d0p1/ack
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity
7269 commits 2 branches 0 tags 43 MiB
Commit graph

2432 commits

Author SHA1 Message Date
David Given f8fa3ece42 inn on ncg now passes the CPU tests. 2016-11-20 19:35:34 +01:00
David Given 953c08839f inn works now; add a helper for it. 2016-11-20 12:53:44 +01:00
David Given 196fa914b3 lxa now works, I hope; traps are better (and stubbed out on qemuppc). 2016-11-20 11:57:21 +01:00
David Given d5328492d7 Better handling of float conversions; more tests; converting to unsigned ints 
works now.
 2016-11-20 11:27:40 +01:00
David Given 454a7494bb cif8 and cuf8 work now. More tests. 2016-11-19 11:42:30 +01:00
David Given cc660b230f Floats and doubles are now written out correctly. 2016-11-19 11:39:13 +01:00
David Given d31bc6a3f9 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 10:55:41 +01:00
David Given 5208e5f751 Yet another OB1 stack format fix. 2016-11-19 10:42:22 +01:00
David Given 43439c6d0c Remember to push the result of lor onto the stack. 2016-11-17 22:04:32 +01:00
David Given 81bc2c74c5 A bb's regsin are no longer the same as those of its first instruction; 
occasionally the first hop of a block would try to rearrange its registers (due
to evicted throughs), resulting in the phi moves copying values into the wrong
registers.
 2016-11-16 20:52:15 +01:00
David Given 581fa4a457 Reenable eviction of corrupted registers, which had been broken by a previous 
change. Change the register move code to get swaps right, or at least righter.
 2016-11-15 21:55:10 +01:00
David Given 86c832ef86 Put saved registers in *actually* the write place. I hope. 2016-11-15 21:54:15 +01:00
David Given cc686ded62 Get subtractions the right way round. 2016-11-15 20:25:11 +01:00
David Given 0289b1004e Allow values left on the stack at the end of the procedure (it's legal!). 2016-11-14 21:47:49 +01:00
David Given e7132183fb Fix buffer overrun: if LABEL_STARTER is seen but LABEL_TERMINATOR is not, the 
label parser will keep going forever looking for the end of the label. It now
stops at the end of the string.
 2016-11-13 14:04:58 +01:00
David Given 852d3a691d Update the table to return call output values in the right registers. Fix the 
register allocator so the corrupted registers only apply to throughs
(otherwise, you can't put output registers in corrupted registers).
 2016-11-11 21:48:36 +01:00
David Given b5c1d622f5 Rework the way stack frames are laid out to be simpler and, hopefully, more 
correct. Saved registers are now placed in what may be the right place.
 2016-11-11 21:17:45 +01:00
David Given 84ee75ec07 Merge from default. 2016-11-11 20:17:54 +01:00
David Given d82df74a7a Rename addr_t to address_t to avoid clashes with the system addr_t. 2016-11-11 20:17:10 +01:00
David Given fd91851005 Add enough return types to the K&R C that the ACK builds (on Linux) using clang 
now.
 2016-11-10 22:04:18 +01:00
David Given 4fa2c94a4a Correctly mangle labels used in initialisers. 2016-10-31 23:21:33 +01:00
David Given 9261cd978d Typo fix. 2016-10-31 23:16:02 +01:00
David Given 941072e0d7 Add, I hope, patterns for fmsub, fnmadd, and fnmsub (also float versions). 2016-10-31 22:36:54 +01:00
David Given 44f0cea6ca Also use fmadd for single-precision floats. 2016-10-31 19:55:16 +01:00
David Given 064d1a5d5d Use fmadd for multiply-and-add instructions. 2016-10-31 19:52:17 +01:00
David Given e19850b114 Fix a few c11isms. 2016-10-30 16:51:06 +01:00
David Given ca5b6e07bb Properly export symbols. 2016-10-29 23:52:17 +02:00
David Given 8c3670483f Get top working with the PowerPC; use it to eliminate useless branches and 
moves.
 2016-10-29 23:37:11 +02:00
David Given a8c4dac67c Merge from default (merging in George Koehler's PowerPC changes). 2016-10-29 22:40:40 +02:00
David Given a311e61360 Add support for preserved registers. 2016-10-29 20:22:44 +02:00
David Given e3ebf986e9 More opcodes. 2016-10-29 13:32:09 +02:00
David Given 1ae8b90238 More opcodes. 2016-10-29 12:55:34 +02:00
David Given acaae765af Emit negative constants correctly. 2016-10-29 12:55:21 +02:00
David Given 61349389fb More opcodes. sti can now cope with non-standard sizes (really need a better 
fix for this). Hack in crude support for mismatched stack pushes and pops (ints
vs longs).
 2016-10-29 12:48:05 +02:00
David Given 68419da235 Actually, the locals need to go above the spills and saved regs, so fp == lb. 2016-10-29 12:00:33 +02:00
David Given 2cc2c0ae98 Lots more opcodes. Rearrange the stack layout so that fp->ab is a fixed value 
(needed for CHAINFP and FPTOAB). Wire up lfrs to calls via a phi when
necessary, to allow call-bra-lfr chains.
 2016-10-29 11:57:56 +02:00
David Given bfa65168e2 Don't generate phis if unnecessary (because this breaks the 
critical-edge-splitting guarantee and causes insertion of phi copies to fail).
 2016-10-29 10:55:48 +02:00
David Given 658db4ba71 Mangle label names (turns out that the ACK assembler can't really cope with 
labels that are the same name as instructions...).
 2016-10-27 23:17:16 +02:00
David Given 81525c0f2c Swaps work (at least for registers). More opcodes. Rearrange the stack layout 
so we can always trivially find fp, which lets CHAINFP work.
 2016-10-27 21:50:58 +02:00
David Given be3dece5af Allow emission of strings containing ". 2016-10-27 21:48:46 +02:00
David Given 51bd3ee4dd Fix bug where some phis weren't being inserted when a given variable definition 
needed more than one phi (due to the dominance frontier containing more than
one basic block).
 2016-10-27 21:40:25 +02:00
David Given 9977ce841a Remove the bytes1, bytes2, bytes4, bytes8 attributes; remove the concept of a 
register 'type'; now use int/float/long/double throughout to identify
registers. Lots of register allocator tweaks and table bugfixes --- we now get
through the dreading Mathlib.mod!
 2016-10-25 23:04:20 +02:00
David Given 45a7f2e993 Phi copies are now inserted as part of type inference. More opcodes. 2016-10-24 22:14:08 +02:00
David Given 111c13e253 More opcodes. 2016-10-24 20:15:22 +02:00
David Given a4644dee4d More opcodes. 2016-10-24 12:08:40 +02:00
David Given b22780c075 More opcodes, including the difficult and fairly stupid los/sts. 2016-10-23 22:24:08 +02:00
David Given abd0cedd61 Massive change to how IR types are handled; we use the type code for matching 
rather than the size. Much cleaner and simpler.
 2016-10-23 21:54:14 +02:00
David Given b1a3d76d6f Re-re-add the type inference layer, now I know more about how things work. 
Remove that terrible float promotion code.
 2016-10-22 23:04:13 +02:00
David Given 11b0bc1055 More opcodes. 2016-10-22 20:32:51 +02:00
David Given 2d52b1fdaa Remove GETRET; values are now returned directly by CALL. Fix a bug in 
convertstackops which was resulting in duplicate IR groups.
 2016-10-22 12:13:57 +02:00
David Given ceb938fb3c More opcodes. 2016-10-22 11:26:28 +02:00
David Given 7ae888b754 Hacky workaround the way the Modula-2 compiler generates non-standard sized 
loads and saves. More opcodes; simplified table using macros.
 2016-10-22 10:48:22 +02:00
David Given 90d0661639 Typo fix. 2016-10-22 00:48:55 +02:00
David Given f851ab83af Better (and more correct) floating point conversions; fif; various new opcodes. 2016-10-22 00:48:26 +02:00
David Given d535be87b1 fef4 and fef8 is now cleaner, albeit slower; add some more register alias 
stuff.
 2016-10-22 00:02:15 +02:00
David Given 4db402f229 Add (pretty crummy) support for register aliases and static pairs of registers. 
We should have enough functionality now for rather buggy 8-bit ints and
doubles. Rework the table and the platform.c to match.
 2016-10-21 23:31:00 +02:00
David Given e4fec71f9c Lots more opcodes; better eviction behaviour; better register moves. Lots more 
PowerPC stuff (some working).
 2016-10-19 23:29:05 +02:00
David Given ffb1eabf45 Floating point promotion is less buggy. 2016-10-19 23:27:53 +02:00
George Koehler 99dee0ad24 Remove f14 to f31 from FREG and FSREG. 
This would have happened later, if f14 to f31 became regvar (like r13
to r31 are now).  I am doing it now because ncg is too slow for rules
"with FREG FREG uses FREG".  We use such rules for adf 8 and other EM
instructions that operate on 2 floats.  Like my last commit cfbc537,
this commit speeds ncg by removing choices for register allocation.
 2016-10-18 21:16:47 -04:00
David Given d5071e7df1 Promote values accessed via NOP. 2016-10-18 23:58:03 +02:00
David Given 5413d47029 '!' tracing is now always emitted; tracing goes to stderr. 2016-10-18 22:32:09 +02:00
David Given 3520704ea8 Add support for floating point constants. 2016-10-18 22:29:42 +02:00
George Koehler cfbc537959 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.
 2016-10-17 20:31:59 -04:00
David Given 938fb8c2fc Lots more opcodes. 2016-10-18 00:31:26 +02:00
David Given 4a093b9eba Add li and mr pseudoinstructions. 2016-10-18 00:21:32 +02:00
George Koehler c7b68033ef Add costs to powerpc instructions. 
Also show how andi., andis., or., set condition codes.
 2016-10-17 14:57:21 -04:00
George Koehler f33b30ed3c 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 00:39:59 -04:00
George Koehler e2ccc8f942 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 18:13:39 -04:00
David Given 5f0164db62 Bolt mcg into the PowerPC backend. It doesn't build yet, but it is generating 
*some* code.
 2016-10-17 00:06:06 +02:00
David Given d539389e81 Merge in the unfinished PowerPC branch. 2016-10-16 22:38:27 +02:00
David Given 1e17921208 Implement saving of dirty registers onto the stack. 2016-10-16 22:37:42 +02:00
George Koehler 19f0eb86a4 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 16:33:24 -04:00
George Koehler 5b5f774a64 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 16:02:25 -04:00
George Koehler 7c64dab491 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 13:58:54 -04:00
George Koehler baa152217e Remove unused parts of mach/powerpc/ncg/table 
Remove unused tokens GPRINDIRECTLO, HILABEL, LOLABEL, LABELI.  Also
remove an #if 0 ... #endif group of patterns.
 2016-10-15 20:00:48 -04:00
David Given 6a23906ad8 Various bits of cleanup; we should almost be ready to try sending this to the 
assembler soon...
 2016-10-15 23:39:38 +02:00
David Given 286435a2ed Oops, forgot to add the output option spec to the string! 2016-10-15 23:34:54 +02:00
David Given b36897c299 References to the stack frame are now rendered properly. 2016-10-15 23:33:30 +02:00
David Given a8ee82d197 Stop passing proc around, and use a global instead --- much cleaner. 2016-10-15 23:19:44 +02:00
David Given 7aa60a6451 Register spilling to the stack frame works, more or less. 2016-10-15 22:53:56 +02:00
David Given 0eb32e7553 Fix yet another bug to do with IR register outputs. 2016-10-15 19:14:25 +02:00
David Given 9504aec2bd Function termination gets routed through an exit block; we now have prologues 
and epilogues. mcgg now exports some useful data as headers. Start factoring
out some of the architecture-specific bits into an architecture-specific file.
 2016-10-15 18:38:46 +02:00
David Given 5ad3aa8595 Add a pile of new instructions used by Pascal; I'm going to need to think about 
how locals and the local base are handled.
 2016-10-15 13:07:59 +02:00
David Given 358c44de35 Bytes were sometimes failing to be sign extended correctly. 2016-10-15 12:11:40 +02:00
David Given 517120d0fb Allow asm names for registers which are different from the friendly names shown 
in the tracing (because PowerPC register names are just numbers).
 2016-10-15 11:42:47 +02:00
David Given b2ddf12473 Some more opcodes. 2016-10-15 11:22:40 +02:00
George Koehler 29cb008faa 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.
 2016-10-14 23:59:26 -04:00
David Given bb17aea73a You can now mark a register as corrupting a certain register class; calls work, 
or at least look like they work. The bad news is that the register allocator
has a rare talent for putting things in the wrong register.
 2016-10-15 01:15:08 +02:00
David Given 886adb86d7 Log empty hops. 2016-10-14 23:19:25 +02:00
David Given 4f2177e41f Reworked loads and stores; it's now *different*, maybe not better. 2016-10-14 23:19:02 +02:00
David Given a63052427e Factor out the register allocation routines to make them easier to deal with. 2016-10-14 23:17:06 +02:00
David Given bb53a7fb51 Fix stupid issue where hop output registers were being overwritten, leading to 
invalid SSA form.
 2016-10-14 23:12:29 +02:00
David Given 98fe70a7de Output register equality constraints work. 2016-10-14 22:17:02 +02:00
David Given 216ff5cc43 Make loads and stores in the table nicer; fix a place where it looked like it 
was working but only accidentally.
 2016-10-12 23:12:53 +02:00
David Given f06b51c981 Keep track of register types as well as attributes --- the type being how we 
find new registers when evicting values. Input constraints work (they were
being ignored before). Various bug fixing so they actually work.
 2016-10-12 22:58:46 +02:00
David Given 4723a1442f Add code to remove unused phis, converting to pruned SSA form, to avoid 
confusing the register allocator later.
 2016-10-12 21:50:12 +02:00
David Given df239b3f90 Don't allow the same IR to be added to the sequence list more than once 
(sometimes happens because op_dup, but makes no sense).
 2016-10-12 00:45:36 +02:00
David Given 96dffd2007 Clean up the allocator a bit, in preparation for making it lots more 
complicated; no semantic changes.
 2016-10-11 23:17:30 +02:00
David Given 668cccdff1 A few more opcodes. 2016-10-11 00:29:18 +02:00
David Given 2be1c51885 A little fiddling with store instructions. The PowerPC is not friendly to 
iburg.
 2016-10-11 00:23:35 +02:00