Commits · 200e87c0c59d88b1dddd12295a87f60fa2edfa41 · Roger Ferrer / llvm-epi · GitLab

Oct 04, 2014
- [x86] Fix a bug in the VZEXT DAG combine that I just made more powerful. · 200e87c0
  Chandler Carruth authored Oct 04, 2014
  200e87c0
- [x86] Sink a generic combine of VZEXT nodes from the lowering to VZEXT · 7e26a67f
  Chandler Carruth authored Oct 04, 2014
  7e26a67f
- [x86] Add a really preposterous number of patterns for matching all of · f3e88069
  Chandler Carruth authored Oct 03, 2014
  f3e88069
Oct 03, 2014
- [x86] Adjust the patterns for lowering X86vzmovl nodes which don't · 0adda1e4
  Chandler Carruth authored Oct 03, 2014
  0adda1e4
- [ISel] Keep matching state consistent when folding during X86 address match · ff63a2dc
  Adam Nemet authored Oct 03, 2014
  ff63a2dc
- [x86] Teach the new vector shuffle lowering to aggressively form MOVSS · 19640789
  Chandler Carruth authored Oct 03, 2014
  19640789
- [x86] Refactor the element insertion logic in the new vector shuffle · 4bf341de
  Chandler Carruth authored Oct 03, 2014
  4bf341de
- [x86] Significantly improve the ability of the new vector shuffle · 971a560c
  Chandler Carruth authored Oct 03, 2014
  971a560c
- [x86] Unbreak SSE1 with the new vector shuffle lowering. We can't widen · e91b3162
  Chandler Carruth authored Oct 03, 2014
  e91b3162
- [AVX512] Pull pattern for subvector insert into the instruction definition · 4dca3ce4
  Adam Nemet authored Oct 02, 2014
  4dca3ce4
- [AVX512] Refactor subvector inserts · 4e2ef472
  Adam Nemet authored Oct 02, 2014
  4e2ef472
- [AVX512] Fix i256mem->f256mem typo in VINSERTF64x4rm · dc87aea1
  Adam Nemet authored Oct 02, 2014
  dc87aea1
- [Stackmaps] Make ithe frame-pointer required for stackmaps. · 99bd3cba
  Juergen Ributzka authored Oct 02, 2014
  99bd3cba
Oct 02, 2014
- [x86] Teach the new vector shuffle lowering to widen floating point · 75e182b4
  Chandler Carruth authored Oct 02, 2014
  75e182b4
- [x86] Improve and correct how the new vector shuffle lowering was · 8a16802d
  Chandler Carruth authored Oct 01, 2014
  8a16802d
Oct 01, 2014
- Lower FNEG ( FABS (x) ) -> FNABS (x) [X86 codegen] PR20578 · 9ebfbb96
  Sanjay Patel authored Oct 01, 2014
  9ebfbb96
- constify TargetMachine parameter for X86TargetLowering. · 12f4a785
  Eric Christopher authored Oct 01, 2014
  12f4a785
- Don't repeat function/variable name in comment. NFC. · 0e4a83e8
  Sanjay Patel authored Oct 01, 2014
  0e4a83e8
- Move the complex address expression out of DIVariable and into an extra · 87b7eb9d
  Adrian Prantl authored Oct 01, 2014
  87b7eb9d
- Revert r218778 while investigating buldbot breakage. · b458dc2e
  Adrian Prantl authored Oct 01, 2014
  b458dc2e
- Move the complex address expression out of DIVariable and into an extra · 25a7174e
  Adrian Prantl authored Oct 01, 2014
  25a7174e
- [x86] Fix a few more tiny patterns with the new vector shuffle lowering · 6c02c031
  Chandler Carruth authored Oct 01, 2014
  6c02c031
- [x86] Delete some extraneous logic from the new vector shuffle lowering. · 04848610
  Chandler Carruth authored Oct 01, 2014
  04848610
- Fix typo in comment from r218733 · 5f75f4dd
  Nick Lewycky authored Oct 01, 2014
  5f75f4dd
- [x86] Teach the new vector shuffle lowering to be even more aggressive · 26cb9b8d
  Chandler Carruth authored Oct 01, 2014
  26cb9b8d
- [x86] Hoist the zext-lowering up in the v4i32 lowering routine -- it is · 846baf2c
  Chandler Carruth authored Oct 01, 2014
  846baf2c
- [AVX512] Remove space before \t in AsmStrings. · 05d8c8e6
  Adam Nemet authored Oct 01, 2014
  05d8c8e6
- [x86] Teach the new vector shuffle lowering about VBROADCAST and · b9d3fa1e
  Chandler Carruth authored Oct 01, 2014
  b9d3fa1e
Sep 30, 2014
- [AVX512] Added intrinsics for 128-, 256- and 512-bit versions of VCMPGT{BWDQ}. · 28a7df0b
  Robert Khasanov authored Sep 30, 2014
  28a7df0b
- [AVX512] Added intrinsics for 128- and 256-bit versions of VCMPEQ{BWDQ} · 5aa4445b
  Robert Khasanov authored Sep 30, 2014
  5aa4445b
- [AVX512] Added intrinsics for VPCMPEQB and VPCMPEQW. · b25e562d
  Robert Khasanov authored Sep 30, 2014
  b25e562d
- [AVX512] Enabled intrinsics for VPCMPEQD and VPCMPEQQ. · a27c8e0f
  Robert Khasanov authored Sep 30, 2014
  a27c8e0f
- [x86] Revert r218588, r218589, and r218600. These patches were pursuing · aaf8e03d
  Chandler Carruth authored Sep 30, 2014
  aaf8e03d
- [AVX512] Use X86VectorVTInfo in the masking helper classes and the FMAs · 6bddb8c3
  Adam Nemet authored Sep 29, 2014
  6bddb8c3
Sep 29, 2014
- [x86] Make the new vector shuffle lowering lower blends as VSELECT · 6cbf4316
  Chandler Carruth authored Sep 29, 2014
  6cbf4316
- [x86] Delete a bunch of really bad and totally unnecessary code in the · b1cc7a85
  Chandler Carruth authored Sep 29, 2014
  b1cc7a85
- [x86] Refactor all of the VSELECT-as-blend lowering code to avoid domain · d639c7a8
  Chandler Carruth authored Sep 29, 2014
  d639c7a8
- [x86] Improve naming and comments for VSELECT lowering. · 2f9e56e5
  Chandler Carruth authored Sep 29, 2014
  2f9e56e5
- [x86] Add the dispatch skeleton to the new vector shuffle lowering for · c7129276
  Chandler Carruth authored Sep 29, 2014
  c7129276
- [x86] Make the split-and-lower routine fully generic by relaxing the · 32a3ebda
  Chandler Carruth authored Sep 29, 2014
  32a3ebda