Commit graph

8015 commits

Author SHA1 Message Date
George Koehler 66f93f08c5 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 14:04:15 -05:00
George Koehler f1304e1a3c Use extended mnemonics and ha16/lo16.
Remove wrong comment: that's a right shift, not a left shift.
2018-01-11 20:04:27 -05:00
George Koehler 103d44c27c Rewrite sigaction() to prevent another race.
A signal handler might call sigaction().  We must block all signals,
not only our signal, to prevent a race between us and the next signal
handler.

Use /* comments */ because cpp might expand macros in ! comments
though such expansion is probably harmless.

The bridge is now shorter by 2 instructions.
2018-01-11 17:59:02 -05:00
George Koehler 2b09d3756c These are EM trap numbers.
Remove .sect; absolute symbols don't go in a section.
2018-01-09 00:40:11 -05:00
George Koehler de2c7c3f25 Hide some i386 stuff from linux68k, linuxppc.
Rename plat/linux/libsys/errno.s to plat/linux386/libsys/trapno.s and
stop building it for linux68k and linuxppc.  It defines symbols for
mach/i386/libem.

In syscalls.h, the numbers after 165 are only for i386, so hide them
from 68k, ppc.  These numbers are unused, because the system calls now
in libsys use the lower numbers.

Also teach the build system that libsys depends on the internal
headers in plat/linux/libsys/*.h
2018-01-08 23:46:30 -05:00
George Koehler 64b50b3a45 Shrink .cfu8
With my PowerBook G4, a program that converts values from 1.0 to
4000000.0 runs in about 0.32s with the old .cfu8 and 0.29s with this
shrunken .cfu8

Leave a comment about other ways to implement .cfu8
2018-01-07 16:03:55 -05:00
George Koehler b90c97b00b 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 17:55:50 -05:00
George Koehler 720af48d8a 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.
2018-01-04 20:40:35 -05:00
George Koehler d6938108a6 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 14:51:14 -05:00
George Koehler 26de4c1ab1 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_.
2017-12-24 22:37:52 -05:00
George Koehler 5f2a7b260f 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
2017-12-22 22:32:16 -05:00
George Koehler c964eeddba 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.
2017-12-22 21:18:58 -05:00
George Koehler f96f918a29 Generate shorter code for ret 4 and ret 8. 2017-12-22 20:37:39 -05:00
George Koehler 5867ca2f2c Remove two obsolete patterns.
These patterns seem to have no effect on the generated code.
2017-12-22 19:57:42 -05:00
George Koehler 2eeee36f78 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 17:04:16 -05:00
George Koehler 4bb31c296d Revise the comments in the EM tests.
You can cheat these tests if _cms_ and _cmu_ always push zero.
2017-12-21 18:24:28 -05:00
George Koehler 787fdeaaa9 Add some tests for Modula-2. 2017-12-21 18:24:17 -05:00
George Koehler aa9418c029 Pass 4 bytes to fail(uint32_t)
This would become necessary if something failed on a platform with
16-bit int (EM_WSIZE == 2).

Remove unreachable `ret 0`.  If reached, it wouldn't work to return
from _m_a_i_n.
2017-12-18 21:58:57 -05:00
George Koehler a5e8dc8a06 Simplify code by using cms EM_WSIZE to compare bytes.
This should work because the C compiler does it.
2017-12-18 21:52:13 -05:00
George Koehler a4e6595032 Remove '\0' from output. Fix a compiler warning.
Don't output '\0' in "@@FINISHED\0".

Cast code to unsigned int.  This helps platforms with 16-bit int, by
doing only the low 16 bits of the bitwise-and.  It also removes the
"(warning) conversion of long to pointer loses accuracy".
2017-12-18 21:17:42 -05:00
George Koehler ad47fa5fe3 Add splitting coercions for IND_ALL_D.
Delete my wrong comment (from commits cfbc537, a8f62f4, 5432bd0) which
claimed that such coercions are not possible.
2017-12-18 20:59:04 -05:00
George Koehler 24abaf6a25 Enable conditional expressions in splitting coercions.
ncgg has parsed the optional conditional expression (optexpr) of each
splitting coercion since commit 72b83cc in 1985; but for almost 33
years, ncg has ignored the expression in c2_expr.

Few tables had conditional coercions (I only found them in arm and
m68020), and no tables had conditional splitting coercions, so this
only becomes a problem now as I try to add a conditional splitting
coercion to powerpc.
2017-12-18 20:39:56 -05:00
George Koehler 5e99baabdf Rename two tokens. CONST_HZ was not hertz (Hz). 2017-12-18 12:36:10 -05:00
George Koehler d8fa9d1b2a In coercions, try to reuse a register with the same token.
This reduces code size.
2017-12-17 12:45:27 -05:00
George Koehler b0d75fed37 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.
2017-12-17 11:56:02 -05:00
George Koehler 5ba83100d6 Delete rules for sti 8 with REG IND_RC_D, with REG IND_RR_D.
Prefer the rule with REG FREG, by coercing IND_RC_D or IND_RR_D to
FREG.  This rule looks better to ncg.  When ncg chose between coercion
to REG IND_RC_D or coercion to REG FREG, it chose REG FREG.  It only
chose REG IND_RC_D if the stack had exact REG IND_RC_D.
2017-12-12 13:36:43 -05:00
George Koehler 11a54e0a7c These instructions write to the CR. 2017-12-10 14:01:14 -05:00
George Koehler 504d2aa34e 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 18:57:10 -05:00
George Koehler 48788287b8 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 17:19:26 -05:00
George Koehler 6b933db90b 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_.
2017-12-07 19:24:09 -05:00
George Koehler a1d1f38691 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 17:16:21 -05:00
George Koehler c95bcac91d Correct the stack pointer when i80 shrinks an integer.
The code used `sphl` to set the stack pointer, but the correct value
was in de, not hl.  Fix by swapping the values of de and hl, so `sphl`
is now correct.  When we shrink an integer from 4 to 2 bytes, both
registers de and hl point to copies of the result, but only one
register preserves the stack below the result.

This fixes writehex() in tests/plat/lib/test.c, when I compile it with
ack -mcpm, so it preserves the pointer to "0123456789abcdef", so it
writes hexadecimal digits and not garbage.

This bug goes back to commit 157b243 of Mar 18, 1985, so the bug is
32 years old, and probably the oldest bug that I ever fixed.
2017-12-07 15:39:41 -05:00
George Koehler 34cf0c8b63 Kill registers a, de, when i80 ncg does Call libem.
I compiled tests/plat/lib/test.c with ack -mcpm, but i80 ncg did emit
wrong code in writehex(uint32_t) for

    "0123456789abcdef"[code & 0xf]

The code called '.and' to evaluate `code & 0xf`, then tried to call
'.cii' to narrow the result from 4 to 2 bytes, but it passed garbage
instead of 4 to '.cii'.  The rule for '.and' was

    pat and defined($1)
    kills ALL
    uses dereg={const2,$1}
    gen Call {label,".and"}

This failed to kill register de={const2,4}, so ncg pushed de,
expecting to push 4, but actually pushing garbage.

Fix such rules using `mvi a,...` or `lxi de,...` so ncg doesn't track
the token in the register.  This is like the i86 table.  A different
fix would use a dummy instruction `killreg a` or `killreg de` like the
m68020 table.

Also correct 1 to $1 when calling '.exg'.
2017-12-06 22:14:00 -05:00
George Koehler 88207db638 Use <stdarg.h> in util/misc/convert.c
I made a syntax error in some .e file, and em_encode dumped core
because a 64-bit pointer didn't fit in a 32-bit int.  Now use stdarg
to pass pointers to error() and fatal().

Stop using the number of errors as the exit status.  Many systems use
only the low 8 bits of the exit status, so 256 errors would become 0.

Also change modules/src/print to accept const char *buf
2017-12-06 17:09:12 -05:00
David Given 6d91bdbbbd
Merge pull request #69 from kernigh/kernigh-stdc
use libc assert, strcmp; declare more functions; fewer clang warnings
2017-11-19 12:00:40 +01:00
David Given fb90b7b8d8
Merge pull request #66 from davidgiven/dtrg-warnings
lang/basic/lib: fewer warnings
2017-11-19 11:39:45 +01:00
George Koehler 760da1f421 Fix build with gcc.
gcc gave an error because the `char *` parameter doesn't match the
`const char *` in the prototype of regsave().  clang didn't give an
error.  I added the prototype in commit 5301cce.
2017-11-17 17:52:37 -05:00
George Koehler 11d48be49e Fix my typo from commit 5bbbaf4. 2017-11-17 15:46:24 -05:00
George Koehler d99a0682fc Switch ego to libc <assert.h>
I also tried, in types.h, to switch ego to libc <stdbool.h>, but that
causes an infinite loop in the IL phase.
2017-11-15 19:48:53 -05:00
George Koehler 9037d137f5 Add prototypes, void in util/ego/share
This uncovers a problem in il/il_aux.c: it passes 3 arguments to
getlines(), but the function expects 4 arguments.  I add FALSE as the
4th argument.  TRUE would fill in the list of mesregs.  IL uses
mesregs during phase 1, but this call to getlines() is in phase 2.
TRUE would leak memory unless I added a call to Ldeleteset(mesregs).
So I pass FALSE.

Functions passed to go() now have a `void *` parameter because
no_action() now takes a `void *`.
2017-11-15 17:19:56 -05:00
David Given caa9df5449 LLgen won't update the output file timestamp if the file contents doesn't
change, which confuses ninja no end. Fix this.

Fixes: #68
2017-11-15 19:41:39 +01:00
George Koehler 5bbbaf4919 Use size_t and void with memory allocation in ego.
alloc.h now needs to #include <stdlib.h> to find type size_t and
function free().
2017-11-14 20:35:18 -05:00
George Koehler 87a2315037 strcmp, strncmp are in <string.h>
*Important:*  Do `make clean` to work around a problem and prevent
infinite rebuilds, https://github.com/davidgiven/ack/issues/68

I edit tokens.g in util/LLgen/src, so I regenerate tokens.c.  The
regeneration script bootstrap.sh can't find LLgen, but I can run the
same command by typing the path to llgen.
2017-11-14 17:35:35 -05:00
George Koehler 229b80a004 Free buf in GetFile().
aprintf() returns a const char *; the assignment to char * caused both
clang and gcc to warn of the dropped const.

Commit 893471a introduced a tiny memory leak, because GetFile()
stopped freeing buf.  The const return type of aprintf() suggests that
the buffer must not be freed.

Now use Malloc() to allocate the buffer and free() to free it.  This
also checks if we are out of memory, because Malloc() does the check
and aprintf() currently doesn't.
2017-11-13 21:34:31 -05:00
George Koehler fbff3a4790 Check each format string in tabgen.c
Silence warning from clang at `if (ch2 = ...)`

Delete `|| rm %{outs}` in build.lua, because it hid the exit status of
tabgen, so if tabgen failed, the build continued and failed later.
2017-11-13 20:59:03 -05:00
George Koehler 974fa93491 Silence warning about pointer cast to int.
This cast was safe because the pointer is a small constant integer,
but it causes warning from gcc.
2017-11-13 20:58:57 -05:00
George Koehler ddf9de3f49 strcpy() is in <string.h> 2017-11-13 17:57:02 -05:00
George Koehler 5301cceee3 Declare machine-dependent functions in mach/proto/ncg
This breaks all machines because the declared return type void
disagrees with the implicit return type int (when I compile mach.c
with clang).  Unbreak i386, i80, i86, m68020, powerpc, vc4 by adding
the return types to mach.c.  We don't build any other machines; they
are broken since commit a46ee91 (May 19, 2013) declared void prolog()
and commit fd91851 (Nov 10, 2016) declared void mes(), with both
declarations in mach/proto/ncg/fillem.c.

Also fix mach/vc4/ncg/mach.c where type full is long, so fprintf()
must use "%ld" not "%d" to print full nlocals.
2017-11-13 14:23:44 -05:00
George Koehler e04166b85d More prototypes, less register in mach/proto/ncg
Files that #include "equiv.h" must do so after including "data.h", now
that a function prototype in equiv.h uses type rl_p from data.h.

Adjust style, changing some `for(...)` to `for (...)`.  The style in
mach/proto/ncg is less than consistent; the big annoyance now is that
some files want tabs at 4 spaces, others want tabs at 8 spaces.
2017-11-13 12:44:17 -05:00
George Koehler 909b0d5bf3 Add prototypes to functions in subr.c
Put the declarations in "data.h", because that header declares the
types cost_t and token_p.  Also #include <cgg_cg.h> from "data.h" to
get types c3_p and set_p, and guard <cgg_cg.h> against multiple
inclusion.
2017-11-12 16:11:05 -05:00