becd957d29
compiler.cfg: merge all alien accessors into ##load-memory-imm and ##store-memory-imm
2010-05-03 17:34:06 -04:00
3c965afd10
compiler.cfg: remove ##set-string-nth-fast instruction since it can be expressed just as efficiently using other instructions
2010-05-03 17:34:04 -04:00
6fdcd9fb02
Untagged fixnums work in progress
2010-05-03 17:34:02 -04:00
1d7089dc04
compiler: combine ##load-constant followed by ##alien-double into a ##load-double on x86-32, saving an integer register
2010-04-18 21:42:45 -05:00
dbd7489894
compiler.cfg: fix some bugs introduced by the ##compare-imm fusion patch
2010-04-18 21:42:45 -05:00
c4cf2a4085
compiler: change how 'f' is represented in low level IR to simplify some code, and fuse a ##load-constant of a word with a ##compare into a ##compare-imm on x86-32. This eliminates a spill from binary-search
2010-04-18 21:42:45 -05:00
3b4f8758b5
Merge branch 'master' of git://factorcode.org/git/factor into abi-symbols
...
Conflicts:
basis/compiler/tests/alien.factor
2010-04-06 12:30:15 -07:00
c3b7a04b50
update %load-param-reg, %save-param-reg, and %cleanup to take register args into account
2010-04-01 17:36:09 -07:00
4f585f991d
compiler: new set-special-object intrinsic; more efficient special-object intrinsic
2010-04-01 20:09:14 -04:00
adeb6fb68f
vm: actually use context callstacks when running code
2010-03-26 22:44:43 -04:00
af47e5e8d8
Compiled code size optimizations for x86-64 resulting in development image size reduction of about 350kb
...
- RIP-relative LEA is now used to load program counter in prolog and tail calls; saves 3 bytes of code and 4 bytes of relocation information over an immediate MOV
- VM field accesses no longer involve a redundant addition
2010-02-03 23:11:35 +13:00
59f0e7a197
alien.c-types: ensure that long and ulong types are defined with define-primitive-type and not typedef
2010-01-23 00:15:08 +13:00
80e64b8379
Store VM object in a register on x86-64
2010-01-11 01:20:32 +13:00
98b296a021
cpu.x86: sse_version was renamed to (sse-version)
2010-01-07 18:30:01 +13:00
d2baa6dde5
vm: move c_to_factor, lazy_jit_compile_impl, throw_impl, set_callstack assembly routines into non-optimizing compiler for x86-64
2010-01-06 15:47:36 +13:00
9399a68786
Change how non-volatile register preservation is done in alien callbacks, with the aim of fixing callbacks on PowerPC, and to eventually eliminate assembly code from VM
...
- Simplify calculation of offset in relocation table
- Open-code %alien-callback
- Remove magic_frame hack from context objects
- Move magical return instruction from optimizing compiler backend into callback entry stub
2010-01-03 01:11:51 +13:00
8f449f1a82
Get optimizing compiler working without global register variables in VM
2009-12-22 15:42:49 +13:00
1ad9dc14b4
fix simd tests
2009-12-05 17:17:16 -08:00
1fbbe91e73
implement %scalar>integer for longlong reps on x86.64
2009-12-05 15:38:43 -08:00
d674ff8191
simd intrinsic implementation for v*high, v*hs+, vavg, and vsad
2009-12-05 14:52:18 -08:00
4116e2aa4e
don't use intrinsics for simd vector element operations when the component size doesn't fit in a fixnum
2009-12-03 12:46:56 -08:00
ebcaaa0d64
Merge branch 'master' into simd-cleanup
2009-11-26 16:14:46 -08:00
8095704010
change name of 'unsign-rep' to more sensible 'signed-rep'
2009-11-24 22:44:12 -08:00
48d6a86455
fill in longlong cases for %scalar>integer (won't work on 32bit yet!)
2009-11-19 12:15:24 -08:00
3323072613
Merge branch 'master' into startup
...
Conflicts:
core/bootstrap/primitives.factor
vm/run.hpp
2009-11-15 02:52:50 -06:00
993631e44c
Merge branch 'master' of git://factorcode.org/git/factor into simd-cleanup
...
Conflicts:
basis/math/vectors/simd/functor/functor.factor
2009-11-14 21:02:39 -06:00
4168ffbcee
New identity-hashcode primitive
2009-11-10 21:18:54 -06:00
a279b360eb
Merge branch 'master' into simd-cleanup
...
Conflicts:
basis/math/vectors/simd/functor/functor.factor
2009-11-05 11:27:08 -06:00
cf28782533
more intrinsic madness
2009-11-05 09:52:57 -06:00
54e9baa7e4
change ##horizontal-add-vector insn to better match what the HADD SSE instructions do (add adjacent pairs, pack results)
2009-11-04 12:18:01 -06:00
39c192c829
Align stack pointer on non-Mac OS X x86-32 platforms, and use aligned loads/stores for SIMD values
2009-11-03 23:51:44 -06:00
81b17590bd
make horizontal shift available to float vectors (it'd still be faster than the software fallback despite pipeline penalty)
2009-11-03 21:38:29 -06:00
f102bd7ded
cpu.x86: update %box-displaced-alien for introduction of address field
2009-11-03 03:17:43 -06:00
6299c42564
Minor bug fixes for 4-bit tags
2009-11-02 17:41:36 -06:00
9ac1cec6c4
move all simd intrinsics to compiler.cfg.intrinsics.simd, and only load it when math.vectors.simd is loaded
2009-11-02 15:17:34 -06:00
d65296b334
vm: 4 bit tags, new representation of alien objects makes unbox-any-c-ptr more efficient (work in progress)
2009-11-02 04:25:54 -06:00
55eb76f695
Merge branch 'master' into new_gc
2009-11-01 22:17:27 -06:00
6e9bd9c631
typo in cpu.x86
2009-11-01 19:39:57 -06:00
0c7f036437
clear destination register before doing CVTS* to break dependency chains
2009-11-01 18:29:12 -06:00
a0396f919a
SSE code generation improvements: always use MOVAPS to move float xmm registers to kill dependency chains, and favor -PS versions of logical, move, and shuffle ops to shrink code size
2009-11-01 16:09:44 -06:00
051dcb6849
Merge branch 'master' into new_gc
2009-10-30 03:03:05 -05:00
ca8d4c15f4
add ##shl-vector-imm and ##shr-vector-imm insn variants. use merge/shr instead of compare/merge to do signed unpacks
2009-10-30 00:41:19 -05:00
b5fd809209
memq? -> member-eq?, sorted-memq? -> sorted-member-eq?
2009-10-28 15:02:00 -05:00
b78202e816
Merge branch 'master' into new_gc
2009-10-22 05:40:57 -05:00
1bff41e35e
oops, longlong comparison is sse4.2, not 4.1
2009-10-21 23:00:02 -05:00
a2976083be
generate unsigned vector comparison fallbacks using min/max or xor/signed compare
2009-10-20 22:30:57 -05:00
0f39bfd4df
Merge branch 'master' into startup
2009-10-20 15:01:06 -04:00
f442d1949b
Change data heap alignment to 16 bytes
2009-10-20 12:45:00 -05:00
a71212f9c4
cpu.x86.32: only create 16-byte parameter area if the word calls into the VM
2009-10-20 05:02:42 -05:00
6fa23d7ea4
change add-init-hook to add-startup-hook, new add-shutdown-hook word
2009-10-19 22:17:02 -04:00
0a3029d9f2
compiler: on PPC, ANDI, ORI and XORI instructions take an unsigned 16-bit immediate, unlike ADDI, SUBI and MULLI which take a signed 16-bit immediate. The code generator was not aware of this, and so for example '[ >fixnum -16 bitand ]' would generate incorrect code. Split up small-enough? hook into immediate-arithmetic? and immediate-bitwise? and update value numbering to be aware of this. Fixes classes.struct bitfields test failure
2009-10-19 04:58:29 -05:00
0726ec30f0
cpu.x86: eliminate 2 instructions form write barrier on x86-32
2009-10-15 22:07:03 -05:00
649f7531fd
cpu.x86: just a cleanup
2009-10-15 05:43:28 -05:00
22e79e8495
compiler: tweak ##write-barrier-imm
2009-10-15 02:40:23 -05:00
bfd1f0d6d2
vm: rt-vm relocation now supports accessing a field directly
2009-10-14 19:24:23 -05:00
10ad5cad53
Working on adding support for the new write barrier to optimized code
2009-10-14 02:06:01 -05:00
b82c8b4416
use TEST reg, reg to compare integer equality with zero
2009-10-10 13:13:53 -05:00
2577ab83a6
only emit ##alien-vector/##set-alien-vector insns if the rep is available
2009-10-10 12:53:10 -05:00
1e3c9321ae
don't use MOVSLDUP/MOVSHDUP to do specialized shuffles unless sse3 is available
2009-10-10 12:00:47 -05:00
a7a77cd03e
fix x86 uchar %scalar>integer
2009-10-10 10:39:23 -05:00
5158a12d32
rename ##shuffle-vector to ##shuffle-vector-imm, and add a new ##shuffle-vector for dynamic shuffles. have vshuffle use ##shuffle-vector to do word and byte shuffles on x86
2009-10-09 21:26:27 -05:00
98836a9e2e
break vector compare intrinsics into %compare, %or, and %not instructions that map directly to cpu instructions
2009-10-07 15:27:03 -05:00
43b51ef2eb
decompose %unpack-vector-head/tail into %compare-vector/%merge-vector-head/tail or %tail>head-vector/%unpack-vector-head insns when there isn't an actual unpack insn; get rid of fake x86 implementations
2009-10-07 14:09:46 -05:00
5152c3b06d
sse doesn't actually have an unsigned->unsigned pack instruction
2009-10-07 12:00:31 -05:00
a13e75f4f4
don't generate a ##not-vector instruction if the cpu doesn't have one; instead, fall back to a ##fill-vector/##xor-vector combo. get rid of pretend %not-vector in cpu.x86
2009-10-07 11:59:36 -05:00
444624e79f
fix x86 %unpack-vector insns
2009-10-06 20:38:51 -05:00
2edccca0bb
oops...PACKUSDW is sse4 only
2009-10-06 20:09:50 -05:00
425ea05529
%float>integer-vector should truncate
2009-10-06 13:57:54 -05:00
84ecb1266d
add insns for vector pack, unpack, integer>float, and float>integer
2009-10-05 22:34:14 -05:00
931107397c
compiler.cfg: remove _gc instruction, it doesn't need to exist, and change GC checks to ensure that the right amount of space is available instead of blindly checking for 1Kb
2009-10-05 05:27:49 -05:00
dca9d3e535
add %merge-vector-head and %merge-vector-tail instructions to back vmerge
2009-10-03 21:48:53 -05:00
335df20713
add intrinsics for v<=, v<, v=, v>, v>=, vunordered?
2009-10-03 11:29:34 -05:00
b1ec36a324
extend x86 %compare-vector to cover all comparison codes, sometimes stupidly for now
2009-10-02 23:19:56 -05:00
e2e75c6b3a
add intrinsic for vnot/vbitnot
2009-10-02 20:04:28 -05:00
9d424a1092
Merge branch 'master' of git://factorcode.org/git/factor
...
Conflicts:
basis/compiler/codegen/codegen.factor
2009-10-01 23:14:16 -05:00
7b13fa4283
fold test-vector/branch sequences into a test-vector-branch instruction
2009-10-01 19:53:30 -05:00
228ad950bb
%test-vector instruction for vany?, vall?, vnone?
2009-10-01 15:35:38 -05:00
94070c11aa
%compare-vector instruction (only does v= for now)
2009-10-01 14:31:37 -05:00
3ba79be651
Revert "add a %blend-vector intrinsic for v?"
...
This reverts commit 21e4b28b67
2009-09-30 23:40:37 -05:00
37a091a188
Merge branch 'master' of git://factorcode.org/git/factor
2009-09-30 23:04:04 -05:00
21e4b28b67
add a %blend-vector intrinsic for v?
2009-09-30 23:03:59 -05:00
65421b111b
math.vectors.simd: use fallbacks for hlshift, hrshift, vshuffle if parameter is not a literal;al; element access in int-4 on x86-64 now sign-extends the value; don't throw error at compile time if parameter for vshuffle does not have enough elements
2009-09-30 20:04:37 -05:00
8e201ca4b7
Various minor compiler tweaks: Combine address calculation with dereferencing in alien accessors; convert SIMD XOR of a vector with itself into an XOR of the destination with itself; convert SIMD unbox of zero vector into XOR of the destination with itself; fix SIMD indexing on x86-64
2009-09-30 05:00:36 -05:00
2b13245704
math.vectors.simd: add fast intrinsic for 'nth', replace broadcast primitive with shuffles
2009-09-29 04:48:11 -05:00
a6e8277b2c
math.vectors.simd: add vshuffle intrinsic
2009-09-28 23:12:13 -05:00
db217295b0
Work in progress
2009-09-28 17:31:34 -05:00
49dba53760
cpu.x86: cleanups
2009-09-28 16:38:35 -05:00
4e2e45b70d
use PSHUFD for longlong-2 broadcast when dst != src to avoid a %copy
2009-09-28 12:04:08 -05:00
3f90473f09
use MOVDDUP for double-2 broadcast to eliminate a %copy
2009-09-28 12:00:03 -05:00
467c389948
cpu.x86.assembler: make SSE shuffle instructions accept an array of indexes so they're easier to use
2009-09-28 11:45:45 -05:00
f7d416a4e4
SSE integer gather and broadcast
2009-09-28 11:24:08 -05:00
f08521bf83
Fixing various test failures caused by C type parser change, and clarify C type docs some more
2009-09-28 08:48:39 -05:00
1109fb5725
math.vectors.simd: add intrinsic for int-4-boa, uint-4-boa, fix tests for C type parser change, fix software fallback for horizontal shifts
2009-09-28 06:34:22 -05:00
dc1b6043dc
cpu.x86: shifts didn't work if dst != src1; re-organize file a bit
2009-09-28 05:39:53 -05:00
daf8f0ebba
cpu.x86: fix regression: fsqrt intrinsic wasn't used
2009-09-28 02:27:55 -05:00
10c5fe5933
math.vectors.simd: add hlshift, hrshift (128-bit shift), vbitandn intrinsics
2009-09-28 02:17:46 -05:00
e8cfaccef0
compiler.cfg: nuke ##bignum>integer and ##integer>bignum since they were unused
2009-09-27 20:36:05 -05:00
6dd8e4657e
Merge branch 'master' into more_aggressive_coalescing
2009-09-27 19:29:50 -05:00
6f2a4eba51
compiler.cfg.linear-scan: fix partial sync point logic in case where dst == src, and clean up spilling code
2009-09-27 19:28:20 -05:00
2efab6efad
cpu.x86.32: implement %unary-float-function and %binary-float-function; speeds up partial-sums and struct-arrays benchmarks
2009-09-27 18:06:30 -05:00
Slava Pestov
Joe Groff
Doug Coleman
Slava Pestov