6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-11-11 08:07:22 +00:00
Code Issues Projects Releases Wiki Activity
86,558 Commits 39 Branches 0 Tags 277 MiB
01c6de341c
Commit Graph

24 Commits

Author SHA1 Message Date
Elena Demikhovsky
3251020738 This patch optimizes shuffle instruction - generates 2 instructions instead of 4. 
Since this specific shuffle is widely used in many workloads we have ~10% performance on them.

shufflevector <8 x float> %A, <8 x float> %B, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14>

vmovaps (%rdx), %ymm0
vshufps $8, %ymm0, %ymm0, %ymm0
vmovaps (%rcx), %ymm1
vshufps $8, %ymm0, %ymm1, %ymm1
vunpcklps       %ymm0, %ymm1, %ymm0

vmovaps (%rcx), %ymm0
vmovsldup       (%rdx), %ymm1
vblendps        $85, %ymm0, %ymm1, %ymm0


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@163134 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-09-04 12:49:02 +00:00
Elena Demikhovsky
1596373671 Shuffle optimization for AVX/AVX2. 
The current patch optimizes frequently used shuffle patterns and gives these instruction sequence reduction.
Before:
      vshufps $-35, %xmm1, %xmm0, %xmm2 ## xmm2 = xmm0[1,3],xmm1[1,3]
       vpermilps       $-40, %xmm2, %xmm2 ## xmm2 = xmm2[0,2,1,3]
       vextractf128    $1, %ymm1, %xmm1
       vextractf128    $1, %ymm0, %xmm0
       vshufps $-35, %xmm1, %xmm0, %xmm0 ## xmm0 = xmm0[1,3],xmm1[1,3]
       vpermilps       $-40, %xmm0, %xmm0 ## xmm0 = xmm0[0,2,1,3]
       vinsertf128     $1, %xmm0, %ymm2, %ymm0
After:
      vshufps $13, %ymm0, %ymm1, %ymm1 ## ymm1 = ymm1[1,3],ymm0[0,0],ymm1[5,7],ymm0[4,4]
      vshufps $13, %ymm0, %ymm0, %ymm0 ## ymm0 = ymm0[1,3,0,0,5,7,4,4]
      vunpcklps       %ymm1, %ymm0, %ymm0 ## ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@159188 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-06-26 08:04:10 +00:00
Jakob Stoklund Olesen
5984d2b31f Run ProcessImplicitDefs on SSA form where it can be much simpler. 
Implicitly defined virtual registers can simply have the <undef> bit set
on all uses, and copies can be turned into implicit defs recursively.

Physical registers are a bit trickier. We handle the common case where a
physreg def is used by a nearby instruction in the same basic block. For
more complicated cases, just leave the IMPLICIT_DEF instruction in.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@159149 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-06-25 18:12:18 +00:00
Craig Topper
703c38bf58 Don't insert 128-bit UNDEF into 256-bit vectors. Just keep the 256-bit vector. Original patch by Elena Demikhovsky. Tweaked by me to allow possibility of covering more cases. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@158792 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-06-20 05:39:26 +00:00
Craig Topper
8ae97baef2 Allow 256-bit shuffles to still be split even if only half of the shuffle comes from two 128-bit pieces. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@157175 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-05-21 06:40:16 +00:00
Craig Topper
f3640d7ec1 Allow v16i16 and v32i8 shuffles to be rewritten as narrower shuffles. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@156156 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-05-04 04:44:49 +00:00
Nadav Rotem
e611378a6e Reapply 154396 after fixing a test. 
Original message:
Modify the code that lowers shuffles to blends from using blendvXX to vblendXX.
blendV uses a register for the selection while Vblend uses an immediate.
On sandybridge they still have the same latency and execute on the same execution ports.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154483 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-04-11 06:40:27 +00:00
Eric Christopher
a139051654 Temporarily revert this patch to see if it brings the buildbots back. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154425 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-04-10 19:33:16 +00:00
Nadav Rotem
50e64cfe6e Modify the code that lowers shuffles to blends from using blendvXX to vblendXX. 
blendv uses a register for the selection while vblend uses an immediate.
On sandybridge they still have the same latency and execute on the same execution ports.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154396 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-04-10 14:33:13 +00:00
Nadav Rotem
e80aa7c783 Lower some x86 shuffle sequences to the vblend family of instructions. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154313 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-04-09 08:33:21 +00:00
Craig Topper
f85cb768fe Test case for PR12413 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154172 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-04-06 14:38:25 +00:00
Chad Rosier
abd6674166 Fix a regression from r147481. 
Original commit message from r147481:
DAGCombine for transforming 128->256 casts into a vmovaps, rather
then a vxorps + vinsertf128 pair if the original vector came from a load.

Fix:
Unaligned loads need to generate a vmovups.
rdar://10974078



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152366 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-03-09 02:00:48 +00:00
Craig Topper
dbd98a4b1b Add instruction selection for 256-bit VPSHUFD and 128-bit VPERMILPS/VPERMILPD. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@149968 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-02-07 06:28:42 +00:00
Nadav Rotem
7e413e9c94 The type-legalizer often scalarizes code. One of the common patterns is extract-and-truncate. 
In this patch we optimize this pattern and convert the sequence into extract op of a narrow type.
This allows the BUILD_VECTOR dag optimizations to construct efficient shuffle operations in many cases.




git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@149692 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-02-03 13:18:25 +00:00
Craig Topper
cc30006391 Fix pattern for memory form of PSHUFD for use with FP vectors to remove bitcast to an integer vector that normal code wouldn't have. Also remove bitcasts from code that turns splat vector loads into a shuffle as it was making the broken pattern necessary. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@149232 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-01-30 07:50:31 +00:00
Nadav Rotem
ba05c91ed2 Transform: (EXTRACT_VECTOR_ELT( VECTOR_SHUFFLE )) -> EXTRACT_VECTOR_ELT. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@148337 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-01-17 21:44:01 +00:00
Nadav Rotem
0b94b5f52b Fix 11769. 
In CanXFormVExtractWithShuffleIntoLoad we assumed that EXTRACT_VECTOR_ELT can be later handled by the DAGCombiner.
However, in some cases on AVX, the EXTRACT_VECTOR_ELT is legalized to EXTRACT_SUBVECTOR + EXTRACT_VECTOR_ELT, which
currently is not handled by the DAGCombiner. In this patch I added a check that we only extract from the XMM part.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@148298 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-01-17 09:13:19 +00:00
Chad Rosier
4f00c08062 Cleanup test case by adding checks for test names. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@148166 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-01-14 01:46:51 +00:00
Elena Demikhovsky
16db710898 Fixed a bug in LowerVECTOR_SHUFFLE caused assertion failure 
lc: X86ISelLowering.cpp:6480: llvm::SDValue llvm::X86TargetLowering::LowerVECTOR_SHUFFLE(llvm::SDValue, llvm::SelectionDAG&) const: Assertion `V1.getOpcode() != ISD::UNDEF&&  "Op 1 of shuffle should not be undef"' failed.
Added a test.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@148044 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-01-12 20:33:10 +00:00
Chad Rosier
3d1161e9ae Enhance DAGCombine for transforming 128->256 casts into a vmovaps, rather 
then a vxorps + vinsertf128 pair if the original vector came from a load.
rdar://10594409

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@147481 91177308-0d34-0410-b5e6-96231b3b80d8
 2012-01-03 21:05:52 +00:00
Elena Demikhovsky
021c0a2ee7 Fixed a bug in LowerVECTOR_SHUFFLE and LowerBUILD_VECTOR. 
Matching MOVLP mask for AVX (265-bit vectors) was wrong.
The failure was detected by conformance tests.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@147308 91177308-0d34-0410-b5e6-96231b3b80d8
 2011-12-28 08:14:01 +00:00
Evan Cheng
b3e6c70c84 Update test to something more sensible. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@146282 91177308-0d34-0410-b5e6-96231b3b80d8
 2011-12-09 21:54:10 +00:00
Evan Cheng
e955726a0e Add 256-bit variant vmovss and vmovsd patterns. rdar://10538417 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@146196 91177308-0d34-0410-b5e6-96231b3b80d8
 2011-12-08 22:30:45 +00:00
Eli Friedman
dca62d53b7 Make sure the X86 backend doesn't explode on 128-bit shuffles in AVX mode. Fixes PR11102. 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@141585 91177308-0d34-0410-b5e6-96231b3b80d8
 2011-10-10 22:28:47 +00:00