Commits · f4d14f781f0923fba574a6a6c9b4e32552a71d68 · Roger Ferrer / llvm-epi · GitLab

Oct 08, 2015
- [SystemZ] Fix another assertion failure in tryBuildVectorShuffle · f4d14f78
  Ulrich Weigand authored Oct 08, 2015
  f4d14f78
- [InstCombine] transform masking off of an FP sign bit into a fabs() intrinsic call (PR24886) · f61a08fb
  Sanjay Patel authored Oct 08, 2015
  f61a08fb
- [ValueTracking] teach computeKnownBits that a fabs() clears sign bits · 9115cf8c
  Sanjay Patel authored Oct 08, 2015
  9115cf8c
- Fix a bug in llvm-objdump’s printing of Objective-C meta data · aac75382
  Kevin Enderby authored Oct 08, 2015
  aac75382
- Support for llvm-bcanalyzer dumping of record array strings. · b1cfcd4a
  Teresa Johnson authored Oct 08, 2015
  b1cfcd4a
- [X86] Disable X86CallFrameOptimization on Darwin in presence of EH · 263b772b
  Frederic Riss authored Oct 08, 2015
  263b772b
- AVX512: vpextrb/w/d/q and vpinsrb/w/d/q implementation. · defab3c1
  Igor Breger authored Oct 08, 2015
  defab3c1
- Compute demanded bits for icmp instructions · e9d50dc9
  James Molloy authored Oct 08, 2015
  e9d50dc9
- Treat Mul just like Add and Subtract · bcd7f0ac
  James Molloy authored Oct 08, 2015
  bcd7f0ac
- [X86] Fix wrong treatment of multi-lane blends in BUILD_VECTORtoBlendMask() · 04e79329
  Michael Kuperstein authored Oct 08, 2015
  04e79329
- Do not assert on first non-prologue instruction being a CFI directive. · 2b3c16ca
  Michael Kuperstein authored Oct 08, 2015
  2b3c16ca
- [SystemZ] SystemZElimCompare pass improved. · 5d3fbd37
  Jonas Paulsson authored Oct 08, 2015
  5d3fbd37
- [SystemZ] Use load-and-test for fp compare with 0 if vector support is present. · 7c5ce10a
  Jonas Paulsson authored Oct 08, 2015
  7c5ce10a
- Revert "Handle inline stacks in gcov-encoded sample profiles." · a082040d
  Diego Novillo authored Oct 08, 2015
  a082040d
- Handle inline stacks in gcov-encoded sample profiles. · b7fca574
  Diego Novillo authored Oct 08, 2015
  b7fca574
- [WinEH] Add missing test case for llvm.eh.exceptioncode · 94fe836a
  Reid Kleckner authored Oct 07, 2015
  94fe836a
- [WinEH] Fix 32-bit funclet epilogues in the presence of dynamic allocas · 97797419
  Reid Kleckner authored Oct 07, 2015
  97797419
Oct 07, 2015
- [WinEH] Refer to filter funclets using their symbol-table symbol · 6af5f82c
  David Majnemer authored Oct 07, 2015
  6af5f82c
- [WinEH] Undo the effect of r249578 for 32-bit · 70bf6bb5
  Reid Kleckner authored Oct 07, 2015
  70bf6bb5
- [WinEH] Remove unreachable blocks before preparation · c289c9ff
  David Majnemer authored Oct 07, 2015
  c289c9ff
- [WinEH] Set NoModuleLevelChanges in clone flags · 39234fc6
  Joseph Tremoulet authored Oct 07, 2015
  39234fc6
- [X86]Update test to use FileCheck. · 99e8c0ff
  Kevin B. Smith authored Oct 07, 2015
  99e8c0ff
- Revert "Revert "This patch builds on top of D13378 to handle constant condition."" · 044cb34b
  Mehdi Amini authored Oct 07, 2015
  044cb34b
- [AArch64] Fold a floating-point divide by power of two into fp conversion. · 7c6ac2b8
  Chad Rosier authored Oct 07, 2015
  7c6ac2b8
- [WinEH] Fix two minor issues in __CxxFrameHandler3 tables · 33bd2d99
  Reid Kleckner authored Oct 07, 2015
  33bd2d99
- [AArch64] Fold a floating-point multiply by power of two into fp conversion. · fa30c9b4
  Chad Rosier authored Oct 07, 2015
  fa30c9b4
- [IndVars] Preserve LCSSA in `eliminateIdentitySCEV` · 0015e5a0
  Sanjoy Das authored Oct 07, 2015
  0015e5a0
- [ARM] Promote helper function to SelectionDAG. · 169865ff
  Chad Rosier authored Oct 07, 2015
  169865ff
- [ARM] Use correct half-precision functions in EABI mode · d3d114ba
  Oliver Stannard authored Oct 07, 2015
  d3d114ba
- Move test back to Generic now it's fixed the right way (thanks Eric!) · 30f07f93
  David Blaikie authored Oct 07, 2015
  30f07f93
- [ARM] Prevent PerformVDIVCombine from combining a vcvt/vdiv with 8 lanes. · 17436bf6
  Chad Rosier authored Oct 07, 2015
  17436bf6
- Teach computeKnownBits to use new align attribute/metadata · d94903c9
  Artur Pilipenko authored Oct 07, 2015
  d94903c9
- [ARM][AArch64] Only lower to interleaved load/store if the target has NEON · aebca095
  Jeroen Ketema authored Oct 07, 2015
  aebca095
- Revert "This patch builds on top of D13378 to handle constant condition." · 47efaeb3
  James Molloy authored Oct 07, 2015
  47efaeb3
- [EarlyCSE] Fix handling of target memory intrinsics for CSE'ing loads. · a6178a17
  Arnaud A. de Grandmaison authored Oct 07, 2015
  a6178a17
- [X86] Emit .cfi_escape GNU_ARGS_SIZE when adjusting the stack before calls · 259f1508
  Michael Kuperstein authored Oct 07, 2015
  259f1508
- AVX512: Change encoding of vpshuflw and vpshufhw instructions. Implement WIG... · 1a6fd1cc
  Igor Breger authored Oct 07, 2015
  1a6fd1cc
- Remove the comdat-ness from the testcase as it won't lower properly · 97b9189e
  Eric Christopher authored Oct 07, 2015
  97b9189e
- Update test to use FileCheck and clean up run lines to match the · ab2802c5
  Eric Christopher authored Oct 07, 2015
  ab2802c5
- AMDGPU: Use explicit register size indirect pseudos · 28419273
  Matt Arsenault authored Oct 07, 2015
  28419273