Commits · e35244cf636b684e0725738cab900ee8f024d044 · Roger Ferrer / llvm-epi · GitLab

Nov 08, 2015
- [WinEH] Update PHIs of CATCHRET successors · e35244cf
  David Majnemer authored Nov 08, 2015
  e35244cf
Nov 07, 2015
- [X86] Fold (trunc (i32 (zextload i16))) into vbroadcast. · b49eb3ab
  Ahmed Bougacha authored Nov 06, 2015
  b49eb3ab
- [X86] SRL non-LSB extracts when folding to truncating broadcasts. · 05a0514b
  Ahmed Bougacha authored Nov 06, 2015
  05a0514b
- [X86] Don't fold non-LSB extracts into truncating broadcasts. · 68614a36
  Ahmed Bougacha authored Nov 06, 2015
  68614a36
Nov 06, 2015
- [ShrinkWrapping] Teach shrink-wrapping how to analyze RegMask. · 9a8efc08
  Quentin Colombet authored Nov 06, 2015
  9a8efc08
- Improved the operands commute transformation for X86-FMA3 instructions. · 4731bea3
  Andrew Kaylor authored Nov 06, 2015
  4731bea3
- [WinEH] Mark funclet entries and exits as clobbering all registers · b8fd162f
  Reid Kleckner authored Nov 06, 2015
  b8fd162f
- Bring r252305 back with a test fix. · 889d7bb4
  Rafael Espindola authored Nov 06, 2015
  889d7bb4
Nov 05, 2015
- DI: Reverse direction of subprogram -> function edge. · d4bff303
  Peter Collingbourne authored Nov 05, 2015
  d4bff303
- [WinEH] Fix funclet prologues with stack realignment · 6ddae310
  Reid Kleckner authored Nov 05, 2015
  6ddae310
- Add cfi instr for CFA calculation when movpc is expanded to call and pop · 99fba3c1
  Petar Jovanovic authored Nov 05, 2015
  99fba3c1
- revert rev. 252153 due to build failure on ubuntu · f99c054e
  Asaf Badouh authored Nov 05, 2015
  f99c054e
- [X86][AVX512] add comi with Sae · 7fdabf0a
  Asaf Badouh authored Nov 05, 2015
  7fdabf0a
Nov 04, 2015
- [x86] Teach the shrink-wrapping hooks to do the proper thing with Win64. · 421723cd
  Quentin Colombet authored Nov 04, 2015
  421723cd
- [X86][SSE] Add general memory folding for (V)INSERTPS instruction · 7e6606f4
  Simon Pilgrim authored Nov 04, 2015
  7e6606f4
- Created new X86 FMA3 opcodes (FMA*_Int) that are used now for lowering of scalar FMA intrinsics. · e41a8c41
  Andrew Kaylor authored Nov 04, 2015
  e41a8c41
- [X86] DAGCombine should not introduce FILD in soft-float mode · b34de722
  Michael Kuperstein authored Nov 04, 2015
  b34de722
- [StatepointLowering] Remove distinction between call and invoke safepoints · 35fe6920
  Igor Laevsky authored Nov 04, 2015
  35fe6920
Nov 03, 2015
- [X86][AVX] Tweaked shuffle stack folding tests · b0d860a3
  Simon Pilgrim authored Nov 03, 2015
  b0d860a3
- [X86][AVX512] Fixed shuffle test name to match shuffle · df993479
  Simon Pilgrim authored Nov 03, 2015
  df993479
- [X86][XOP] Add support for the matching of the VPCMOV bit select instruction · e88dc04c
  Simon Pilgrim authored Nov 03, 2015
  e88dc04c
- Remove unnecessary dependency on section and string positions. · 41302723
  Rafael Espindola authored Nov 03, 2015
  41302723
- [X86] Generate .cfi_adjust_cfa_offset correctly when pushing arguments · 73dc8529
  Michael Kuperstein authored Nov 03, 2015
  73dc8529
Nov 02, 2015
- In MachineBlockPlacement, filter cold blocks off the loop chain when profile data is available. · b90b9e05
  Cong Hou authored Nov 02, 2015
  b90b9e05
- Fix two issues in MergeConsecutiveStores: · 646c4032
  James Y Knight authored Nov 02, 2015
  646c4032
- AVX512: Implemented encoding and intrinsics for VBROADCASTI32x2 and VBROADCASTF32x2 instructions. · fa798a9d
  Igor Breger authored Nov 02, 2015
  fa798a9d
- [X86] Don't pass a scale value of 0 to scatter/gather intrinsics. This causes... · cc56e3a1
  Craig Topper authored Nov 02, 2015
  cc56e3a1
Nov 01, 2015
- AVX-512: Optimized SIMD truncate operations for AVX512F set. · db738d9c
  Elena Demikhovsky authored Nov 01, 2015
  db738d9c
Oct 29, 2015
- [X86][SSE] Added load+sext tests for 16i1->16i8 and 32i1->32i8 · 6baace03
  Simon Pilgrim authored Oct 29, 2015
  6baace03
- [X86][SSE] Shuffle blends with zero · ca56a72a
  Simon Pilgrim authored Oct 29, 2015
  ca56a72a
Oct 27, 2015
- [X86][AVX512] Test UNPCK with non-sequential scalars · 94c49435
  Simon Pilgrim authored Oct 27, 2015
  94c49435
- Use the 'arcp' fast-math-flag when combining repeated FP divisors · bbd4c79c
  Sanjay Patel authored Oct 27, 2015
  bbd4c79c
- [X86][AVX512] [X86][AVX512] add convert float to half · c7cb8806
  Asaf Badouh authored Oct 27, 2015
  c7cb8806
- [X86] Make elfiamcu an OS, not an environment. · e1194bdb
  Michael Kuperstein authored Oct 27, 2015
  e1194bdb
- [x86] replace integer logic ops with packed SSE FP logic ops · 309c4f93
  Sanjay Patel authored Oct 27, 2015
  309c4f93
- Fix llc crash processing S/UREM for -Oz builds caused by rL250825. · fee370be
  Steve King authored Oct 27, 2015
  fee370be
- add FP logic test cases to show current codegen (PR22428) · 28d1598e
  Sanjay Patel authored Oct 26, 2015
  28d1598e
Oct 26, 2015
- [x86] Make the vselect-minmax test 2x to 3x faster by deleting all the · 5a14186b
  Chandler Carruth authored Oct 26, 2015
  5a14186b
- fix test errors (on windows) for commit r251287 · 2e919c89
  Igor Breger authored Oct 26, 2015
  2e919c89
- AVX512: Enabled VPBROADCASTB lowering for v64i8 vectors. · e4ddc3f4
  Igor Breger authored Oct 26, 2015
  e4ddc3f4