Commits

Branch selector

 master

User selector

nhaehnle

Datepicker

All time

Commit History

Commits on Jan 31, 2017

• [DAGCombine] require UnsafeFPMath for re-association of addition

  Show description for ab43652
  nhaehnle

  committed
  ab43652

Commits on Dec 12, 2016

• AMDGPU: llvm.amdgcn.interp.mov is a source of divergence

  Show description for ddba2a5
  nhaehnle

  committed
  ddba2a5

Commits on Dec 8, 2016

• [SelectionDAG] Add expansion and promotion of [US]MUL_LOHI

  Show description for 616da7e
  nhaehnle

  committed
  616da7e

• X86: Add checks for fma_patterns[_wide].ll with -enable-no-infs-fp-math

  Show description for 09d35d9
  nhaehnle

  committed
  09d35d9

• AMDGPU: Properly implement SIRegisterInfo::isFrameOffsetLegal and needsFrameBaseReg

  Show description for e299df1
  nhaehnle

  committed
  e299df1

Commits on Dec 2, 2016

• [DAGCombiner] do not fold (fmul (fadd X, 1), Y) -> (fmad X, Y, Y) by default

  Show description for 42285f5
  nhaehnle

  committed
  42285f5

Commits on Dec 1, 2016

• [SelectionDAG] Rename and clarify visitFMULForFMADCombine (NFC)

  Show description for 77151a0
  nhaehnle

  committed
  77151a0

Commits on Nov 30, 2016

• [SelectionDAG] Refactor TargetLowering::expandMUL (NFC)

  Show description for 989c694
  nhaehnle

  committed
  989c694

Commits on Nov 23, 2016

• [SelectionDAG] Early-out in TargetLowering::expandMUL (NFC)

  Show description for da8f0a5
  nhaehnle

  committed
  da8f0a5

Commits on Nov 18, 2016

• AMDGPU: Fix legalization of MUBUF instructions in shaders

  Show description for 1e27a61
  nhaehnle

  committed
  1e27a61

Commits on Nov 3, 2016

• DAGCombiner: fix use-after-free when merging consecutive stores

  Show description for de5ede6
  nhaehnle

  committed
  de5ede6

Commits on Nov 2, 2016

• AMDGPU: Allow additional implicit operands on MOVRELS instructions

  Show description for 08f7b24
  nhaehnle

  committed
  08f7b24

Commits on Oct 27, 2016

• AMDGPU: Fix SILoadStoreOptimizer when writes cannot be merged due register dependencies

  Show description for ce0a523
  nhaehnle

  committed
  ce0a523

Commits on Oct 24, 2016

• AMDGPU: Fix Two Address problems with v_movreld

  Show description for e07033a
  nhaehnle

  committed
  e07033a

Commits on Oct 14, 2016

• AMDGPU: Select 64-bit {ADD,SUB}{C,E} nodes

  Show description for 877e3be
  nhaehnle

  committed
  877e3be

• Fix use-after-frees

  Show description for 9b21983
  nhaehnle

  committed
  9b21983

• AMDGPU: Fix use-after-frees

  Show description for 5d66203
  nhaehnle

  committed
  5d66203

Commits on Oct 7, 2016

• AMDGPU: Fix use-after-free in SIOptimizeExecMasking

  Show description for f3907ed
  nhaehnle

  committed
  f3907ed

Commits on Sep 12, 2016

• AMDGPU: Do not clobber SCC in SIWholeQuadMode

  Show description for 01a133c
  nhaehnle

  committed
  01a133c

Commits on Sep 3, 2016

• AMDGPU: Reduce the duration of whole-quad-mode

  Show description for f9f3b70
  nhaehnle

  committed
  f9f3b70

• AMDGPU: Fix an interaction between WQM and polygon stippling

  Show description for c280d74
  nhaehnle

  committed
  c280d74

Commits on Aug 17, 2016

• LiveIntervals: add removeRegUnit

  Show description for 6811218
  nhaehnle

  committed
  6811218

Commits on Aug 10, 2016

• LiveIntervalAnalysis: fix a crash in repairOldRegInRange

  Show description for be7124c
  nhaehnle

  committed
  be7124c

Commits on Aug 5, 2016

• [InstCombine] try to fold (select C, (sext A), B) into logical ops

  Show description for fd14e43
  nhaehnle

  committed
  fd14e43

Commits on Aug 3, 2016

• [InstCombine] Cleanup select-bitext.ll tests

  Show description for d22c167
  nhaehnle

  committed
  d22c167

• [InstCombine] Add select-bitext.ll tests

  Show description for 41fc0ec
  nhaehnle

  committed
  41fc0ec

Commits on Aug 2, 2016

• AMDGPU: Stay in WQM for non-intrinsic stores

  Show description for 87d2983
  nhaehnle

  committed
  87d2983

• AMDGPU: Track physical registers in SIWholeQuadMode

  Show description for a846118
  nhaehnle

  committed
  a846118

Commits on Jul 28, 2016

• AMDGPU: add execfix flag to SI_ELSE

  Show description for b18ca96
  nhaehnle

  committed
  b18ca96

Commits on Jul 18, 2016

• AMDGPU: Disable AMDGPUPromoteAlloca pass for shader calling conventions.

  Show description for 0c05ce4
  nhaehnle

  committed
  0c05ce4

Commits on Jul 12, 2016

• AMDGPU: Unify MOVRELSOffset and MOVRELDOffset

  Show description for 4d01feb
  nhaehnle

  committed
  4d01feb

Commits on Jul 11, 2016

• AMDGPU: Treat texture gather instructions more like other MIMG instructions

  Show description for ac1e47a
  nhaehnle

  committed
  ac1e47a

• AMDGPU: fix local stack slot allocation bugs

  Show description for a48da98
  nhaehnle

  committed
  a48da98

• [Sink] Don't move calls to readonly functions across stores

  Show description for bd10da1
  nhaehnle

  committed
  bd10da1

• AliasAnalysis: unify getModRefInfo(I, CS) semantics with other overloads

  Show description for 0f5a9cf
  nhaehnle

  committed
  0f5a9cf

Pagination

PreviousNext