Commits

Branch selector

 master

User selector

jvesely

Datepicker

All time

Commit History

Commits on May 27, 2015

• R600: Rely on TypeLegalizer to use divrem instead of div/rem

  Show description for e93c0d6
  jvesely

  committed
  e93c0d6

  Copy full SHA for e93c0d6

• SelectionDAG: Don't do libcall on div/rem if divrem is custom

  Show description for 72be0e4
  jvesely

  committed
  72be0e4

  Copy full SHA for 72be0e4

Commits on May 26, 2015

• R600: Use SIGN_EXTEND_INREG for SEXT loads

  Show description for 176d1fa
  jvesely

  committed
  176d1fa

  Copy full SHA for 176d1fa

• R600: Add comments to subword private address load lowering code

  Show description for 0704171
  jvesely

  committed
  0704171

  Copy full SHA for 0704171

Commits on Apr 30, 2015

• Reinstate revisions r234755, r234759, r234760

  Show description for e3d17d7
  jvesely

  committed
  e3d17d7

  Copy full SHA for e3d17d7

Commits on Apr 29, 2015

• CodeGen: Default overflow operations to expand so we don't have to assume targets are lying

  Show description for 42eeb1d
  jvesely

  committed
  42eeb1d

  Copy full SHA for 42eeb1d

Commits on Apr 13, 2015

• Revert revisions r234755, r234759, r234760

  Show description for a017ce2
  jvesely

  committed
  a017ce2

  Copy full SHA for a017ce2

• R600: Add carry and borrow instructions. Use them to implement UADDO/USUBO

  Show description for 4ce8b9c
  jvesely

  committed
  4ce8b9c

  Copy full SHA for 4ce8b9c

• LegalizeDAG: Try to use Overflow operations when expanding ADD/SUB

  Show description for 187ac42
  jvesely

  committed
  187ac42

  Copy full SHA for 187ac42

Commits on Apr 12, 2015

• R600: Make FMIN/MAXNUM legal on all asics

  Show description for b7239d3
  jvesely

  committed
  b7239d3

  Copy full SHA for b7239d3

• R600: remove manual BFE optimization

  Show description for 8332e14
  jvesely

  committed
  8332e14

  Copy full SHA for 8332e14

Commits on Mar 13, 2015

• r600: Clear visited structure before running.

  Show description for 2ccb127
  jvesely

  committed
  2ccb127

  Copy full SHA for 2ccb127

• r600: Use deque and simplify loops in AMDGPUCFGStructurizer

  Show description for 915e3b4
  jvesely

  committed
  915e3b4

  Copy full SHA for 915e3b4

Commits on Mar 2, 2015

• Support: Use const pointers for reads.

  Show description for ef20302
  jvesely

  committed
  ef20302

  Copy full SHA for ef20302

• R600: Use c++11 style for loop

  Show description for e8676a2
  jvesely

  committed
  e8676a2

  Copy full SHA for e8676a2

Commits on Jan 22, 2015

• R600: Try to use lower types for 64bit division if possible

  Show description for 1d07592
  jvesely

  committed
  1d07592

  Copy full SHA for 1d07592

• SelectionDAG: Add KnownBits and SignBits computation for EXTRACT_ELEMENT

  Show description for 3ff611b
  jvesely

  committed
  3ff611b

  Copy full SHA for 3ff611b

• R600: Simplify LowerUDIVREM

  Show description for cd0d34f
  jvesely

  committed
  cd0d34f

  Copy full SHA for cd0d34f

Commits on Oct 17, 2014

• Mips: Only set divrem i64 to custom on 64bit

  Show description for 7eee3b0
  jvesely

  committed
  7eee3b0

  Copy full SHA for 7eee3b0

• R600: Add EG to FMA test

  Show description for 9c1d1fa
  jvesely

  committed
  9c1d1fa

  Copy full SHA for 9c1d1fa

• SelectionDAG: Add sext_inreg optimizations

  Show description for cef793e
  jvesely

  committed
  cef793e

  Copy full SHA for cef793e

Commits on Oct 14, 2014

• Reapply "R600: Add new intrinsic to read work dimensions"

  Show description for d6315ea
  jvesely

  committed
  d6315ea

  Copy full SHA for d6315ea

• R600: Add new intrinsic to read work dimensions

  Show description for 6a52985
  jvesely

  committed
  6a52985

  Copy full SHA for 6a52985

• R600: FMA is VecALU only instruction

  Show description for 787e3ca
  jvesely

  committed
  787e3ca

  Copy full SHA for 787e3ca

Commits on Sep 5, 2014

• R600: Fix FROUND

  Show description for 286f644
  jvesely

  committed
  286f644

  Copy full SHA for 286f644

Commits on Sep 4, 2014

• build/cmake: Fix CMP0023 warning with libffi

  Show description for d3a511e
  jvesely

  committed
  d3a511e

  Copy full SHA for d3a511e

Commits on Aug 13, 2014

• utils: Fix segfault in flattencfg

  Show description for d3fa093
  jvesely

  committed
  d3fa093

  Copy full SHA for d3fa093

• Initialize FlattenCFG pass

  Show description for f7a325b
  jvesely

  committed
  f7a325b

  Copy full SHA for f7a325b

Commits on Aug 12, 2014

• R600: Use optimized 24bit path in udivrem

  Show description for 3c57820
  jvesely

  committed
  3c57820

  Copy full SHA for 3c57820

• R600: Remove unused code.

  Show description for 1001344
  jvesely

  committed
  1001344

  Copy full SHA for 1001344

• R600: Use i24 optimized path for SREM

  Show description for b40562c
  jvesely

  committed
  b40562c

  Copy full SHA for b40562c

Commits on Jul 31, 2014

• R600: Modernize work item intrinsics test

  Show description for 22c99ee
  jvesely

  committed
  22c99ee

  Copy full SHA for 22c99ee

Commits on Jul 15, 2014

• R600: Implement zero undef variants of ctlz/cttz

  Show description for 7b5f8a5
  jvesely

  committed
  7b5f8a5

  Copy full SHA for 7b5f8a5

Commits on Jul 10, 2014

• R600: Implement float to long/ulong

  Show description for 865527a
  jvesely

  committed
  865527a

  Copy full SHA for 865527a

• SelectionDAG: Factor FP_TO_SINT lower code out of DAGLegalizer

  Show description for bb917c2
  jvesely

  committed
  bb917c2

  Copy full SHA for bb917c2

Pagination

Previous

Next