Commits · f7866fad5415ff8406fe7fa453fed1c4b08f3b92 · Roger Ferrer / llvm-epi · GitLab

Dec 02, 2016
- AMDGPU: Use wider scalar spills for SGPR spilling · c47701c0
  Matt Arsenault authored Dec 02, 2016
  c47701c0
Dec 01, 2016
- RegisterCoalscer: Only coalesce complete reserved registers. · 709a4cc2
  Matthias Braun authored Dec 01, 2016
  709a4cc2
Nov 30, 2016
- AMDGPU: Move mir tests into mir test directory · 387afc93
  Matt Arsenault authored Nov 30, 2016
  387afc93
Nov 29, 2016
- AMDGPU: Disallow exec as SMEM instruction operand · 640c44b8
  Matt Arsenault authored Nov 29, 2016
  640c44b8
- AMDGPU: Materialize frame index before add · f96eeec0
  Matt Arsenault authored Nov 29, 2016
  f96eeec0
- AMDGPU/SI: Avoid moving PHIs to VALU when phi values are defined in scalar branches · 0bc68811
  Tom Stellard authored Nov 29, 2016
  0bc68811
Nov 28, 2016
- [AMDGPU] Allow hoisting of comparisons out of a loop and eliminate condition copies · 0ee250ee
  Stanislav Mekhanoshin authored Nov 28, 2016
  0ee250ee
Nov 26, 2016
- AMDGPU/SI: Use float as the operand type for amdgcn.interp intrinsics · 1473f07c
  Tom Stellard authored Nov 26, 2016
  1473f07c
Nov 25, 2016
- AMDGPU/SI: Add back reverted SGPR spilling code, but disable it · 79c05871
  Marek Olsak authored Nov 25, 2016
  79c05871
- Revert "AMDGPU: Implement SGPR spilling with scalar stores" · e3895bfb
  Marek Olsak authored Nov 25, 2016
  e3895bfb
- Revert "AMDGPU: Fix MMO when splitting spill" · dad553a5
  Marek Olsak authored Nov 25, 2016
  dad553a5
- Revert "AMDGPU: Make m0 unallocatable" · a45dae45
  Marek Olsak authored Nov 25, 2016
  a45dae45
- Revert "AMDGPU: Preserve m0 value when spilling" · 18a95bcb
  Marek Olsak authored Nov 25, 2016
  18a95bcb
Nov 24, 2016
- AMDGPU: Preserve m0 value when spilling · 7b54dd03
  Matt Arsenault authored Nov 24, 2016
  7b54dd03
- AMDGPU: Make m0 unallocatable · 9e5c7b10
  Matt Arsenault authored Nov 24, 2016
  9e5c7b10
Nov 23, 2016
- AMDGPU: Fix MMO when splitting spill · 2669a76f
  Matt Arsenault authored Nov 23, 2016
  2669a76f
Nov 22, 2016
- [AMDGPU] Fix multiple vreg definitions in si-lower-control-flow · ae0f6620
  Stanislav Mekhanoshin authored Nov 22, 2016
  ae0f6620
Nov 21, 2016
- DAG: Ignore call site attributes when emitting target intrinsic · b30d2aca
  Matt Arsenault authored Nov 21, 2016
  b30d2aca
Nov 18, 2016
- [AMDGPU] Change frexp.exp intrinsic to return i16 for f16 input · aefee42e
  Konstantin Zhuravlyov authored Nov 18, 2016
  aefee42e
- AMDGPU/SI: Remove zero_extend patterns for i16 ops selected to 32-bit insts · 01e65d2c
  Tom Stellard authored Nov 18, 2016
  01e65d2c
- AMDGPU: Fix legalization of MUBUF instructions in shaders · ce2b589d
  Nicolai Haehnle authored Nov 18, 2016
  ce2b589d
- AMDGPU: Fix crash on illegal type for inlineasm · 742deb24
  Matt Arsenault authored Nov 18, 2016
  742deb24
Nov 17, 2016
- Revert "AMDGPU: Enable ConstrainCopy DAG mutation" · 0a1a7b6b
  Konstantin Zhuravlyov authored Nov 17, 2016
  0a1a7b6b
- [AMDGPU] Add missing test for rL287203 · 20ba24e2
  Konstantin Zhuravlyov authored Nov 17, 2016
  20ba24e2
- [AMDGPU] Promote f16/i16 conversions to f32/i32 · 3f0cdc7a
  Konstantin Zhuravlyov authored Nov 17, 2016
  3f0cdc7a
- [AMDGPU] Expand `br_cc` for f16 · 662e01df
  Konstantin Zhuravlyov authored Nov 17, 2016
  662e01df
Nov 16, 2016
- AMDGPU: Enable ConstrainCopy DAG mutation · 3b36bb1d
  Matt Arsenault authored Nov 16, 2016
  3b36bb1d
- AMDGPU/SI: Avoid creating unnecessary copies in the SIFixSGPRCopies pass · 0d162b1c
  Tom Stellard authored Nov 16, 2016
  0d162b1c
- [AMDGPU] Handle f16 select{_cc} · 2a87a420
  Konstantin Zhuravlyov authored Nov 16, 2016
  2a87a420
- AMDGPU/GCN: Exit early in hazard recognizer if there is no vreg argument · e8cc395e
  Jan Vesely authored Nov 15, 2016
  e8cc395e
Nov 15, 2016
- AMDGPU/SI: Fix pattern for i16 = sign_extend i1 · d23de360
  Tom Stellard authored Nov 15, 2016
  d23de360
- AMDGPU: Enable store clustering · d4bb5e48
  Matt Arsenault authored Nov 15, 2016
  d4bb5e48
- AMDGPU: Analyze mubuf with immediate soffset · 36666298
  Matt Arsenault authored Nov 15, 2016
  36666298
- [AMDGPU] Add wave barrier builtin · ea91cca5
  Stanislav Mekhanoshin authored Nov 15, 2016
  ea91cca5
- AMDGPU: Fix f16 fabs/fneg · c79dc70d
  Matt Arsenault authored Nov 15, 2016
  c79dc70d
- AMDGPU: Fix formatting of 1/2pi immediate · 972034bd
  Matt Arsenault authored Nov 15, 2016
  972034bd
Nov 14, 2016
- AMDGPU/SI: Support data types other than V4f32 in image intrinsics · 8236fe10
  Changpeng Fang authored Nov 14, 2016
  8236fe10
Nov 13, 2016
- AMDGPU: Implement SGPR spilling with scalar stores · dc45274d
  Matt Arsenault authored Nov 13, 2016
  dc45274d
- [AMDGPU] Add f16 support (VI+) · f86e4b72
  Konstantin Zhuravlyov authored Nov 13, 2016
  f86e4b72
Nov 12, 2016
- AMDGPU/SI: Promote i16 = fp_[us]int f32 for VI · b4c8e8e3
  Tom Stellard authored Nov 12, 2016
  b4c8e8e3