}
pat lxl sdf nicelx($1) && smalls($2) && smalls($2+4)
with REG REG
kills IND_V, FRAME_V %level==$1 && fover($2, 8)
uses REG={LXFRAME, $1}
gen
move %1, {FRAME_W, $1, %a, $2, 4}
move %2, {FRAME_W, $1, %a, $2+4, 4}
with FREG
kills IND_V, FRAME_V %level==$1 && fover($2, 8)
uses REG={LXFRAME, $1}
gen move %1, {FRAME_D, $1, %a, $2, 8}
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
pat lxl nicelx($1) /* Local base on static chain */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
uses REG={LXFRAME, $1}
yields %a /* Can't yield LXFRAME. */
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
pat lxl stl nicelx($1) && inreg($2)==reg_any
kills regvar($2)
gen move {LXFRAME, $1}, {REG_EXPR, regvar($2)}
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
pat lxl $1==0 /* Our local base */
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
yields fp
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
pat lxa $1==0 /* Our argument base */
yields {SUM_RC, fp, EM_BSIZE}
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Global variables */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat lpi /* Load address of function */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lae $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat lae /* Load address of external */
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
uses REG={LABEL_HA, $1}
yields {SUM_RL, %a, $1}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat loe /* Load word external */
leaving
lae $1
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
loi 4
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat ste /* Store word external */
leaving
lae $1
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
sti 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat lde /* Load double-word external */
leaving
lae $1
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
loi 8
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat sde /* Store double-word external */
leaving
lae $1
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
sti 8
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat zre /* Zero external */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
loc 0
ste $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat ine /* Increment external */
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
leaving
loe $1
inc
ste $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat dee /* Decrement external */
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
leaving
loe $1
dec
ste $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Structures */
pat lof /* Load word offsetted */
leaving
adp $1
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
loi 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat ldf /* Load double-word offsetted */
leaving
adp $1
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
loi 8
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat stf /* Store word offsetted */
leaving
adp $1
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
sti 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat sdf /* Store double-word offsetted */
leaving
adp $1
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
sti 8
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Loads and stores */
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loi $1==1 /* Load byte indirect */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
yields {IND_RC_B, %1, 0}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with exact SUM_RC
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
yields {IND_RC_B, %1.reg, %1.off}
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
with exact SUM_RL
yields {IND_RL_B, %1.reg, %1.adr}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with exact SUM_RR
yields {IND_RR_B, %1.reg1, %1.reg2}
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
/* Load half-word indirect and sign-extend */
pat loi loc loc cii $1==2 && $2==2 && $3==4
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
yields {IND_RC_H_S, %1, 0}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with exact SUM_RC
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
yields {IND_RC_H_S, %1.reg, %1.off}
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
with exact SUM_RL
yields {IND_RL_H_S, %1.reg, %1.adr}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with exact SUM_RR
yields {IND_RR_H_S, %1.reg1, %1.reg2}
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loi $1==2 /* Load half-word indirect */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
yields {IND_RC_H, %1, 0}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with exact SUM_RC
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
yields {IND_RC_H, %1.reg, %1.off}
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
with exact SUM_RL
yields {IND_RL_H, %1.reg, %1.adr}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with exact SUM_RR
yields {IND_RR_H, %1.reg1, %1.reg2}
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loi $1==4 /* Load word indirect */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {IND_RC_W, %1, 0}
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
with exact SUM_RC
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {IND_RC_W, %1.reg, %1.off}
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
with exact SUM_RL
yields {IND_RL_W, %1.reg, %1.adr}
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
with exact SUM_RR
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {IND_RR_W, %1.reg1, %1.reg2}
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loi $1==8 /* Load double-word indirect */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {IND_RC_D, %1, 0}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with exact SUM_RC
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {IND_RC_D, %1.reg, %1.off}
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
with exact SUM_RL
yields {IND_RL_D, %1.reg, %1.adr}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with exact SUM_RR
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {IND_RR_D, %1.reg1, %1.reg2}
pat loi /* Load arbitrary size */
leaving
loc $1
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
los 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
pat los $1==4 /* Load arbitrary size */
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
with REG3 STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
kills ALL
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen bl {LABEL, ".los4"}
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat sti $1==1 /* Store byte indirect */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RC_B, %1, 0}
with SUM_RC REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RC_B, %1.reg, %1.off}
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
with SUM_RL REG
kills MEMORY
gen move %2, {IND_RL_B, %1.reg, %1.adr}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with SUM_RR REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RR_B, %1.reg1, %1.reg2}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat sti $1==2 /* Store half-word indirect */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RC_H, %1, 0}
with SUM_RC REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RC_H, %1.reg, %1.off}
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
with SUM_RL REG
kills MEMORY
gen move %2, {IND_RL_H, %1.reg, %1.adr}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with SUM_RR REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RR_H, %1.reg1, %1.reg2}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat sti $1==4 /* Store word indirect */
Use ha16/lo16 to emit pairs of lis/stw, lis/lfs, lis/stfs. A 4-byte load from a label yields a token IND_RL_W. This token emits either lis/lwz or lis/lfs, if we want a general-purpose register or a floating-point register.
2017-02-02 15:48:25 +00:00
with REG REG+FSREG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RC_W, %1, 0}
with SUM_RC REG+FSREG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RC_W, %1.reg, %1.off}
In PowerPC ncg, allocate register for ha16[label]. Use it to generate code like lis r12,ha16[__II0] lis r11,ha16[_f] lfs f1,lo16[_f](r11) lfs f2,lo16[__II0](r12) fadds f13,f2,f1 stfs f13,lo16[_f](r11) Here ncg has allocated r11 for ha16[_f]. We use r11 in lfs and again in stfs. Before this change, we needed an extra lis before stfs, because ncg did not remember that ha16[_f] was in a register. This example has a gap between ha16[__II0] and lo16[__II0], because the lo16 is not in the next instruction. This requires my previous commit 1bf58cf for RELOLIS. There is a gap because ncg emits the lis as soon as I allocate it. The "lfs f2,lo16[__II0](r12)" happens in a coercion from IND_RL_W to FSREG. The coercion allocates one FSREG but may not allocate any other registers. So I must allocate r12 earlier. I allocate r12 in pat lae, but this causes a gap.
2017-02-08 17:23:06 +00:00
with SUM_RL REG+FSREG
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RL_W, %1.reg, %1.adr}
with SUM_RR REG+FSREG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RR_W, %1.reg1, %1.reg2}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat sti $1==8 /* Store double-word indirect */
In powerpc ncg, add a speed hack for sti 8. ncg is too slow with this many registers. A stack pattern "with GPR GPR GPR" or "with REG REG REG" takes too long to pick registers, causing ncg 8 to take about 2 seconds on each sti 8. I introduce REG_PAIR and there are only 4 such pairs. For programs that use sti 8 (including C programs that copy 8-byte structs), this speed hack improves the ncg run from several seconds to almost instantaneous. Also add a few COMMENT(...) lines in stacking rules.
2016-10-18 00:31:59 +00:00
with REG FREG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RC_D, %1, 0}
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with SUM_RC FREG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
gen move %2, {IND_RC_D, %1.reg, %1.off}
Use ha16/lo16 to load or store 1, 2, 8 bytes from labels. Add the tokens IND_RL_B, IND_RL_H, IND_RL_H_S, IND_RL_D, along with the rules to use them. These rules emit shorter code. For example, loading a byte becomes lis, lbz instead of lis, addi, lbz. While making this, I wrongly set IND_RL_D to size 4. Then ncg made infinite recursion in codegen() and stackupto(), until it crashed by stack overflow. I correctly set IND_RL_D to size 8, preventing the crash.
2017-02-08 17:31:14 +00:00
with SUM_RL FREG
kills MEMORY
gen move %2, {IND_RL_D, %1.reg, %1.adr}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with SUM_RR FREG
kills MEMORY
gen move %2, {IND_RR_D, %1.reg1, %1.reg2}
Remove REG_PAIR. I added REG_PAIR in cfbc537 to speed up the register allocator, because ncg was taking about 2 seconds on each sti 8. I defined only 4 such pairs, so allocating REG_PAIR was much faster than allocating REG REG. After my last commit c5bb3be, allocation of REG REG is fast, and REG_PAIR seems unnecessary.
2017-02-13 23:11:27 +00:00
with REG REG REG
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
kills MEMORY
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Remove REG_PAIR. I added REG_PAIR in cfbc537 to speed up the register allocator, because ncg was taking about 2 seconds on each sti 8. I defined only 4 such pairs, so allocating REG_PAIR was much faster than allocating REG REG. After my last commit c5bb3be, allocation of REG REG is fast, and REG_PAIR seems unnecessary.
2017-02-13 23:11:27 +00:00
move %2, {IND_RC_W, %1, 0}
move %3, {IND_RC_W, %1, 4}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat sti /* Store arbitrary size */
leaving
loc $1
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
sts 4
Add "kills MEMORY" to powerpc sti rules. Adjust some of the loi rules (and associated moves) so we can identify the tokens that must be in MEMORY. With this commit, I can navigate the Enterprise even if I comment out my work-around from e22c888.
2016-10-16 22:13:39 +00:00
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
pat sts $1==4 /* Store arbitrary size */
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
with REG3 STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
kills ALL
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen bl {LABEL, ".sts4"}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Arithmetic wrappers */
pat ads $1==4 /* Add var to pointer */
leaving adi $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat sbs $1==4 /* Subtract var from pointer */
leaving sbi $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat adp /* Add constant to pointer */
leaving
loc $1
adi 4
pat adu /* Add unsigned */
leaving
adi $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat sbu /* Subtract unsigned */
leaving
sbi $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat inc /* Add 1 */
leaving
loc 1
adi 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat dec /* Subtract 1 */
leaving
loc 1
sbi 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Use "mr" and make a few other tweaks in PowerPC ncg table. Use extended "mr" instead of basic "or" to move registers. Both "mr" and "or" encode the same machine instruction. With "mr", I can more easily search the assembly output for register moves. Fold several stacking rules into a single rule ANY_BHW-REG to STACK. Remove the EM patterns for loc mlu $2==2 and loc slu. The first pattern had the wrong size (should be $2==4, not $2==2). Both patterns were redundant. They rewrote loc mlu as loc mli and loc slu as loc sli, but this table doesn't have patterns for loc mli or loc sli, so it is enough to rewrite mlu as mli and slu as sli.
2017-02-10 16:45:50 +00:00
pat mlu /* Multiply unsigned */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
mli $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Use "mr" and make a few other tweaks in PowerPC ncg table. Use extended "mr" instead of basic "or" to move registers. Both "mr" and "or" encode the same machine instruction. With "mr", I can more easily search the assembly output for register moves. Fold several stacking rules into a single rule ANY_BHW-REG to STACK. Remove the EM patterns for loc mlu $2==2 and loc slu. The first pattern had the wrong size (should be $2==4, not $2==2). Both patterns were redundant. They rewrote loc mlu as loc mli and loc slu as loc sli, but this table doesn't have patterns for loc mli or loc sli, so it is enough to rewrite mlu as mli and slu as sli.
2017-02-10 16:45:50 +00:00
pat slu /* Shift left unsigned */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
sli $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Word arithmetic */
Add some comments before I forget how this stuff works.
2018-01-24 20:17:32 +00:00
/* Like most back ends, this one doesn't trap EIOVFL, so it
* ignores overflow in signed integers.
*/
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat adi $1==4 /* Add word (second + top) */
with REG REG
yields {SUM_RR, %1, %2}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with CONST2 REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {SUM_RC, %2, %1.val}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with REG CONST2
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {SUM_RC, %1, %2.val}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with CONST_HI_ZR REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {SUM_RIS, %2, his(%1.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with REG CONST_HI_ZR
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {SUM_RIS, %1, his(%2.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with CONST_STACK-CONST2-CONST_HI_ZR REG
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
uses reusing %2, REG={SUM_RIS, %2, his(%1.val)}
yields {SUM_RC, %a, los(%1.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with REG CONST_STACK-CONST2-CONST_HI_ZR
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
uses reusing %1, REG={SUM_RIS, %1, his(%2.val)}
yields {SUM_RC, %a, los(%2.val)}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat sbi $1==4 /* Subtract word (second - top) */
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {SUB_RR, %2, %1}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with CONST2_WHEN_NEG REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {SUM_RC, %2, 0-%1.val}
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
with REG CONST2
yields {SUB_CR, %2.val, %1}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with CONST_HI_ZR REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {SUM_RIS, %2, his(0-%1.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with CONST_STACK-CONST2_WHEN_NEG-CONST_HI_ZR REG
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
uses reusing %2, REG={SUM_RIS, %2, his(0-%1.val)}
yields {SUM_RC, %a, los(0-%1.val)}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat ngi $1==4 /* Negate word */
with REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {NEG_R, %1}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat mli $1==4 /* Multiply word (second * top) */
Use subfic (val - reg) and mulli (reg * val). In the instruction list, put /* kills xer */ for sraw, srawi, subfic; and correct the (now unused) "addi." and "lfdu". Change MACHOPT_F from -m3 to -m2. This changes the code for 15 * i from slwi r3,r4,4 subfic r5,r4,0 add r3,r3,r5 to mulli r3,r4,15 If the sequence "slwi subfic addi" takes 3 cycles and 12 bytes, and mulli takes 3 cycles and 4 bytes, then mulli is better.
2018-01-27 20:33:43 +00:00
with CONST2 REG
yields {MUL_RC, %2, %1.val}
with REG CONST2
yields {MUL_RC, %1, %2.val}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
with REG REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {MUL_RR, %2, %1}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat dvi $1==4 /* Divide word (second / top) */
with REG REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {DIV_RR, %2, %1}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
pat dvu $1==4 /* Divide unsigned word (second / top) */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
with REG REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {DIV_RR_U, %2, %1}
/* To calculate a remainder: a % b = a - (a / b * b) */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat rmi $1==4 /* Remainder word (second % top) */
with REG REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
uses REG={DIV_RR, %2, %1}, REG
gen move {MUL_RR, %a, %1}, %b
yields {SUB_RR, %2, %b}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
pat rmu $1==4 /* Remainder unsigned word (second % top) */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
with REG REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
uses REG={DIV_RR_U, %2, %1}, REG
gen move {MUL_RR, %a, %1}, %b
yields {SUB_RR, %2, %b}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add test for EM rol, ror. Fix i80, i86, powerpc. EM instructions _rol_ and _ror_ do rotate an integer left or right. Our compilers and optimizers never emit _rol_ nor _ror_, but I might want to use them in the future. Add _rol_ and _ror_ to powerpc. Fix `rol 4` and `ror 4` in both i80 and i86, where the rules for `rol 4` and `ror 4` seem to have never been tested until now.
2017-12-07 22:16:21 +00:00
/* Bitwise logic */
Add some comments before I forget how this stuff works.
2018-01-24 20:17:32 +00:00
/* This back end doesn't know how to combine shifts and
* bitwise ops to emit rlwinm, rlwnm, or rlwimi instructions.
*/
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat and $1==4 /* AND word */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG NOT_R
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {ANDC_RR, %1, %2.reg}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with NOT_R REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {ANDC_RR, %2, %1.reg}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {AND_RR, %1, %2}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG UCONST2
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {AND_RC, %1, %2.val}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with UCONST2 REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {AND_RC, %2, %1.val}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with REG CONST_HI_ZR
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {AND_RIS, %1, hi(%2.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with CONST_HI_ZR REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {AND_RIS, %2, hi(%1.val)}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
pat and defined($1) /* AND set */
Add a bunch more set operations to the PowerPC backends, and the Pascal test for the same.
2017-01-17 21:31:38 +00:00
leaving
loc $1
cal ".and"
pat and !defined($1)
leaving
cal ".and"
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat ior $1==4 /* OR word */
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with REG NOT_R
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {ORC_RR, %1, %2.reg}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with NOT_R REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {ORC_RR, %2, %1.reg}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with REG REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {OR_RR, %1, %2}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with REG UCONST2
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {OR_RC, %1, %2.val}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with UCONST2 REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {OR_RC, %2, %1.val}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with REG CONST_HI_ZR
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {OR_RIS, %1, hi(%2.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with CONST_HI_ZR REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {OR_RIS, %2, hi(%1.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with REG CONST_STACK-UCONST2-CONST_HI_ZR
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
uses reusing %1, REG={OR_RIS, %1, hi(%2.val)}
Fix issue with logical operations and small signed constants which turned out to affect multiple platforms.
2019-02-13 19:56:10 +00:00
yields {OR_RC, %a, lo(%2.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with CONST_STACK-UCONST2-CONST_HI_ZR REG
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
uses reusing %2, REG={OR_RIS, %2, hi(%1.val)}
Fix issue with logical operations and small signed constants which turned out to affect multiple platforms.
2019-02-13 19:56:10 +00:00
yields {OR_RC, %a, lo(%1.val)}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
pat ior defined($1) /* OR set */
Add a test (currently failing) to check that Pascal char sets can store all 256 possible values. Add the PowerPC ncg and mcg backend support to let the test actually run, including modifying a bunch of PowrePC libem functions so that they can be called from both ncg and mcg.
2017-01-15 21:28:14 +00:00
leaving
loc $1
cal ".ior"
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
/* OR set (variable), used in lang/m2/libm2/LtoUset.e */
pat ior !defined($1)
Add a test (currently failing) to check that Pascal char sets can store all 256 possible values. Add the PowerPC ncg and mcg backend support to let the test actually run, including modifying a bunch of PowrePC libem functions so that they can be called from both ncg and mcg.
2017-01-15 21:28:14 +00:00
leaving
cal ".ior"
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat xor $1==4 /* XOR word */
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with REG REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {XOR_RR, %1, %2}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with REG UCONST2
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {XOR_RC, %1, %2.val}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
with UCONST2 REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {XOR_RC, %2, %1.val}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with REG CONST_HI_ZR
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XOR_RIS, %1, hi(%2.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with CONST_HI_ZR REG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XOR_RIS, %2, hi(%1.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with REG CONST_STACK-UCONST2-CONST_HI_ZR
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
uses reusing %1, REG={XOR_RIS, %1, hi(%2.val)}
Fix issue with logical operations and small signed constants which turned out to affect multiple platforms.
2019-02-13 19:56:10 +00:00
yields {XOR_RC, %a, lo(%2.val)}
Rename two tokens. CONST_HZ was not hertz (Hz).
2017-12-18 17:36:10 +00:00
with CONST_STACK-UCONST2-CONST_HI_ZR REG
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
uses reusing %2, REG={XOR_RIS, %2, hi(%1.val)}
Fix issue with logical operations and small signed constants which turned out to affect multiple platforms.
2019-02-13 19:56:10 +00:00
yields {XOR_RC, %a, lo(%1.val)}
Refactor how powerpc ncg pushes constants. When loc (load constant) pushes a constant, it now checks the value of the constant and pushes any of 7 tokens. These tokens allow stack patterns to recognize 16-bit signed integers (CONST2), 16-bit unsigned integers (UCONST2), multiples of 0x10000 (CONST_HZ), and other interesting forms of constants. Use the new constant tokens in the rules for adi, sbi, and, ior, xor. Adjust a few other rules to understand the new tokens. Require that SUM_RC has a signed 16-bit constant, and OR_RC and XOR_RC each have an unsigned 16-bit constant. The moves from SUM_RC, OR_RC, XOR_RC to GPR no longer touch the scratch register, because the constant is not too big.
2016-10-16 17:58:54 +00:00
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
pat xor defined($1) /* XOR set */
Edit PowerPC assembly for .and, .cms, .ior, .xor, .zer Remove one addi instruction from some loops. These loops had increased 2 pointers, they now increase 1 index. I must initialize the index, so I add "li r6, 0" before each loop. Change .zer to use subf instead of neg, add. Change .xor to take the size on the real stack, as .and and .or have done since 81c677d.
2017-02-11 23:00:56 +00:00
leaving
loc $1
cal ".xor"
pat xor !defined($1)
leaving
cal ".xor"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat com $1==4 /* NOT word */
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
with exact AND_RR
yields {NAND_RR, %1.reg1, %1.reg2}
with exact OR_RR
yields {NOR_RR, %1.reg1, %1.reg2}
with exact XOR_RR
yields {EQV_RR, %1.reg1, %1.reg2}
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields {NOT_R, %1}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
pat com defined($1) /* NOT set */
Add a bunch more set operations to the PowerPC backends, and the Pascal test for the same.
2017-01-17 21:31:38 +00:00
leaving
loc $1
cal ".com"
pat com !defined($1)
leaving
cal ".com"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
pat zer $1==4 /* Push zero */
leaving
loc 0
Revise stack shuffles and integer conversions in PowerPC ncg. Allow asp 4, exg 4 to shuffle tokens without coercing them into registers; but comment why dup 4, dup 8 coerce tokens into registers. Allow dup, dus, exg with larger sizes; and add tests dup_e.e and exg_e.e to check that dup 20, dus, exg 20 work as well in powerpc as in i80 and i86. Then powerpc failed to compile loc 2 loc 4 cuu in dup_e.e. Revise the integer conversions, so powerpc can compile and pass the test.
2017-12-09 22:21:06 +00:00
pat zer defined($1) /* Create empty set */
Add a test (currently failing) to check that Pascal char sets can store all 256 possible values. Add the PowerPC ncg and mcg backend support to let the test actually run, including modifying a bunch of PowrePC libem functions so that they can be called from both ncg and mcg.
2017-01-15 21:28:14 +00:00
leaving
loc $1
cal ".zer"
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
Add test for EM rol, ror. Fix i80, i86, powerpc. EM instructions _rol_ and _ror_ do rotate an integer left or right. Our compilers and optimizers never emit _rol_ nor _ror_, but I might want to use them in the future. Add _rol_ and _ror_ to powerpc. Fix `rol 4` and `ror 4` in both i80 and i86, where the rules for `rol 4` and `ror 4` seem to have never been tested until now.
2017-12-07 22:16:21 +00:00
/* Shifts and rotations */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat sli $1==4 /* Shift left (second << top) */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with CONST_STACK REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %2, REG
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen slwi %a, %2, {C, %1.val & 0x1F}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields %a
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen slw %a, %2, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
pat sli stl $1==4 && inreg($2)==reg_any
with CONST_STACK REG
gen slwi {LOCAL, $2}, %2, {C, %1.val & 0x1F}
with REG REG
gen slw {LOCAL, $2}, %2, %1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
pat sri $1==4 /* Shift right signed (second >> top) */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with CONST_STACK REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %2, REG
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen srawi %a, %2, {C, %1.val & 0x1F}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields %a
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen sraw %a, %2, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
pat sri stl $1==4 && inreg($2)==reg_any
with CONST_STACK REG
gen srawi {LOCAL, $2}, %2, {C, %1.val & 0x1F}
with REG REG
gen sraw {LOCAL, $2}, %2, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
pat sru $1==4 /* Shift right unsigned (second >> top) */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with CONST_STACK REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %2, REG
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen srwi %a, %2, {C, %1.val & 0x1F}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields %a
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen srw %a, %2, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
yields %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
pat sru stl $1==4 && inreg($2)==reg_any
with CONST_STACK REG
gen srwi {LOCAL, $2}, %2, {C, %1.val & 0x1F}
with REG REG
gen srw {LOCAL, $2}, %2, %1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add test for EM rol, ror. Fix i80, i86, powerpc. EM instructions _rol_ and _ror_ do rotate an integer left or right. Our compilers and optimizers never emit _rol_ nor _ror_, but I might want to use them in the future. Add _rol_ and _ror_ to powerpc. Fix `rol 4` and `ror 4` in both i80 and i86, where the rules for `rol 4` and `ror 4` seem to have never been tested until now.
2017-12-07 22:16:21 +00:00
pat rol $1==4 /* Rotate left word */
with CONST_STACK REG
uses reusing %2, REG
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen rotlwi %a, %2, {C, %1.val & 0x1F}
Add test for EM rol, ror. Fix i80, i86, powerpc. EM instructions _rol_ and _ror_ do rotate an integer left or right. Our compilers and optimizers never emit _rol_ nor _ror_, but I might want to use them in the future. Add _rol_ and _ror_ to powerpc. Fix `rol 4` and `ror 4` in both i80 and i86, where the rules for `rol 4` and `ror 4` seem to have never been tested until now.
2017-12-07 22:16:21 +00:00
yields %a
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG
Add test for EM rol, ror. Fix i80, i86, powerpc. EM instructions _rol_ and _ror_ do rotate an integer left or right. Our compilers and optimizers never emit _rol_ nor _ror_, but I might want to use them in the future. Add _rol_ and _ror_ to powerpc. Fix `rol 4` and `ror 4` in both i80 and i86, where the rules for `rol 4` and `ror 4` seem to have never been tested until now.
2017-12-07 22:16:21 +00:00
gen rotlw %a, %2, %1
yields %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
pat rol stl $1==4 && inreg($2)==reg_any
with CONST_STACK REG
gen rotlwi {LOCAL, $2}, %2, {C, %1.val & 0x1F}
with REG REG
gen rotlw {LOCAL, $2}, %2, %1
Add test for EM rol, ror. Fix i80, i86, powerpc. EM instructions _rol_ and _ror_ do rotate an integer left or right. Our compilers and optimizers never emit _rol_ nor _ror_, but I might want to use them in the future. Add _rol_ and _ror_ to powerpc. Fix `rol 4` and `ror 4` in both i80 and i86, where the rules for `rol 4` and `ror 4` seem to have never been tested until now.
2017-12-07 22:16:21 +00:00
/*
* ror 4 -> ngi 4, rol 4
* because to rotate right by n bits is to rotate left by
* (32 - n), which is to rotate left by -n. PowerPC rotlw
* handles -n as (-n & 0x1F).
*/
pat ror $1==4 /* Rotate right word */
with CONST_STACK REG
uses reusing %2, REG
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen rotrwi %a, %2, {C, %1.val & 0x1F}
Add test for EM rol, ror. Fix i80, i86, powerpc. EM instructions _rol_ and _ror_ do rotate an integer left or right. Our compilers and optimizers never emit _rol_ nor _ror_, but I might want to use them in the future. Add _rol_ and _ror_ to powerpc. Fix `rol 4` and `ror 4` in both i80 and i86, where the rules for `rol 4` and `ror 4` seem to have never been tested until now.
2017-12-07 22:16:21 +00:00
yields %a
with /* anything */
leaving
ngi 4
rol 4
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
pat ror stl $1==4 && inreg($2)==reg_any
with CONST_STACK REG
gen rotrwi {LOCAL, $2}, %2, {C, %1.val & 0x1F}
with /* anything */
leaving
ngi 4
rol 4
stl $2
Add test for EM rol, ror. Fix i80, i86, powerpc. EM instructions _rol_ and _ror_ do rotate an integer left or right. Our compilers and optimizers never emit _rol_ nor _ror_, but I might want to use them in the future. Add _rol_ and _ror_ to powerpc. Fix `rol 4` and `ror 4` in both i80 and i86, where the rules for `rol 4` and `ror 4` seem to have never been tested until now.
2017-12-07 22:16:21 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Arrays */
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
pat aar $1==4 /* Address of array element */
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
leaving cal ".aar4"
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
pat lar $1==4 /* Load from array */
with STACK
Add some missing clauses to los, sts, aar, inn, cmi, cmu. We only implement 'los 4', 'sts 4', 'cmi 4', 'cmu 4', not for sizes other than 4. Add clause $1==4. We only implement inn when defined($1). The rule for aar needs 'kills ALL' because it kills many registers, like other rules that call libem.
2016-12-10 00:49:50 +00:00
kills ALL
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
bl {LABEL, ".aar4"}
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
/* pass r3 = size from .aar4 to .los4 */
bl {LABEL, ".los4"}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
pat lae lar $2==4 && nicesize(rom($1, 3))
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lae $1
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
aar 4
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
loi rom($1, 3)
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
pat sar $1==4 /* Store to array */
with STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
kills ALL
gen
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
bl {LABEL, ".aar4"}
/* pass r3 = size from .aar4 to .sts4 */
bl {LABEL, ".sts4"}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
pat lae sar $2==4 && nicesize(rom($1, 3))
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
lae $1
Use .los4 in lar 4 and .sts4 in sar 4. Our libem had two implementations of loading a block from a stack, one for lar 4 and one for los 4. Now lar 4 and los 4 share the code in .los4. Likewise, sar 4 and sts 4 share the code in .sts4. Rename .los to .los4 and .sts to .sts4, because they implement los 4 and sts 4. Remove the special case for loading or storing 4 bytes, because we can do it with 1 iteration of the loop. Remove the lines to "align size" where the size must already be a multiple of 4. Fix the upper bound check in .aar4. Change .aar4, .lar4, .los4, .sar4, .sts4 to pass all operands on the real stack, except that .los4 and .sts4 take the size in register r3. Have .aar4 set r3 to the size of the array element. So lar 4 is just .aar4 then .los4, and sar 4 is just .aar4 then .sts4. ncg no longer calls .lar4 and .sar4 in libem, because it inlines the code; but I keep .lar4 and .sar4 in libem, because mcg references them. They might or might not work in mcg.
2017-02-13 20:22:00 +00:00
aar 4
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
sti rom($1, 3)
/* Sets */
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
pat set defined($1) /* Create singleton set */
Add a test (currently failing) to check that Pascal char sets can store all 256 possible values. Add the PowerPC ncg and mcg backend support to let the test actually run, including modifying a bunch of PowrePC libem functions so that they can be called from both ncg and mcg.
2017-01-15 21:28:14 +00:00
leaving
loc $1
cal ".set"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
/* Create set (variable), used in lang/m2/libm2/LtoUset.e */
pat set !defined($1)
Add a test (currently failing) to check that Pascal char sets can store all 256 possible values. Add the PowerPC ncg and mcg backend support to let the test actually run, including modifying a bunch of PowrePC libem functions so that they can be called from both ncg and mcg.
2017-01-15 21:28:14 +00:00
leaving
cal ".set"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add some missing clauses to los, sts, aar, inn, cmi, cmu. We only implement 'los 4', 'sts 4', 'cmi 4', 'cmu 4', not for sizes other than 4. Add clause $1==4. We only implement inn when defined($1). The rule for aar needs 'kills ALL' because it kills many registers, like other rules that call libem.
2016-12-10 00:49:50 +00:00
pat inn defined($1) /* Test for set bit */
Add a test (currently failing) to check that Pascal char sets can store all 256 possible values. Add the PowerPC ncg and mcg backend support to let the test actually run, including modifying a bunch of PowrePC libem functions so that they can be called from both ncg and mcg.
2017-01-15 21:28:14 +00:00
leaving
loc $1
cal ".inn"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add a bunch more set operations to the PowerPC backends, and the Pascal test for the same.
2017-01-17 21:31:38 +00:00
pat inn !defined($1)
leaving
cal ".inn"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Boolean resolutions */
pat teq /* top = (top == 0) */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %1, REG
gen
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
test %1
mfcr %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XEQ, %a}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat tne /* top = (top != 0) */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %1, REG
gen
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
test %1
mfcr %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XNE, %a}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat tlt /* top = (top < 0) */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %1, REG
gen
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
test %1
mfcr %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLT, %a}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat tle /* top = (top <= 0) */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %1, REG
gen
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
test %1
mfcr %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLE, %a}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat tgt /* top = (top > 0) */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %1, REG
gen
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
test %1
mfcr %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGT, %a}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat tge /* top = (top >= 0) */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %1, REG
gen
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
test %1
mfcr %a
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGE, %a}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmi teq $1==4 /* Signed second == top */
with REG CONST2
uses reusing %1, REG={COND_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XEQ, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with CONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={COND_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XEQ, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={COND_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XEQ, %a}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmi tne $1==4 /* Signed second != top */
with REG CONST2
uses reusing %1, REG={COND_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XNE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with CONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={COND_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XNE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={COND_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XNE, %a}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmi tgt $1==4 /* Signed second > top */
with REG CONST2
uses reusing %1, REG={COND_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with CONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={COND_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={COND_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGT, %a}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmi tge $1==4 /* Signed second >= top */
with REG CONST2
uses reusing %1, REG={COND_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with CONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={COND_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={COND_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGE, %a}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmi tlt $1==4 /* Signed second < top */
with REG CONST2
uses reusing %1, REG={COND_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with CONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={COND_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={COND_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLT, %a}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmi tle $1==4 /* Signed second <= top */
with REG CONST2
uses reusing %1, REG={COND_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with CONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={COND_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={COND_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLE, %a}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmu teq $1==4 /* Unsigned second == top */
with REG UCONST2
uses reusing %1, REG={CONDL_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XEQ, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with UCONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={CONDL_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XEQ, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={CONDL_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XEQ, %a}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmu tne $1==4 /* Unsigned second != top */
with REG UCONST2
uses reusing %1, REG={CONDL_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XNE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with UCONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={CONDL_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XNE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={CONDL_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XNE, %a}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmu tgt $1==4 /* Unsigned second > top */
with REG UCONST2
uses reusing %1, REG={CONDL_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with UCONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={CONDL_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={CONDL_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGT, %a}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmu tge $1==4 /* Unsigned second >= top */
with REG UCONST2
uses reusing %1, REG={CONDL_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with UCONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={CONDL_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={CONDL_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmu tlt $1==4 /* Unsigned second < top */
with REG UCONST2
uses reusing %1, REG={CONDL_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with UCONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={CONDL_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={CONDL_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLT, %a}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmu tle $1==4 /* Unsigned second <= top */
with REG UCONST2
uses reusing %1, REG={CONDL_RC, %1, %2.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with UCONST2 REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %2, REG={CONDL_RC, %2, %1.val}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={CONDL_RR, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
/* Simple branches */
proc zxx example zeq
with REG STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
test %1
bxx* {LABEL, $1}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
/* Pop signed int, branch if... */
pat zeq call zxx("beq") /* top == 0 */
pat zne call zxx("bne") /* top != 0 */
pat zgt call zxx("bgt") /* top > 0 */
pat zge call zxx("bge") /* top >= 0 */
pat zlt call zxx("blt") /* top < 0 */
pat zle call zxx("ble") /* top >= 0 */
/* The peephole optimizer rewrites
* cmi 4 zeq
* as beq, and does same for bne, bgt, and so on.
*/
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
proc bxx example beq
with REG CONST2 STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
cmpwi %1, %2
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
bxx[2] {LABEL, $1}
with CONST2 REG STACK
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
cmpwi %2, %1
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
bxx[1] {LABEL, $1}
with REG REG STACK
gen
cmpw %2, %1
bxx[1] {LABEL, $1}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
/* Pop two signed ints, branch if... */
pat beq call bxx("beq", "beq") /* second == top */
pat bne call bxx("bne", "bne") /* second != top */
pat bgt call bxx("bgt", "blt") /* second > top */
pat bge call bxx("bge", "ble") /* second >= top */
pat blt call bxx("blt", "bgt") /* second < top */
pat ble call bxx("ble", "bge") /* second >= top */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
proc cmu4zxx example cmu zeq
with REG CONST2 STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
cmplwi %1, %2
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
bxx[2] {LABEL, $2}
with CONST2 REG STACK
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
cmplwi %2, %1
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
bxx[1] {LABEL, $2}
with REG REG STACK
gen
cmplw %2, %1
bxx[1] {LABEL, $2}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
/* Pop two unsigned ints, branch if... */
pat cmu zeq $1==4 call cmu4zxx("beq", "beq")
pat cmu zne $1==4 call cmu4zxx("bne", "bne")
pat cmu zgt $1==4 call cmu4zxx("bgt", "blt")
pat cmu zge $1==4 call cmu4zxx("bge", "ble")
pat cmu zlt $1==4 call cmu4zxx("blt", "bgt")
pat cmu zle $1==4 call cmu4zxx("ble", "bge")
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
/* Comparisons */
/* Each comparison extracts the lt and gt bits from cr0.
* extlwi %a, %a, 2, 0
* puts lt in the sign bit, so lt yields a negative result,
* gt yields positive.
* rlwinm %a, %a, 1, 31, 0
* puts gt in the sign bit, to reverse the comparison.
*/
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat cmi $1==4 /* Signed tristate compare */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG CONST2
uses reusing %1, REG={COND_RC, %1, %2.val}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen rlwinm %a, %a, {C, 1}, {C, 31}, {C, 0}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
yields %a
with CONST2 REG
uses reusing %2, REG={COND_RC, %2, %1.val}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen extlwi %a, %a, {C, 2}, {C, 0}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
yields %a
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={COND_RR, %2, %1}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen extlwi %a, %a, {C, 2}, {C, 0}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
yields %a
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat cmu $1==4 /* Unsigned tristate compare */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with REG UCONST2
uses reusing %1, REG={CONDL_RC, %1, %2.val}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen rlwinm %a, %a, {C, 1}, {C, 31}, {C, 0}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
yields %a
with UCONST2 REG
uses reusing %2, REG={CONDL_RC, %2, %1.val}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen extlwi %a, %a, {C, 2}, {C, 0}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
yields %a
with REG REG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, REG={CONDL_RR, %2, %1}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
gen extlwi %a, %a, {C, 2}, {C, 0}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
yields %a
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat cmp /* Compare pointers */
leaving
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
cmu 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat cms $1==4 /* Compare blocks (word sized) */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
leaving
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
cmi 4
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add Modula-2 set operations to PowerPC ncg. This provides and, ior, xor, com, zer, set, cms when defined($1) and ior, set when !defined($1). I don't provide the other operations !defined($1) because our Modula-2 compiler hasn't used them. I wrote a Modula-2 example in https://gist.github.com/kernigh/add79662bb3c63ffb7c46d01dc8ae788 Put a dummy comment in mach/powerpc/libem/build.lua so git checkout will touch that file. Without the touch, the build system doesn't see the new *.s files.
2016-12-10 17:23:07 +00:00
pat cms defined($1)
Add pat cms !defined($1) Switch .cms to pass inputs and outputs on the real stack, not in registers; like we do with .and, .or (81c677d) and .xor (c578c49). At this point, nearly all functions in libem use the real stack, not registers, for passing inputs and outputs. This simplifies the ncg table (which needs fewer lists of specific registers) but slows calls to libem. For example, after ba9b021, each call to .aar4 is about 10 instructions slower. I moved 3 inputs and 1 output from registers to the real stack. A program would take 4 instructions to move registers to stack, 4 to move stack to registers, and perhaps 2 to adjust the stack pointer.
2017-02-13 21:52:32 +00:00
leaving
loc $1
cal ".cms"
pat cms !defined($1)
leaving
cal ".cms"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Other branching and labelling */
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
/* During an unconditional jump, if the top element on the
* stack has 4 bytes, then we hold it in register r3.
*/
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat lab topeltsize($1)==4 && !fallthrough($1)
Add constraints for pat lab, as done in the m68020 table. Always use 'kills ALL' when reaching a label, because our registers and tokens have the wrong values if the program jumps to this label from somewhere else. When falling through a label, if the top element is in r3, then require that the rest of the stack is in the real STACK, not in registers or tokens. I'm doing this to be certain that the missing constraints are not causing bugs. I did not find any such bug, perhaps because the labels are usually near other instructions (like conditional branches and function calls) that stack or kill tokens.
2017-01-24 16:26:35 +00:00
kills ALL
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen labeldef $1
yields r3
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat lab topeltsize($1)==4 && fallthrough($1)
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
with REG3 STACK
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
kills ALL
gen labeldef $1
yields r3
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat lab topeltsize($1)!=4 /* Label without r3 */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
with STACK
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
kills ALL
gen labeldef $1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat bra topeltsize($1)==4 /* Branch with r3 */
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
with REG3 STACK
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen b {LABEL, $1}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat bra topeltsize($1)!=4 /* Branch without r3 */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
with STACK
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen b {LABEL, $1}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* Miscellaneous */
pat cal /* Call procedure */
with STACK
kills ALL
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen bl {LABEL, $1}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat cai /* Call procedure indirect */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG STACK
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
kills ALL
gen
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
mtspr ctr, %1
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
bctrl.
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat lfr $1==4 /* Load function result, word */
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
yields r3
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat lfr $1==8 /* Load function result, double-word */
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
yields r4 r3
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat ret $1==0 /* Return from procedure */
gen
Remove .ret from libem and inline the code. This removes a wrong-way dependency of libsys on libem. The C functions in libsys called .ret, but libsys is after libem in the linker arguments, so the linker didn't find .ret unless something else had called .ret. Almost everything called .ret, but I got a linker error when I wrote an assembly program using the EM runtime, because my assembly program didn't call .ret. Add a dummy comment to build.lua, so git checkout touches that file, the build system reconfigures itself, and the *.s glob sees that ret.s has gone.
2017-02-17 02:18:39 +00:00
/* Restore saved registers. */
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
return
Remove .ret from libem and inline the code. This removes a wrong-way dependency of libsys on libem. The C functions in libsys called .ret, but libsys is after libem in the linker arguments, so the linker didn't find .ret unless something else had called .ret. Almost everything called .ret, but I got a linker error when I wrote an assembly program using the EM runtime, because my assembly program didn't call .ret. Add a dummy comment to build.lua, so git checkout touches that file, the build system reconfigures itself, and the *.s glob sees that ret.s has gone.
2017-02-17 02:18:39 +00:00
/* Epilog: restore lr and fp. */
lwz r0, {IND_RC_W, fp, 4}
mtspr lr, r0
lwz r0, {IND_RC_W, fp, 0}
/* Free our stack frame. */
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
addi sp, fp, {C, 8}
Remove .ret from libem and inline the code. This removes a wrong-way dependency of libsys on libem. The C functions in libsys called .ret, but libsys is after libem in the linker arguments, so the linker didn't find .ret unless something else had called .ret. Almost everything called .ret, but I got a linker error when I wrote an assembly program using the EM runtime, because my assembly program didn't call .ret. Add a dummy comment to build.lua, so git checkout touches that file, the build system reconfigures itself, and the *.s glob sees that ret.s has gone.
2017-02-17 02:18:39 +00:00
mr fp, r0
blr.
Teach top to merge or delete "addi sp, sp, X". This reduces code size, because ncg emits too many "addi sp, sp, X" instructions when unstacking things. Now top lowers "addi sp, sp, X" by lifting other instructions. This sometimes creates chances to merge or delete _addi_ instructions. If no such chance is found, the _addi_ remains uselessly lowered. Edit ncg/table to remove something that top now does. Edit ncg/mach.c to remove some spaces after commas. This removes a whitespace difference between *.s and *.so files, because top removes the space.
2018-01-05 22:55:50 +00:00
/* If "ret" coerces STACK to REG3, then top will delete the
* extra "addi sp, sp, 4".
*/
Remove .ret from libem and inline the code. This removes a wrong-way dependency of libsys on libem. The C functions in libsys called .ret, but libsys is after libem in the linker arguments, so the linker didn't find .ret unless something else had called .ret. Almost everything called .ret, but I got a linker error when I wrote an assembly program using the EM runtime, because my assembly program didn't call .ret. Add a dummy comment to build.lua, so git checkout touches that file, the build system reconfigures itself, and the *.s glob sees that ret.s has gone.
2017-02-17 02:18:39 +00:00
pat ret $1==4 /* Return from procedure, word */
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
with REG3
Remove .ret from libem and inline the code. This removes a wrong-way dependency of libsys on libem. The C functions in libsys called .ret, but libsys is after libem in the linker arguments, so the linker didn't find .ret unless something else had called .ret. Almost everything called .ret, but I got a linker error when I wrote an assembly program using the EM runtime, because my assembly program didn't call .ret. Add a dummy comment to build.lua, so git checkout touches that file, the build system reconfigures itself, and the *.s glob sees that ret.s has gone.
2017-02-17 02:18:39 +00:00
leaving ret 0
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove .ret from libem and inline the code. This removes a wrong-way dependency of libsys on libem. The C functions in libsys called .ret, but libsys is after libem in the linker arguments, so the linker didn't find .ret unless something else had called .ret. Almost everything called .ret, but I got a linker error when I wrote an assembly program using the EM runtime, because my assembly program didn't call .ret. Add a dummy comment to build.lua, so git checkout touches that file, the build system reconfigures itself, and the *.s glob sees that ret.s has gone.
2017-02-17 02:18:39 +00:00
pat ret $1==8 /* Return from proc, double-word */
Generate shorter code for ret 4 and ret 8.
2017-12-23 01:37:39 +00:00
with REG3 INT_W
Reorder registers. Fix problem with ret 8. After c5bb3be, ncg began to allocate regvars from r13 up. I reorder the regvars so ncg again allocates them from r31 down. I also reorder the other registers. This exposed a bug in my rule for ret 8. It was wrong if item %2 was in r3, because I moved %1 to r3 before %2 to r4. Fix it by adding back an individual register class for r3 (called REG3 here, GPR3 in c5bb3be). Also fix my typo in mach.c that made a syntax error in assembly.
2017-02-18 00:32:27 +00:00
gen move %2, r4
leaving ret 0
Generate shorter code for ret 4 and ret 8.
2017-12-23 01:37:39 +00:00
with REG3 STACK
gen lwz r4, {IND_RC_W, sp, 0}
leaving ret 0
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Use lwzu, stwu to tighten more loops. Because lwzu or stwu moves the pointer, I can remove an addi instruction from the loop, so the loop is slightly faster. I wrote a benchmark in Modula-2 that exercises some of these loops. I measured its time on my old PowerPC Mac. Its user time decreases from 8.401s to 8.217s with the tighter loops.
2017-10-18 16:12:42 +00:00
/*
* These rules for blm/bls are wrong if length is zero.
* So are several procedures in libem.
*/
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat blm /* Block move constant length */
Implement blm and bls using an inline loop. This replaces a call to memmove() in libc. That was working for me, but it can fail because EM programs don't always link to libc. blm and bls only need to copy aligned words. They don't need to copy bytes, and they don't need to copy between overlapping buffers, as memmove() does. So the new loop is simpler than memmove().
2017-02-12 00:30:12 +00:00
leaving
loc $1
bls
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat bls /* Block move variable length */
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
with REG SPFP+REG SPFP+REG
/* allows sp as %2, %3 */
Use lwzu, stwu to tighten more loops. Because lwzu or stwu moves the pointer, I can remove an addi instruction from the loop, so the loop is slightly faster. I wrote a benchmark in Modula-2 that exercises some of these loops. I measured its time on my old PowerPC Mac. Its user time decreases from 8.401s to 8.217s with the tighter loops.
2017-10-18 16:12:42 +00:00
/* ( src%3 dst%2 len%1 -- ) */
uses reusing %1, REG, REG, REG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
srwi %a, %1, {C, 2}
Use lwzu, stwu to tighten more loops. Because lwzu or stwu moves the pointer, I can remove an addi instruction from the loop, so the loop is slightly faster. I wrote a benchmark in Modula-2 that exercises some of these loops. I measured its time on my old PowerPC Mac. Its user time decreases from 8.401s to 8.217s with the tighter loops.
2017-10-18 16:12:42 +00:00
mtspr ctr, %a
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
addi %b, %3, {C, 0-4}
addi %c, %2, {C, 0-4}
Use lwzu, stwu to tighten more loops. Because lwzu or stwu moves the pointer, I can remove an addi instruction from the loop, so the loop is slightly faster. I wrote a benchmark in Modula-2 that exercises some of these loops. I measured its time on my old PowerPC Mac. Its user time decreases from 8.401s to 8.217s with the tighter loops.
2017-10-18 16:12:42 +00:00
1: lwzu %a, {IND_RC_W, %b, 4}
stwu %a, {IND_RC_W, %c, 4}
Implement blm and bls using an inline loop. This replaces a call to memmove() in libc. That was working for me, but it can fail because EM programs don't always link to libc. blm and bls only need to copy aligned words. They don't need to copy bytes, and they don't need to copy between overlapping buffers, as memmove() does. So the new loop is simpler than memmove().
2017-02-12 00:30:12 +00:00
bdnz {LABEL, "1b"}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat csa /* Array-lookup switch */
Made csa and csb work with mcg; adjust the libem functions and the corresponding invocation in the ncg table so the same helpers can be used for both mcg and ncg. Add a new IR opcode, FARJUMP, which jumps to a helper function but saves volatile registers.
2016-11-19 09:55:41 +00:00
with STACK
Add missing instances of "kills ALL" or "with STACK".
2017-02-13 21:38:26 +00:00
kills ALL
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen b {LABEL, ".csa"}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat csb /* Table-lookup switch */
Made csa and csb work with mcg; adjust the libem functions and the corresponding invocation in the ncg table so the same helpers can be used for both mcg and ncg. Add a new IR opcode, FARJUMP, which jumps to a helper function but saves volatile registers.
2016-11-19 09:55:41 +00:00
with STACK
Add missing instances of "kills ALL" or "with STACK".
2017-02-13 21:38:26 +00:00
kills ALL
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen b {LABEL, ".csb"}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
/* EM specials */
pat fil /* Set current filename */
leaving
lae $1
Remove .linenumber and .filename; use hol0 and hol0+4. We need this because some .e files in lang/ are using 'loe 0' and 'lae 4' to load the line number from hol0 and filename from hol0+4.
2016-09-30 17:40:36 +00:00
ste "hol0+4"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat lin /* Set current line number */
leaving
loc $1
Remove .linenumber and .filename; use hol0 and hol0+4. We need this because some .e files in lang/ are using 'loe 0' and 'lae 4' to load the line number from hol0 and filename from hol0+4.
2016-09-30 17:40:36 +00:00
ste "hol0"
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat lni /* Increment line number */
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
leaving ine "hol0"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat lim /* Load EM trap ignore mask */
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
leaving loe ".ignmask"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat sim /* Store EM trap ignore mask */
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
leaving ste ".ignmask"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add test for EM _rck_. Fix traps in PowerPC ncg. The new test rck_e.e segfaults on PowerPC unless I make some changes. The inline code for _rck_ was wrong because it didn't allow the trap handler to return. _sig_ forgot to push the old trap handler. Move plat/linuxppc/libsys/trap.s to mach/powerpc/libem/trp.s and rewrite it with simplified/extended mnemonics. Remove .trap alias for .trp procedure. Add a missing `mtspr lr, r0` so we can return from the trap handler. Call write() and _exit() so trp.s works with both linuxppc and osxppc. Before, Mac OS X was wrongly using the trap.s for Linux. In powerpc/libem, simplify .aar4; teach .csa and .csb to raise the trap if the default target is zero. C programs don't need these changes. You may relink your C programs with the changed .csa and .csb, but C code doesn't raise the trap. Modula-2 code can raise traps, so you may want to relink your Modula-2 programs with the changed libem, but you might keep your old .o files from Modula-2. You may need to recompile your Pascal programs (delete old .o files from Pascal) because the Pascal compiler might use _rck_.
2017-12-25 03:37:52 +00:00
pat sig /* Set trap handler, yield old */
leaving
loe ".trppc"
exg 4
ste ".trppc"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
pat trp /* Raise EM trap */
with REG3
kills ALL
gen bl {LABEL, ".trp"}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat rtt /* Return from trap */
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
leaving ret 0
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
pat rck $1==4 /* Range check */
leaving cal ".rck"
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
/* Our caller's local base, "lxl 0 dch", appears in
* lang/cem/libcc.ansi/setjmp/setjmp.e, lang/m2/libm2/par_misc.e
Implement lxl for PowerPC ncg. This fixes lxl 1 (so it follows the static chain, not the dynamic chain) and provides lxl 2 and greater. The Modula-2 compiler uses lxl for nested procedures, so they can access the variables of the enclosing procedures.
2017-02-14 04:22:31 +00:00
*/
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
pat lxl dch $1==0
yields {IND_RC_W, fp, FP_OFFSET}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Implement lxl for PowerPC ncg. This fixes lxl 1 (so it follows the static chain, not the dynamic chain) and provides lxl 2 and greater. The Modula-2 compiler uses lxl for nested procedures, so they can access the variables of the enclosing procedures.
2017-02-14 04:22:31 +00:00
pat dch /* Dynamic chain: LB -> caller's LB */
Trimming mach/powerpc/ncg/table Remove coercion from LABEL to REG. The coercion never happens because I have stopped putting LABEL on the stack. Also remove LABEL from set ANY_BHW. Retain the move from LABEL to REG because pat gto uses it. Remove li32 instruction, unused after the switch to the hi16, ha16, lo16 syntax. Remove COMMENT(...) lines from most moves. In my opinion, they took too much space, both in the table and in the assembly output. The stacking rules and coercions keep their COMMENT(...) lines. In test GPR, don't write to RSCRATCH. Fold several coercions into a single coercion from ANY_BHW uses REG. Use REG instead of GPR in stack patterns. REG and GPR act the same, because every GPR on the stack is a REG, but I want to be clear that I expect a REG, not r0. In code rules, sort SUM_RC before SORT_RR, so I can add SUM_RL later. Remove rules to optimize loc loc cii loc loc cii. If $2==$4, the peephole optimizer can optimize it. If $2!=$4, then the EM program is missing a conversion from size $2 to size $4. Remove rules to store a SEX_B with sti 1 or a SEX_H with sti 2. These rules would never get used, unless the EM program is missing a conversion from size 4 to size 1 or 2.
2017-02-08 17:27:16 +00:00
with REG
Implement lxl for PowerPC ncg. This fixes lxl 1 (so it follows the static chain, not the dynamic chain) and provides lxl 2 and greater. The Modula-2 compiler uses lxl for nested procedures, so they can access the variables of the enclosing procedures.
2017-02-14 04:22:31 +00:00
yields {IND_RC_W, %1, FP_OFFSET}
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Implement lxl for PowerPC ncg. This fixes lxl 1 (so it follows the static chain, not the dynamic chain) and provides lxl 2 and greater. The Modula-2 compiler uses lxl for nested procedures, so they can access the variables of the enclosing procedures.
2017-02-14 04:22:31 +00:00
pat lpb /* LB -> argument base */
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
leaving adp EM_BSIZE
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add tests for C <setjmp.h> and Modula-2 Semaphores. Fix PowerPC ncg so setjmp() returns the correct value. I got unlucky when ncg picked r3 for "uses REG"; this destroyed the return value in r3 and caused the new test to fail.
2018-01-03 19:51:14 +00:00
/* "gto" must preserve the function result for "lfr", so
* longjmp() can pass the return value to setjmp().
* - See lang/cem/libcc.ansi/setjmp/setjmp.e
*
* Must preserve r3 and r4, so no "uses REG".
* PowerPC can't add r0 + constant. Use r12.
*/
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
pat gto /* longjmp */
Add missing instances of "kills ALL" or "with STACK".
2017-02-13 21:38:26 +00:00
with STACK
gen
Add tests for C <setjmp.h> and Modula-2 Semaphores. Fix PowerPC ncg so setjmp() returns the correct value. I got unlucky when ncg picked r3 for "uses REG"; this destroyed the return value in r3 and caused the new test to fail.
2018-01-03 19:51:14 +00:00
move {LABEL, $1}, r12
move {IND_RC_W, r12, 8}, fp
move {IND_RC_W, r12, 4}, sp
move {IND_RC_W, r12, 0}, r12
mtspr ctr, r12
Add missing instances of "kills ALL" or "with STACK".
2017-02-13 21:38:26 +00:00
bctr.
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Implement lxl for PowerPC ncg. This fixes lxl 1 (so it follows the static chain, not the dynamic chain) and provides lxl 2 and greater. The Modula-2 compiler uses lxl for nested procedures, so they can access the variables of the enclosing procedures.
2017-02-14 04:22:31 +00:00
pat lor $1==0 /* Load local base */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
leaving lxl 0
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Implement lxl for PowerPC ncg. This fixes lxl 1 (so it follows the static chain, not the dynamic chain) and provides lxl 2 and greater. The Modula-2 compiler uses lxl for nested procedures, so they can access the variables of the enclosing procedures.
2017-02-14 04:22:31 +00:00
pat lor $1==1 /* Load stack pointer */
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
with STACK
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
yields sp
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
/* Next few patterns for "lor 1" appear in
* lang/m2/libm2/par_misc.e
*/
pat lor adp $1==1 && smalls($2) /* sp + constant */
with STACK
Allow sp and fp on the fake stack. This simplifies parts of the PowerPC table and causes ncg to better decide whether to push sp or fp to the real stack, or coerce it to REG3, or coerce it to REG-REG3, or move it to a regvar. These better decisions remove extra _mr_ instructions. The idea comes from mach/powerpc/arm/table, where SP has a property STACKPOINTER and LB has LOCALBASE. I don't need two properties, so I make one property SPFP for both registers.
2018-01-23 23:18:40 +00:00
yields {SUM_RC, sp, $2}
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
/* Subtract stack pointer by doing %1 - (sp - 4)
* because sp - 4 would point to %1.
*/
pat lor sbs loc adu $1==1 && $2==4 && $4==4
with REG STACK
uses reusing %1, REG
gen subf %a, sp, %1
yields %a
leaving loc $3+4 adu 4
pat lor sbs $1==1 && $2==4
with REG STACK
uses reusing %1, REG
gen subf %a, sp, %1
yields {SUM_RC, %a, 4}
Implement lxl for PowerPC ncg. This fixes lxl 1 (so it follows the static chain, not the dynamic chain) and provides lxl 2 and greater. The Modula-2 compiler uses lxl for nested procedures, so they can access the variables of the enclosing procedures.
2017-02-14 04:22:31 +00:00
pat str $1==0 /* Store local base */
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
with INT_W
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
gen move %1, fp
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
with STACK
gen
lwz fp, {IND_RC_W, sp, 0}
addi sp, sp, {C, 4}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Implement lxl for PowerPC ncg. This fixes lxl 1 (so it follows the static chain, not the dynamic chain) and provides lxl 2 and greater. The Modula-2 compiler uses lxl for nested procedures, so they can access the variables of the enclosing procedures.
2017-02-14 04:22:31 +00:00
pat str $1==1 /* Store stack pointer */
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
with INT_W
kills ALL
Add FRAME_V tokens for local variables. When storing to a local, stop killing the tokens of other locals, unless they might overlap with the stored local. This helps some procedures that juggle locals when the locals aren't in registers. Also use FRAME_V tokens for locals in statically enclosing procedures. Rewrite _lxa_ as _lxl_, to skip the `addi ?,?,8` if we can add 8 to the next constant. The PowerPC code from _lxl_ is now sometimes better, sometimes worse than before. The i386 table provided the idea to use %size to find overlapping locals.
2017-12-22 22:04:16 +00:00
gen move %1, sp
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
with STACK
kills ALL
gen lwz sp, {IND_RC_W, sp, 0}
Add a test (currently failing) to check that Pascal char sets can store all 256 possible values. Add the PowerPC ncg and mcg backend support to let the test actually run, including modifying a bunch of PowrePC libem functions so that they can be called from both ncg and mcg.
2017-01-15 21:28:14 +00:00
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
/* Single-precision floating-point */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat zrf $1==4 /* Push zero */
Add test for EM _rck_. Fix traps in PowerPC ncg. The new test rck_e.e segfaults on PowerPC unless I make some changes. The inline code for _rck_ was wrong because it didn't allow the trap handler to return. _sig_ forgot to push the old trap handler. Move plat/linuxppc/libsys/trap.s to mach/powerpc/libem/trp.s and rewrite it with simplified/extended mnemonics. Remove .trap alias for .trp procedure. Add a missing `mtspr lr, r0` so we can return from the trap handler. Call write() and _exit() so trp.s works with both linuxppc and osxppc. Before, Mac OS X was wrongly using the trap.s for Linux. In powerpc/libem, simplify .aar4; teach .csa and .csb to raise the trap if the default target is zero. C programs don't need these changes. You may relink your C programs with the changed .csa and .csb, but C code doesn't raise the trap. Modula-2 code can raise traps, so you may want to relink your Modula-2 programs with the changed libem, but you might keep your old .o files from Modula-2. You may need to recompile your Pascal programs (delete old .o files from Pascal) because the Pascal compiler might use _rck_.
2017-12-25 03:37:52 +00:00
leaving loe ".fs_00000000"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
pat adf $1==4 /* Add single */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FSREG FSREG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, FSREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fadds %a, %2, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
pat adf stl $1==4 && inreg($2)==reg_float
with FSREG FSREG
gen fadds {LOCAL, $2}, %2, %1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
pat sbf $1==4 /* Subtract single */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FSREG FSREG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, FSREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fsubs %a, %2, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
pat sbf stl $1==4 && inreg($2)==reg_float
with FSREG FSREG
gen fsubs {LOCAL, $2}, %2, %1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
pat mlf $1==4 /* Multiply single */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FSREG FSREG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, FSREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fmuls %a, %2, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
pat mlf stl $1==4 && inreg($2)==reg_float
with FSREG FSREG
gen fmuls {LOCAL, $2}, %2, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat dvf $1==4 /* Divide single */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FSREG FSREG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, FSREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fdivs %a, %2, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
pat dvf stl $1==4 && inreg($2)==reg_float
with FSREG FSREG
gen fdivs {LOCAL, $2}, %2, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat ngf $1==4 /* Negate single */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FSREG
uses reusing %1, FSREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fneg %a, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add more rules for single-precision reg_float. The result of single-precision fadds, fsubs, and such can go into a register variable, like we already do with double precision. This avoids an extra fmr from a temporary register to the regvar.
2017-10-17 21:53:03 +00:00
pat ngf stl $1==4 && inreg($2)==reg_float
with FSREG
gen fneg {LOCAL, $2}, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
/* When a or b is NaN, then a < b, a <= b, a > b, a >= b
* should all be false. We can't make them false, because
* - EM's _cmf_ is only for ordered comparisons.
* - The peephole optimizer assumes (a < b) == !(a >= b).
*
* We do make a == b false and a != b true, by checking the
* eq (equal) bit or un (unordered) bit in cr0.
*/
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat cmf $1==4 /* Compare single */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FSREG FSREG
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
uses REG={COND_FS, %2, %1}
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
/* Extract lt, gt, un; put lt in sign bit. */
gen andisX %a, %a, {C, 0xd000}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
yields %a
pat cmf teq $1==4 /* Single second == top */
with FSREG FSREG
uses REG={COND_FS, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XEQ, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmf tne $1==4 /* Single second == top */
with FSREG FSREG
uses REG={COND_FS, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XNE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmf tgt $1==4 /* Single second > top */
with FSREG FSREG
uses REG={COND_FS, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmf tge $1==4 /* Single second >= top */
with FSREG FSREG
uses REG={COND_FS, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmf tlt $1==4 /* Single second < top */
with FSREG FSREG
uses REG={COND_FS, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmf tle $1==4 /* Single second <= top */
with FSREG FSREG
uses REG={COND_FS, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
proc cmf4zxx example cmf zeq
Fix comparison of 4-byte floats. I broke it in f64b7d8. My stack pattern had the wrong type of registers. The comparison popped too many bytes and corrupted the stack.
2017-02-18 00:29:45 +00:00
with FSREG FSREG STACK
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
gen
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
fcmpo cr0, %2, %1
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
bxx* {LABEL, $2}
/* Pop 2 singles, branch if... */
pat cmf zeq $1==4 call cmf4zxx("beq")
pat cmf zne $1==4 call cmf4zxx("bne")
pat cmf zgt $1==4 call cmf4zxx("bgt")
pat cmf zge $1==4 call cmf4zxx("bge")
pat cmf zlt $1==4 call cmf4zxx("blt")
pat cmf zle $1==4 call cmf4zxx("ble")
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loc loc cff $1==4 && $2==8 /* Convert single to double */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FSREG
yields %1.1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loc loc cfi $1==4 && $2==4 /* Single to signed int */
Experiment with conversions between integers and floats. Switch some conversions from libem calls to inline code. The conversions from integers to floats are now too slow, because each conversion allocates 4 or 5 registers, and the register allocator is too slow. I might use these slow conversions to experiment with the register allocator. I add the missing conversions between 4-byte single floats and integers, simply by going through 8-byte double floats. (These replace the calls to nonexistant functions in libem.) I remove the placeholder for fef 4, because it doesn't exist in libem, and our language runtimes only use fef 8.
2017-02-12 04:23:47 +00:00
leaving
loc 4
loc 8
cff
loc 8
loc 4
cfi
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loc loc cfu $1==4 && $2==4 /* Single to unsigned int */
Experiment with conversions between integers and floats. Switch some conversions from libem calls to inline code. The conversions from integers to floats are now too slow, because each conversion allocates 4 or 5 registers, and the register allocator is too slow. I might use these slow conversions to experiment with the register allocator. I add the missing conversions between 4-byte single floats and integers, simply by going through 8-byte double floats. (These replace the calls to nonexistant functions in libem.) I remove the placeholder for fef 4, because it doesn't exist in libem, and our language runtimes only use fef 8.
2017-02-12 04:23:47 +00:00
leaving
loc 4
loc 8
cff
loc 8
loc 4
cfu
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loc loc cif $1==4 && $2==4 /* Signed int to single */
Experiment with conversions between integers and floats. Switch some conversions from libem calls to inline code. The conversions from integers to floats are now too slow, because each conversion allocates 4 or 5 registers, and the register allocator is too slow. I might use these slow conversions to experiment with the register allocator. I add the missing conversions between 4-byte single floats and integers, simply by going through 8-byte double floats. (These replace the calls to nonexistant functions in libem.) I remove the placeholder for fef 4, because it doesn't exist in libem, and our language runtimes only use fef 8.
2017-02-12 04:23:47 +00:00
leaving
loc 4
loc 8
cif
loc 8
loc 4
cff
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loc loc cuf $1==4 && $2==4 /* Unsigned int to single */
Experiment with conversions between integers and floats. Switch some conversions from libem calls to inline code. The conversions from integers to floats are now too slow, because each conversion allocates 4 or 5 registers, and the register allocator is too slow. I might use these slow conversions to experiment with the register allocator. I add the missing conversions between 4-byte single floats and integers, simply by going through 8-byte double floats. (These replace the calls to nonexistant functions in libem.) I remove the placeholder for fef 4, because it doesn't exist in libem, and our language runtimes only use fef 8.
2017-02-12 04:23:47 +00:00
leaving
loc 4
loc 8
cuf
loc 8
loc 4
cff
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
pat fef $1==4 /* Split fraction, exponent */
leaving cal ".fef4"
/* Multiply two singles, then split fraction, integer */
pat fif $1==4
leaving cal ".fif4"
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
/* Double-precision floating-point */
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat zrf $1==8 /* Push zero */
leaving lde ".fd_00000000"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat adf $1==8 /* Add double */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FREG FREG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, FREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fadd %a, %2, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat adf sdl $1==8 && inreg($2)==reg_float
with FREG FREG
gen fadd {DLOCAL, $2}, %2, %1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat sbf $1==8 /* Subtract double */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FREG FREG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, FREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fsub %a, %2, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat sbf sdl $1==8 && inreg($2)==reg_float
with FREG FREG
gen fsub {DLOCAL, $2}, %2, %1
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat mlf $1==8 /* Multiply double */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FREG FREG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, FREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fmul %a, %2, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat mlf sdl $1==8 && inreg($2)==reg_float
with FREG FREG
gen fmul {DLOCAL, $2}, %2, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat dvf $1==8 /* Divide double */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FREG FREG
Fix lim. Improve lxl, lxa, lor, str, procs with no locals. _lim_ must use _loe_ (load word external), not _lde_ (load double-word external). The new patterns for _lxl_, _lxa_, _lor_, _str_ emit shorter code in some cases. The change from GPR_EXPR to REG_EXPR allows moving LXFRAME to a register variable. Add more "reusing" clauses. We have enough registers that ncg almost never reuses a register, but sometimes it can reuse r3. In mach.c, emit one fewer instruction in procedures with no locals.
2018-01-05 01:40:35 +00:00
uses reusing %1, reusing %2, FREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fdiv %a, %2, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat dvf sdl $1==8 && inreg($2)==reg_float
with FREG FREG
gen fdiv {DLOCAL, $2}, %2, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat ngf $1==8 /* Negate double */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FREG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
uses reusing %1, FREG
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
gen fneg %a, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Add floating-point register variables to PowerPC ncg. Use f14 to f31 as register variables for 8-byte double-precison. There are no regvars for 4-byte double precision, because all regvar(reg_float) must have the same size. I expect more programs to prefer 8-byte double precision. Teach mach/powerpc/ncg/mach.c to emit stfd and lfd instructions to save and restore 8-byte regvars. Delay emitting the function prolog until f_regsave(), so we can use one addi to make stack space for both local vars and saved registers. Be more careful with types in mach.c; don't assume that int and long and full are the same. In ncg table, add f14 to f31 as register variables, and some rules to use them. Add rules to put the result of fadd, fsub, fmul, fdiv, fneg in a regvar. Without such rules, the result would go in a scratch FREG, and we would need fmr to move it to the regvar. Also add a rule for pat sdl inreg($1)==reg_float with STACK, so we can unstack the value directly into the regvar, again without a scratch FREG and fmr. Edit util/ego/descr/powerpc.descr to tell ego about the new float regvars. This might not be working right; ego usually decides against using any float regvars, so ack -O1 (not running ego) uses the regvars, but ack -O4 (running ego) doesn't use the regvars. Beware that ack -mosxppc runs ego using powerpc.descr but -mlinuxppc and -mqemuppc run ego without a config file (since 8ef7c31). I am testing powerpc.descr with a local edit to plat/linuxppc/descr to run ego with powerpc.descr there, but I did not commit my local edit.
2017-02-16 00:34:07 +00:00
pat ngf sdl $1==8 && inreg($2)==reg_float
with FREG
gen fneg {DLOCAL, $2}, %1
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
/* To compare NaN, see comment above pat cmf $1==4 */
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat cmf $1==8 /* Compare double */
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FREG FREG
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
uses REG={COND_FD, %2, %1}
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
/* Extract lt, gt, un; put lt in sign bit. */
gen andisX %a, %a, {C, 0xd000}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
yields %a
pat cmf teq $1==8 /* Double second == top */
with FREG FREG
uses REG={COND_FD, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XEQ, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
Add fef 4, fif 4. Improve fef 8, fif 8. Other float changes. When I wrote fef 8, I forgot to test denormalized numbers. Oops. Now fix two of my mistakes: - When checking for zero, `extrwi r6, r3, 22, 12` needs to be `extrwi r6, r3, 20, 12`. There are only 20 bits to extract. - After the multiplication by 2**64, I forgot to put the fraction in [0.5, 1) or (-1, 0.5] by setting IEEE exponent = 1022. Teach fif 8 about signed zero and NaN. In ncg/table, change cmf so NaN is not equal to any value, and comment why ordered comparisons don't work with NaN. Also add cost for fctwiz, remove extra `uses REG`. Edit comment in cfu8.s because the conditional branch might be before or after fctwiz.
2018-01-22 19:04:15 +00:00
pat cmf tne $1==8 /* Double second == top */
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
with FREG FREG
uses REG={COND_FD, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XNE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmf tgt $1==8 /* Double second > top */
with FREG FREG
uses REG={COND_FD, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmf tge $1==8 /* Double second >= top */
with FREG FREG
uses REG={COND_FD, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XGE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmf tlt $1==8 /* Double second < top */
with FREG FREG
uses REG={COND_FD, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLT, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
pat cmf tle $1==8 /* Double second <= top */
with FREG FREG
uses REG={COND_FD, %2, %1}
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
yields {XLE, %a}
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
proc cmf8zxx example cmf zeq
with FREG FREG STACK
gen
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
fcmpo cr0, %2, %1
Rewrite how PowerPC ncg does conditional branches and tests. The rewritten code rules bring 3 new features: 1. The new rules compare a small constant with a register by reversing the comparison and using `cmpwi` or `cmplwi`. The old rules put the constant in a register. 2. The new rules emit shorter code to yield the test results, without referencing the tables in mach/powerpc/ncg/tge.s. 3. The new rules use the extended `beq` and relatives, not the basic `bc`, in the assembly output. I delete the old tristate tokens and the old moves, because they confused me. Some of the old moves weren't really moves. For example, `move R3, C0` and then `move C0, R0` did not move r3 to r0. I rename C0 to CR0.
2017-01-26 00:08:55 +00:00
bxx* {LABEL, $2}
/* Pop 2 doubles, branch if... */
pat cmf zeq $1==8 call cmf8zxx("beq")
pat cmf zne $1==8 call cmf8zxx("bne")
pat cmf zgt $1==8 call cmf8zxx("bgt")
pat cmf zge $1==8 call cmf8zxx("bge")
pat cmf zlt $1==8 call cmf8zxx("blt")
pat cmf zle $1==8 call cmf8zxx("ble")
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
/* Convert double to single */
/* reg_float pattern must be first, or it goes unused! */
pat loc loc cff stl $1==8 && $2==4 && inreg($4)==reg_float
with FREG
gen frsp {LOCAL, $4}, %1
pat loc loc cff $1==8 && $2==4
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
with FREG
uses reusing %1, FSREG
Add more chances to put results in register variables. When a rule `uses REG ... yields %a`, the result %a is always a temporary, never a regvar. If the EM code uses _stl_ to put the result in a regvar, then ncg emits _mr_ to move %a to the regvar. There are two ways to put the result in the regvar without %a: 1. Yield a token, as in `yields {MUL_RR, %2, %1}`, so that _stl_ can move the token to the regvar without using %a. 2. Provide a pattern, like `sli stl`, that just puts the result in `{LOCAL, $2}` and not %a. Allow some tokens, like SUM_RIS and XEQ, onto the stack; and add tokens like MUL_RR, and patterns like `sli stl`. Delete patterns for `stl lol` and `sdl ldl` to avoid an extra temporary %a when the local is a regvar. Delete `lal sti lal loi` because it would emit wrong code.
2017-12-08 22:19:26 +00:00
gen frsp %a, %1
Enable the Hall check again, and get powerpc to pass it. Upon enabling the check, mach/powerpc/ncg/table fails to build as ncgg gives many errors of "Previous rule impossible on empty stack". David Given reported this problem in 2013: https://sourceforge.net/p/tack/mailman/message/30814694/ Commit c93cb69 commented out the error in util/ncgg/cgg.y to disable the Hall check. This commit enables it again. In ncgg, the Hall check is checking that a rule is possible with an empty fake stack. It would be possible if ncg can coerce the values from the real stack to the fake stack. The powerpc table defined coercions from STACK to {FS, %a} and {FD, %a}, but the Hall check didn't understand the coercions and rejected each rule "with FS" or "with FD". This commit removes the FS and FD tokens and adds a new group of FSREG registers for single-precision floats, while keeping FREG registers for double precision. The registers overlap, with each FSREG containing one FREG, because it is the same register in PowerPC hardware. FS tokens become FSREG registers and FD tokens become FREG registers. The Hall check understands the coercions from STACK to FSREG and FREG. The idea to define separate but overlapping registers comes from the PDP-11 table (mach/pdp/ncg/table). This commit also removes F0 from the FREG group. This is my attempt to keep F0 off the fake stack, because one of the stacking rules uses F0 as a scratch register (FSCRATCH).
2016-09-18 19:08:55 +00:00
yields %a
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loc loc cfi $1==8 && $2==4 /* Double to signed int */
Experiment with conversions between integers and floats. Switch some conversions from libem calls to inline code. The conversions from integers to floats are now too slow, because each conversion allocates 4 or 5 registers, and the register allocator is too slow. I might use these slow conversions to experiment with the register allocator. I add the missing conversions between 4-byte single floats and integers, simply by going through 8-byte double floats. (These replace the calls to nonexistant functions in libem.) I remove the placeholder for fef 4, because it doesn't exist in libem, and our language runtimes only use fef 8.
2017-02-12 04:23:47 +00:00
with FREG STACK
uses reusing %1, FREG
Archival checkin (semi-working code).
2007-11-02 18:56:58 +00:00
gen
Experiment with conversions between integers and floats. Switch some conversions from libem calls to inline code. The conversions from integers to floats are now too slow, because each conversion allocates 4 or 5 registers, and the register allocator is too slow. I might use these slow conversions to experiment with the register allocator. I add the missing conversions between 4-byte single floats and integers, simply by going through 8-byte double floats. (These replace the calls to nonexistant functions in libem.) I remove the placeholder for fef 4, because it doesn't exist in libem, and our language runtimes only use fef 8.
2017-02-12 04:23:47 +00:00
fctiwz %a, %1
Speed up register allocation by removing some register classes. The table for PowerPC had placed each GPR and FPR into an individual register class (like GPR3, GPR4, FPR1, FPR2), and had used these classes to coerce stack values into specific registers. But ncg does not like having many register classes. In http://tack.sourceforge.net/olddocs/ncg.pdf Hans van Staveren wrote: > Every extra property means the register set is more unorthogonal and > *cg* execution time is influenced by that, because it has to take > into account a larger set of registers that are not equivalent. So > try to keep the number of different register classes to a minimum. Recent changes to the PowerPC table have removed many coercions to specific registers. Many functions in libem switched from taking values in registers to taking them from the stack (see dc05cb2). I now remove all 64 individual register classes of GPR and FPR. In the few cases where I need a stack value in a specific register, I now do a move (as the arm and m68020 tables do). This commit speeds the compilation of some files. For my test file fconv.c, the compilation time goes from over 20 seconds to under 1 second. My fconv.c has 4 conversions from floats to integers, and the table has my experimental rules that do the conversions by allocating 4 or 5 registers.
2017-02-13 22:44:46 +00:00
stfdu %a, {IND_RC_D, sp, 0-8}
Split C from CONST. Rename token CONST to C. Define set CONST = C + CONST_STACK. The instructions with CONST operands can now accept CONST_STACK tokens; some cases of {CONST, %1.val} become %1. Also simplify two of _rlwinm_ into _slwi_ and _srwi_.
2017-12-08 00:24:09 +00:00
addi sp, sp, {C, 4}
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loc loc cfu $1==8 && $2==4 /* Double to unsigned int */
leaving cal ".cfu8"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loc loc cif $1==4 && $2==8 /* Signed int to double */
leaving cal ".cif8"
Remove trailing whitespace in mach/powerpc/ncg/table
2016-12-09 21:36:42 +00:00
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
pat loc loc cuf $1==4 && $2==8 /* Unsigned int to double */
leaving cal ".cuf8"
Rewrite .fif8 to avoid powerpc64 fctid This fixes the SIGILL (illegal instruction) in startrek when firing phasers. The 32-bit processors in my PowerPC Mac and in QEMU don't have fctid, a 64-bit instruction. I got the idea from mach/proto/fp/fif8.c to extract the exponent, clear some bits to get an integer, then subtract the integer from the original value to get the fraction.
2016-10-17 04:39:59 +00:00
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
pat fef $1==8 /* Split fraction, exponent */
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
leaving cal ".fef8"
Rewrite .fif8 to avoid powerpc64 fctid This fixes the SIGILL (illegal instruction) in startrek when firing phasers. The 32-bit processors in my PowerPC Mac and in QEMU don't have fctid, a 64-bit instruction. I got the idea from mach/proto/fp/fif8.c to extract the exponent, clear some bits to get an integer, then subtract the integer from the original value to get the fraction.
2016-10-17 04:39:59 +00:00
Change .fef8 and .fif8 to pass values on the stack. Reorder the code in .fef8 and .fif8 so that in the usual case, we fall through to the blr without taking any branches. The usual case, by my guess, is .fef8 with normalized numbers or .fif8 with small integers. I change .fef8 and .fif8 to pass values on the real stack, not in specific registers. This simplifies the ncg table, and might help me experiment with changes to the ncg table. This change might or might not help mcg. Seems that mcg always uses the stack to pass values to libem, but I have not tested .fef8 or .fif8 with mcg.
2017-02-12 21:44:37 +00:00
/* Multiply two doubles, then split fraction, integer */
pat fif $1==8
Remove INT32 and such. Adjust indentation. I understand `loi 4` more easily than `loi INT32`, because `loi 4` appears in .e files. So remove INT8, INT16, INT32, INT64. Add a comment to explain r3 during unconditional jumps.
2017-12-23 02:18:58 +00:00
leaving cal ".fif8"
Reference in a new issue Copy permalink