Commits · de4b88d9a1b961048864a17123b7a36974dc2b0d · Roger Ferrer / llvm-epi · GitLab

Oct 10, 2017
- [AMDGPU] Lower enqueued blocks and generate runtime metadata · de4b88d9
  Yaxun Liu authored Oct 10, 2017
  de4b88d9
- [DAGCombine] Fix for shuffle to vector extend for non power 2 vectors · 51c1b228
  David Stuttard authored Oct 10, 2017
  51c1b228
- AMDGPU: Split MUBUF offset into aligned components · 312b64f4
  Nicolai Haehnle authored Oct 10, 2017
  312b64f4
Oct 06, 2017
- [AMDGPU] New 64 bit div/rem expansion · de42c29a
  Stanislav Mekhanoshin authored Oct 06, 2017
  de42c29a
Oct 05, 2017
- [PassManager] Run global optimizations after the inliner. · ff829cea
  Davide Italiano authored Oct 05, 2017
  ff829cea
- AMDGPU: Set v2i32 any_extend to expand · 2d3f8f33
  Matt Arsenault authored Oct 05, 2017
  2d3f8f33
- AMDGPU: Add and set AMDGPU-specific e_flags · aa0835a7
  Konstantin Zhuravlyov authored Oct 05, 2017
  aa0835a7
- AMDGPU: Do not fold clamp instructions when sources are different · aafff87d
  Matt Arsenault authored Oct 05, 2017
  aafff87d
- AMDGPU: Fix not accounting for instruction size in bundles · 9ab1fa68
  Matt Arsenault authored Oct 04, 2017
  9ab1fa68
- AMDGPU: Correctly set EI_OSABI based on the os · 8684f7b4
  Konstantin Zhuravlyov authored Oct 04, 2017
  8684f7b4
Oct 03, 2017
- AMDGPU: Expand setcc for v2i32 and v4i32 · 908fa90b
  Konstantin Zhuravlyov authored Oct 03, 2017
  908fa90b
- [AMDGPU] implemented pal metadata · 72800f04
  Tim Renouf authored Oct 03, 2017
  72800f04
- [AMDGPU] Avoid predicated execution of the basic blocks containing scalar · 46513965
  Alexander Timofeev authored Oct 03, 2017
  46513965
- Revert "Re-enable "[MachineCopyPropagation] Extend pass to do COPY source forwarding"" · fabedbad
  Geoff Berry authored Oct 03, 2017
  fabedbad
- Re-enable "[MachineCopyPropagation] Extend pass to do COPY source forwarding" · bfc5fb45
  Geoff Berry authored Oct 02, 2017
  bfc5fb45
Oct 02, 2017
- AMDGPU: Fix potentially incorrectly matching check lines · b866d101
  Matt Arsenault authored Oct 02, 2017
  b866d101
- Eliminate ftrunc if source is know to be rounded · 6deb2125
  Stanislav Mekhanoshin authored Oct 02, 2017
  6deb2125
Sep 30, 2017
- [AMDGPU] Set fast-math flags on functions given the options · 1d8cf2be
  Stanislav Mekhanoshin authored Sep 29, 2017
  1d8cf2be
- CodeGen: Fix pointer info in expandUnalignedLoad/Store · b33607e5
  Yaxun Liu authored Sep 29, 2017
  b33607e5
Sep 29, 2017
- AMDGPU: VALU carry-in and v_cndmask condition cannot be EXEC · ce4ddd06
  Nicolai Haehnle authored Sep 29, 2017
  ce4ddd06
- [AMDGPU] calling conventions for AMDPAL OS type · ef1ae8ff
  Tim Renouf authored Sep 29, 2017
  ef1ae8ff
- [AMDGPU] AMDPAL scratch buffer support · 13229158
  Tim Renouf authored Sep 29, 2017
  13229158
- [Triple] Add AMDPAL operating system type · 9f7ead33
  Tim Renouf authored Sep 29, 2017
  9f7ead33
Sep 21, 2017
- AMDGPU: Add option to stress calls · 1390af2d
  Matt Arsenault authored Sep 21, 2017
  1390af2d
- AMDGPU: Fix crash on immediate operand · fdcdd88d
  Matt Arsenault authored Sep 21, 2017
  fdcdd88d
Sep 20, 2017
- AMDGPU: Start selecting v_mad_mixhi_f16 · 8cbb4884
  Matt Arsenault authored Sep 20, 2017
  8cbb4884
- AMDGPU: Start selecting v_mad_mixlo_f16 · 76935122
  Matt Arsenault authored Sep 20, 2017
  76935122
- AMDGPU: Fix encoding of op_sel for mad_mix* opcodes · 644883ff
  Matt Arsenault authored Sep 20, 2017
  644883ff
- AMDGPU: Match load d16 hi instructions · b81495dc
  Matt Arsenault authored Sep 20, 2017
  b81495dc
- [AMDGPU] Port of HSAIL inliner · 5670e6d4
  Stanislav Mekhanoshin authored Sep 20, 2017
  5670e6d4
- AMDGPU: Match store d16_hi instructions · fcc213fa
  Matt Arsenault authored Sep 20, 2017
  fcc213fa
Sep 19, 2017
- [AMDGPU] Prevent post-RA scheduler from breaking memory clauses · d4ae470d
  Stanislav Mekhanoshin authored Sep 19, 2017
  d4ae470d
- AMDGPU: Run internalize symbols at -O0 · e745d996
  Matt Arsenault authored Sep 19, 2017
  e745d996
Sep 18, 2017
- AMDGPU: Start selecting s_xnor_{b32, b64} · ca8946a3
  Konstantin Zhuravlyov authored Sep 18, 2017
  ca8946a3
Sep 14, 2017
- AMDGPU: Fix violating constant bus restriction · c317287f
  Matt Arsenault authored Sep 14, 2017
  c317287f
- AMDGPU: Fix assert on alloca of array of struct · 37ab4cf8
  Matt Arsenault authored Sep 14, 2017
  37ab4cf8
- AMDGPU: Stop modifying SP in call sequences · defe3717
  Matt Arsenault authored Sep 14, 2017
  defe3717
- AMDGPU: Make frame register caller preserved · 6efd082c
  Matt Arsenault authored Sep 14, 2017
  6efd082c
- [DAGCombine] (shl (or x, c1), c2) -> (or (shl x, c2), c1 << c2) · 8bd2d878
  Simon Pilgrim authored Sep 14, 2017
  8bd2d878
- Fix line endings. NFCI. · 337b2d00
  Simon Pilgrim authored Sep 14, 2017
  337b2d00