0899934220
cpu.x86: use full set of 8-bit, 16-bit and 32-bit registers on x86-64 to avoid clumsy save/restore logic
2009-07-29 21:56:37 -05:00
7831293fda
cpu.x86.assembler: move operands to operands sub-vocabulary, clean up small-reg-* code in compiler backend
2009-07-29 21:44:08 -05:00
c1fd97d515
Merge branch 'dcn'
2009-07-28 12:37:45 -05:00
4c664a469a
SSE4 opcodes for x86 assembler
2009-07-28 12:19:37 -05:00
afd914c808
Merge branch 'master' into dcn
2009-07-28 11:20:43 -05:00
1fe11f7c87
SSE1–SSSE3 opcodes + branch hints for x86 assembler
2009-07-28 00:22:27 -05:00
f0a5ac3fbb
cpu.x86: compile a load of zero, and adds, subs where dst = src1 more efficiently
2009-07-27 22:27:54 -05:00
39a70db831
Improve code generation for shift word: add intrinsics for fixnum-shift-fast in the case where the shift count is not constant, transform 1 swap shift into a more overflow check with open-coded fast case, transform bitand into fixnum-bitand in more cases
2009-07-16 23:50:48 -05:00
99faf3c79f
Overflowing fixnum intrinsics now expand into several CFG nodes. This speeds up the common case since only the uncommon case is now a stack syncpoint
2009-07-16 18:29:40 -05:00
1eae4286cd
compiler.cfg: split off condition codes into a comparisons sub-vocabulary
2009-07-13 14:42:52 -05:00
27c0577c91
cpu.x86.32: don't emit sub %esp,0x0 in prologue on Linux and Windows
2009-07-01 18:13:45 -05:00
554559c0b1
%dispatch: sometimes the generated sequence is one byte longer, so instead of hard-coding it, compute the right length
2009-06-30 18:11:15 -05:00
4782c737ab
cpu.x86: don't clobber src in %dispatch
2009-06-30 16:47:22 -05:00
a61a992bfd
cpu.x86.assembler: IMUL2 instruction was busted for immediate operands
...
When given a register and an immediate, it would generate imul imm,dst,dst however the 64-bit prefix was generated wrong and if dst was an extended register only the first operand would be an extended register. To fix this, change IMUL2 to not work on immediates anymore, and added a new IMUL3 that takes a destination register, source register, and immediate. Also, change compiler.cfg.two-operand to not two-operandize %mul-imm, since this isn't needed anymore.
This fixes the sporadic benchmark.tuple-arrays crash on 64-bit machines.
2009-06-08 21:15:52 -05:00
0d265fe016
Remove %dispatch-label since its tehe same on all platforms; fix %gc on PowerPC
2009-06-07 21:46:28 -05:00
f0b132fa7f
Fix 32-bit bootstrap
2009-06-03 03:23:55 -05:00
fd710385e5
cpu.x86: fix small register intrinsics on x86-64
2009-06-03 03:22:46 -05:00
7aca076408
GC checks now save and restore registers
2009-06-02 18:23:47 -05:00
3de85158de
Merge branch 'master' into global_optimization
2009-06-01 03:12:32 -05:00
096803e58f
Redo compiler.codegen.fixup and get %dispatch to work
2009-06-01 02:32:36 -05:00
64114947d2
Various improvements aimed at getting local optimization regressions fixed:
...
- Rename _gc to ##gc
- Absolute labels are now supported
- Generate _dispatch-label
2009-05-31 23:28:08 -05:00
e2b8b04d15
cpu.x86.features: add RDTSC support. This is a new vocabulary with words: sse2? instruction-counter count-instructions
2009-05-31 15:02:14 -05:00
40949800bf
Fixing various bugs; alias analysis wasn't handling ##phi nodes, stack analysis incorrectly handled height-changing back edges and ##fixnum-*, clean up ##dispatch generation
2009-05-29 01:39:14 -05:00
9c85bc8ce3
fix duplicate using lines
2009-05-17 20:29:32 -05:00
514956537f
Fix cpu.ppc for strict vocabulary search path semantics
2009-05-16 08:58:10 -05:00
ba04d5af1e
Update documentation for stricter vocabulary search path semantics
2009-05-16 00:29:21 -05:00
3e7269731b
cpu.ppc: really fix bool type
2009-05-10 19:10:20 -05:00
9b491d1442
Fix bool type on PowerPC
2009-05-10 19:01:38 -05:00
931db821d1
Merge branch 'master' of git://factorcode.org/git/factor
2009-05-07 23:26:37 -05:00
e007cb56e8
cpu.ppc: fix alien-indirect
2009-05-07 23:26:33 -05:00
5e35f19312
cpu.ppc: bools are 4 bytes on OS X/PowerPC
2009-05-07 23:18:41 -05:00
02bd871575
Merge branch 'master' of git://factorcode.org/git/factor
2009-05-07 19:47:42 -05:00
cb9d50887c
Update PowerPC %jump and %dispatch-label, and add PIC-related functions to cpu-ppc.hpp
2009-05-07 19:47:38 -05:00
e78c043acb
Merge branch 'master' of git://factorcode.org/git/factor
2009-05-07 19:41:54 -05:00
b45284421d
cpu.ppc.bootstrap: updates
2009-05-07 19:40:25 -05:00
db6ae46c47
Fix x86-64 backend
2009-05-07 16:58:18 -05:00
9b419aa0b1
Count megamorphic cache hits
2009-05-07 14:26:08 -05:00
74094142fe
Fix tail call PICs on x86-64
2009-05-06 22:44:30 -05:00
4f0a1b024e
Clean up bootstrap.image, and implement new calling convention for tail calls; tail call sites now have PICs
2009-05-06 22:04:01 -05:00
c1e25f3b43
JIT now supports multiple relocations per code template. This simplifies non-optimizing compiler backends
2009-05-06 20:04:49 -05:00
d3b85c14c9
Working on inline caching for tail call sites
2009-05-06 19:22:22 -05:00
478d29a175
Better separation of concerns: cpu.{x86,ppc}.assembler no longer depends on compiler.codegen.fixup and cpu.architecture. Rename rt-xt-direct to rt-xt-pic to better explain its purpose
2009-05-06 16:14:53 -05:00
fdb87205d4
Merge branch 'master' of git://factorcode.org/git/factor
2009-05-06 15:34:48 -05:00
5e90662337
Working on PowerPC backend
2009-05-06 15:39:03 -05:00
ddf05d82ce
cpu.ppc.bootstrap: working on polymorphic inline caching for PowerPC
2009-05-06 15:30:52 -05:00
bb246d5afc
Remove some unused constants and update an obsolete comment
2009-05-06 14:57:17 -05:00
18454e4e6e
cpu.x86.bootstrap: remove obsolete comment
2009-05-05 20:56:44 -05:00
9e34307f58
cpu.ppc.assembler: update for code_format=1
2009-05-05 20:56:36 -05:00
6f03366e2b
Rename enable/disable-compiler to enable/disable-optimizer, and add with-optimizer and without-optimizer combinators
2009-05-01 05:52:05 -05:00
742d574162
Non-optimizing compiler doesn't need to optimize 'dispatch' primitive anymore since its rarely used
2009-04-30 19:42:08 -05:00
f9ec0a07c5
Fix PICs for x86-64
2009-04-30 18:00:49 -05:00
5c9dc6d753
Tag-only PIC checks would always miss
2009-04-30 06:41:37 -05:00
4ae6269185
PPC make-image fix
2009-04-30 05:59:59 -05:00
5fb5c19d61
Machine code is now stored in a byte array instead of an array for add_code_block()
2009-04-30 04:34:35 -05:00
59d6131c7c
Bum 3 instructions out of megamorphic fast path by switching to fixed-size caches
2009-04-30 04:14:14 -05:00
8c01b79d06
Fix PPC make-image
2009-04-30 04:00:44 -05:00
515c619202
Non-optimizing compiler now open-codes megamorphic dispatch fast path
2009-04-30 03:37:07 -05:00
62db617927
Get rid of two instructions from tag pics and one instruction from all others
2009-04-29 22:24:28 -05:00
5188f4e1f0
Fix TEST opcode in cpu.x86.assembler
2009-04-29 22:23:42 -05:00
bd186b6320
Eliminate 3 instructions from PIC fast path
2009-04-29 21:32:05 -05:00
38a38a57a9
Fix hi-tag-or-tuple PICs
2009-04-29 03:00:09 -05:00
9243316489
Debugging inline caching
2009-04-28 23:05:18 -05:00
c15a4c1c5a
Add new relocation type for call sites which may be replaced by ICs
2009-04-28 17:53:14 -05:00
4d10105802
Working on inline caching
2009-04-28 03:48:37 -05:00
7e84daf0f1
Move (execute) to kernel.private
2009-04-26 02:42:37 -05:00
76281235e7
bootstrap.image: clean up jit-define and define-sub-primitive so that
2009-04-25 22:35:19 -05:00
27c84e89fe
Working on faster (execute)
2009-04-25 21:14:59 -05:00
3353a777f7
Fixing some unit test failures
2009-04-22 07:05:00 -05:00
0f0218b209
Fix some compile warnings in bootstrap
2009-04-17 17:53:26 -05:00
42d164db77
Fix C99 complex number support in FFI on Mac OS X/PPC
2009-03-30 07:19:14 -05:00
75452c842f
cpu.ppc.bootstrap: update for syntax change
2009-03-22 21:37:42 -05:00
ea60f8ae93
Changing : foo ; parsing to SYNTAX: foo ;
2009-03-21 01:27:50 -05:00
7f4c967ace
More compact relocation layout
2009-03-19 20:02:43 -05:00
bf41b187b0
Re-organize code so that with-compilation-unit can infer
...
Fix with-compilation-unit to work in deployed code
2009-03-13 19:39:32 -05:00
990513db60
use CONSTANT: in basis
2009-02-22 19:13:08 -06:00
6b25e99470
Add summary for heaps more vocabs
2009-02-16 21:05:13 -06:00
6b99b04531
Add vocab: for vocab-relative paths
2009-02-15 19:53:21 -06:00
3c1ddde1f9
Fix complex float support on NetBSD
2009-02-12 08:10:21 -06:00
facd455e92
Working on complex float support in FFI on NetBSD
2009-02-12 06:25:07 -06:00
44bfff7c7b
Rename ##load-indirect to ##load-reference since this is more descriptive; value numbering doesn't assign expressions to ##load-reference nodes since this would end up folding literals which were eq? but not =
2009-01-29 01:44:58 -06:00
f34c14a0f5
Remove some usages of tuck
2009-01-23 18:20:47 -06:00
34792a9f23
Remove >r/r>
2008-12-17 19:17:37 -06:00
c5160d76e9
Merge qualified, alias, symbols, constants into core
2008-12-17 18:10:01 -06:00
97fd662d61
Clean up PPC backend a bit
2008-12-16 02:03:21 -06:00
0021cde743
Fix more compile errors
2008-12-15 22:21:56 -06:00
5a45119252
Fix jit-dispatch on x86-64; was broken on windows
2008-12-15 20:55:24 -06:00
09c60f7e90
Fix some vocabularies for new requirement that inline words have a stack effect declaration
2008-12-15 20:31:55 -06:00
d821cd7281
Fix win64 bootstrap
2008-12-10 01:19:22 -06:00
4b927f732f
Clean up JIT backend a little
2008-12-09 17:14:17 -06:00
05e4626c49
Clean up
2008-12-09 03:22:09 -06:00
0359ec8eac
Fix PowerPC backend again
2008-12-06 16:57:28 -06:00
8a8f0c925c
Use BSR instruction to implement fixnum-log2 intrinsic
2008-12-06 15:31:17 -06:00
d2ce4355f8
Fixing PPC backend
2008-12-06 15:30:40 -06:00
a56d480aa6
Various optimizations leading to a 10% speedup on compiling empty EBNF parser:
...
- open-code getenv primitive
- inline tuple predicates in finalization
- faster partial dispatch
- faster built-in type predicates
- faster tuple predicates
- faster lo-tag dispatch
- compile V{ } clone and H{ } clone more efficiently
- add fixnum fast-path to =; avoid indirect branch if two fixnums not eq
- faster >alist on hashtables
2008-12-06 09:16:29 -06:00
b06cfc6225
Update ppc backend for recent string intrinsic changes
2008-12-06 00:12:07 -06:00
82cf6530c6
set-string-nth-fast intrinsic was busted
2008-12-05 23:52:09 -06:00
e256846acd
Tweak string representation; high bit indicates if character has high bits in aux vector. Avoids memory access in common case. Split set-string-nth into two primitives; set-string-nth-fast is open-coded by optimizing compiler. 13% improvement on reverse-complement
2008-12-05 06:38:51 -06:00
e5ed7447ed
Removing more >r/r> usages
2008-12-03 08:46:16 -06:00
e4db2afb7e
Update cpu.x86.assembler to use dip instead of >r/r>
2008-12-02 03:10:01 -06:00
a558f91d5d
Better command-line processing
2008-12-01 16:10:48 -06:00
Slava Pestov
Joe Groff
Slava Pestov
Slava Pestov
U-C4\Administrator
Slava Pestov
sheeple
Doug Coleman
sheeple
Slava Pestov