Commits

Branch selector

 master

User selector

atkassen

Datepicker

All time

Commit History

Commits on Aug 1, 2024

• gpu: generic: sum: update copyright

  atkassen

  committed
  386e88f

  Copy full SHA for 386e88f

• tests: benchdnn: inputs: reorder: add fp8 transpose case

  atkassen

  committed
  c542db9

  Copy full SHA for c542db9

• xe: reorder: allow fp8 without conversion

  atkassen

  committed
  94aaf0b

  Copy full SHA for 94aaf0b

Commits on Jul 22, 2024

• xe: concat: fix write index for zero padding

  atkassen

  committed
  169be1d

  Copy full SHA for 169be1d

Commits on Jul 2, 2024

• xpu: ocl: enable profiling for copy

  atkassen

  committed
  e565857

  Copy full SHA for e565857

Commits on Jun 11, 2024

• gpu: fixup direct copy

  atkassen

  committed
  d690012

  Copy full SHA for d690012

Commits on Jun 7, 2024

• gpu: move direct_copy to gpu::generic

  atkassen

  committed
  fdc3cd2

  Copy full SHA for fdc3cd2

Commits on Jun 4, 2024

• gpu: intel: ocl: reorder: avoid extraneous copies

  atkassen

  committed
  1edbb54

  Copy full SHA for 1edbb54

• gpu: intel: jit: reorder: avoid extraneous copies

  atkassen

  committed
  6264136

  Copy full SHA for 6264136

Commits on May 30, 2024

• tests: benchdnn: allow empty cache blob for direct copy

  atkassen

  committed
  d326047

  Copy full SHA for d326047

• gpu: intel: ocl: add direct copy reorder

  atkassen

  committed
  b3505dc

  Copy full SHA for b3505dc

Commits on May 8, 2024

• gpu: intel: ocl: reorder: explicitly promote offset

  atkassen

  committed
  3fd77ce

  Copy full SHA for 3fd77ce

Commits on Apr 26, 2024

• tests: benchdnn: inputs: binary: add mixed src1/post-op src broadcast cases

  atkassen

  committed
  68cd88d

  Copy full SHA for 68cd88d

• gpu: intel: ocl: binary: fix post-op tensor offset

  atkassen

  committed
  d042a55

  Copy full SHA for d042a55

Commits on Apr 25, 2024

• gpu: add intel namespace for dev mode impls

  atkassen

  committed
  843b6c5

  Copy full SHA for 843b6c5

Commits on Apr 24, 2024

• gpu: intel: jit: ir: fallback to 1D messages for narrow blocks

  atkassen

  committed
  e05ec6e

  Copy full SHA for e05ec6e

• gpu: intel: jit: reorder: enable 2d sends

  atkassen

  committed
  457b8e3

  Copy full SHA for 457b8e3

• gpu: intel: jit: ir: check memory offset alignment for 2d sends

  atkassen

  committed
  32932bf

  Copy full SHA for 32932bf

• gpu: intel: jit: ir: message: refactor get_send_params

  Show description for 49a5b86
  atkassen

  committed
  49a5b86

  Copy full SHA for 49a5b86

Commits on Apr 9, 2024

• gpu: jit: codegen: add missing Xe2 get_binary case

  atkassen

  committed
  7607a4b

  Copy full SHA for 7607a4b

Commits on Mar 13, 2024

• doc: add Xe2 to list of GPU ISA build options

  atkassen

  committed
  6838589

  Copy full SHA for 6838589

• {build,common}: support Xe2 build guard

  Show description for 173de73
  atkassen

  committed
  173de73

  Copy full SHA for 173de73

Commits on Mar 12, 2024

• gpu: ocl: reorder: support large buffers

  atkassen

  committed
  c160cd7

  Copy full SHA for c160cd7

Commits on Feb 27, 2024

• tests: benchdnn: inputs: reorder: add integer overflow test cases

  atkassen

  committed
  5cbc0df

  Copy full SHA for 5cbc0df

• gpu: jit: reorder: avoid exceeding ND range limits

  atkassen

  committed
  02dbca7

  Copy full SHA for 02dbca7

• gpu: jit: ir: avoid integer overflow in loop split

  atkassen

  committed
  f1367fa

  Copy full SHA for f1367fa

Commits on Feb 13, 2024

• gpu: compute: correct max execution size on Xe2

  atkassen

  committed
  3100be3

  Copy full SHA for 3100be3

Commits on Feb 12, 2024

• gpu: jit: ir: ease pointer alignment restriction on Xe2

  atkassen

  committed
  085a282

  Copy full SHA for 085a282

• gpu: jit: ir: fix 2d send alignment on Xe2

  atkassen

  committed
  b43f619

  Copy full SHA for b43f619

Commits on Feb 9, 2024

• gpu: ocl: ref concat: fix status check

  atkassen

  committed
  7b7ab5b

  Copy full SHA for 7b7ab5b

Commits on Feb 7, 2024

• gpu: jit: codegen: reorder: avoid unnecessary copies

  atkassen

  committed
  771de94

  Copy full SHA for 771de94

Commits on Jan 29, 2024

• gpu: ocl: atomic reduction: use public serialized_t ctor

  atkassen

  committed
  4c87e74

  Copy full SHA for 4c87e74

Commits on Jan 26, 2024

• tests: benchdnn: inputs: conv: add non-power-of-2 stride gpu regression tests

  atkassen

  committed
  e36243e

  Copy full SHA for e36243e

• gpu: jit: ir: reorder: use scalar mov with non-power-of-2 strides

  atkassen

  committed
  e62b756

  Copy full SHA for e62b756

Commits on Jan 16, 2024

• tests: benchdnn: inputs: binary: add GPU regression test

  atkassen

  committed
  7551905

  Copy full SHA for 7551905

Pagination

Previous

Next