Commits · a8c24d57b81703b5730460d7cb12af9783a02539 · Roger Ferrer / llvm-epi · GitLab

Nov 07, 2022
- [InstCombine] Remove redundant splats in InstCombineVectorOps · a8c24d57
  Matt Devereau authored Oct 12, 2022
  a8c24d57
- [AArch64][SVE2] Add the SVE2.1 BF16 instructions · cf69895a
  David Sherwood authored Nov 02, 2022
  cf69895a
- [X86] Folded MOVDDUPrm has the same sched behaviour as MOVSHDUPrm/MOVSLDUPrm on Haswell/IceLake · 5c0cb757
  Simon Pilgrim authored Nov 07, 2022
  5c0cb757
- InstCombine: Fold compare with smallest normal if input denormals are flushed · 0f68ffe1
  Matt Arsenault authored Nov 03, 2022
  0f68ffe1
- [NFC] Move getDebugValueLoc from static in Local.cpp to DebugInfo.h · 80378a4c
  OCHyams authored Nov 07, 2022
  80378a4c
- [SLP]Redesign vectorization of the gather nodes. · 8ddd1ccd
  Alexey Bataev authored Sep 16, 2022
  8ddd1ccd
- [Assignment Tracking][5.1/*] Add deleteAssignmentMarkers function · 4c44fa1c
  OCHyams authored Nov 07, 2022
  4c44fa1c
- [AArch64] Add SME2.1 target feature for Armv9-A 2022 Architecture Extension · 12a6572d
  David Sherwood authored Nov 03, 2022
  12a6572d
- [MemCpyOpt] Move lifetime marker before call to enable call slot optimization · 9a45e4be
  Nikita Popov authored Oct 13, 2022
  9a45e4be
- [AMDGPU][MC][GFX11][NFC] Correct VINTERP src operands · 8f689521
  Dmitry Preobrazhensky authored Nov 07, 2022
  8f689521
- [AMDGPU][MC][GFX10+] Enable literal operands with permlane16/permlanex16 · 6e279f5b
  Dmitry Preobrazhensky authored Nov 07, 2022
  6e279f5b
- [X86] Flatten WriteShift/Rotate SchedRW defs · 4aabbc0c
  Simon Pilgrim authored Nov 07, 2022
  4aabbc0c
- [Assignment Tracking][5/*] Add core infrastructure for instruction reference · 171f7024
  OCHyams authored Nov 07, 2022
  171f7024
- [Assignment Tracking][4/*] Add llvm.dbg.assign intrinsic boilerplate · c37f29c4
  OCHyams authored Nov 07, 2022
  c37f29c4
- [InstSimplify] (~A & B) | ~(A | B) --> ~A with logical and · b46427b9
  David Green authored Nov 07, 2022
  b46427b9
- [SWP] Recognize mem carried dep with different base · c8be3529
  Thomas Preud'homme authored Oct 21, 2022
  c8be3529
- Recommit [AArch64] Improve codegen for shifted mask op · 83255c4a
  chenglin.bi authored Nov 07, 2022
  83255c4a
- [Assignment Tracking][3/*] Add DIAssignID metadata boilerplate · a2620e00
  OCHyams authored Nov 03, 2022
  a2620e00
- GlobalISel: Pass through AA metadata for target memory intrinsics · 162d9030
  Matt Arsenault authored Nov 05, 2022
  162d9030
- [PowerPC] Use llvm::is_contained (NFC) · 7c2b76f7
  Kazu Hirata authored Nov 06, 2022
  
  7c2b76f7
- Canonicalize PowerPC detection macros to __powerpc__ · 6c927f2a
  Fangrui Song authored Nov 06, 2022
  
  6c927f2a
Nov 06, 2022
- [IndVars] Forget the SCEV when the instruction has been sunk. · 49143f9d
  luxufan authored Nov 06, 2022
  49143f9d
- [X86] Schedule scalar movsx/movzx load+extend ops as WriteLoad instead of WriteALULd · 4e56aa25
  Simon Pilgrim authored Nov 06, 2022
  4e56aa25
- [SimplifyLibCalls] improve code readability for AttributeList propagation; NFC · bff6880a
  Sanjay Patel authored Nov 06, 2022
  bff6880a
- [InstCombine] reduce multi-use casts+masks · 1c6ebe29
  Sanjay Patel authored Nov 06, 2022
  1c6ebe29
- [X86] Fix scalar load latencies for WriteLoad scheduler class · 08fe55b3
  Simon Pilgrim authored Nov 06, 2022
  08fe55b3
- Revert rG244331ae833aaf33503bbd36890e704afb66a237 "[X86] Fix scalar load... · 6fff3bab
  Simon Pilgrim authored Nov 06, 2022
  6fff3bab
- [X86] Fix scalar load latencies for WriteLoad scheduler class · 244331ae
  Simon Pilgrim authored Nov 06, 2022
  244331ae
- Revert "[SLP] Extend reordering data of tree entry to support PHI nodes" · 656f1d8b
  David Green authored Nov 06, 2022
  656f1d8b
- Revert "[llvm] [cmake] Set EXCLUDE_FROM_ALL on gtest and TestingSupport" · 61068164
  Michał Górny authored Nov 06, 2022
  61068164
- [llvm] [cmake] Set EXCLUDE_FROM_ALL on gtest and TestingSupport · b6be37b0
  Michał Górny authored Oct 30, 2022
  b6be37b0
Nov 05, 2022
- [X86] Replace unnecessary int2float and float2double overrides with better base class defs · edf88553
  Simon Pilgrim authored Nov 05, 2022
  edf88553
- [NFC] Refactor DAGCombiner::foldSelectOfConstants to reduce nesting 2.0 · 82209fd9
  Amaury Séchet authored Nov 05, 2022
  
  82209fd9
- [SimpleLoopUnswitch] Forget block & loop dispos during trivial unswitch. · a41cb8bf
  Florian Hahn authored Nov 05, 2022
  a41cb8bf
- [NFC] Refactor DAGCombiner::foldSelectOfConstants to reduce nesting · 7c05f092
  Amaury Séchet authored Nov 05, 2022
  
  7c05f092
- [X86] Cleanup WriteCvtSD2SS/WriteCvtPD2PS overrides · 2c79186b
  Simon Pilgrim authored Nov 05, 2022
  2c79186b
- [X86] Fix cvtss2si64/cvttss2si64 typo in SkylakeClient · 0b7f3278
  Simon Pilgrim authored Nov 05, 2022
  0b7f3278
- [X86] Fix override for CVTPD2PS/CVTPD2DQ/CVTTPD2DQ AVX variants · b781ca4d
  Simon Pilgrim authored Nov 05, 2022
  b781ca4d
- [X86] Add SchedWriteVecTruncate scheduler per-width wrapper · ee1ad1a6
  Simon Pilgrim authored Nov 05, 2022
  ee1ad1a6
- [X86] Add missing ReadAfterFold attributes from CVTSD2SSrm/CVTSS2SDrm to match... · 184817d0
  Simon Pilgrim authored Nov 05, 2022
  184817d0