Skip to content

Pull requests: openxla/xla

New pull request New
464 Open 34,064 Closed
464 Open 34,064 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[XLA:GPU] Update HLO cublas workspace size after autotuner select the algorithm
#35132 opened Dec 11, 2025 by shawnwang18 Loading…
Support padding in convolution op with YNNPACK enabled.
#35131 opened Dec 11, 2025 by copybara-service bot Loading…
Automated Code Change
#35130 opened Dec 11, 2025 by copybara-service bot Loading…
Automated Code Change
#35129 opened Dec 11, 2025 by copybara-service bot Loading…
Automated Code Change
#35127 opened Dec 11, 2025 by copybara-service bot Loading…
Remove redundant environment variables.
#35126 opened Dec 11, 2025 by copybara-service bot Loading…
@ybaturina
1
Minor readability cleanup.
#35125 opened Dec 11, 2025 by copybara-service bot Loading…
@toli-y
Test nightly TPU.
#35122 opened Dec 10, 2025 by copybara-service bot Loading…
@ybaturina
1
PR #34143: [GPU] Add all-to-all support to S-curve model
#35121 opened Dec 10, 2025 by copybara-service bot Loading…
Integrate LLVM at llvm/llvm-project@48d942c7158a
#35120 opened Dec 10, 2025 by copybara-service bot Loading…
Fix memory leak in MakeShapesInfo() and re-enable layout conversion.
#35119 opened Dec 10, 2025 by copybara-service bot Loading…
@matthiaskramm
Pass HERMETIC_PYTHON_URL in python_register_toolchains as a single URL.
#35116 opened Dec 10, 2025 by copybara-service bot Loading…
@belitskiy
[XLA:GPU] Fix combine decomposition when replica groups are shuffled.
#35115 opened Dec 10, 2025 by copybara-service bot Loading…
@olegshyshkov
Enqueue cross-host send after send buffer definition events are recorded, not complete
#35113 opened Dec 10, 2025 by rao-ashish Loading…
Remove unused CoordinationClientCache.
#35111 opened Dec 10, 2025 by copybara-service bot Loading…
@mwhittaker
[XLA:GPU] ignored dim=1 sizes when swapping broadcast and bitcast
#35109 opened Dec 10, 2025 by copybara-service bot Loading…
@metaflow
[ROCm] Fix crashes in hermetic build tests
#35108 opened Dec 10, 2025 by alekstheod Loading…
@akuegel
1
Remove undefined support of tuple shardings from NumTiles() and TiledDataRank() methods. These methods should only be defined for non-tuple shardings, when sharding is tuple tile_assignment_ is not populated therefore these would be returning undefined results.
#35107 opened Dec 10, 2025 by copybara-service bot Loading…
@KanishAnand
[Autotuner] Fix behavior when cublas is disabled and autotune level is set to 0.
#35106 opened Dec 10, 2025 by copybara-service bot Loading…
@derdrdirk
Add serialization of ExecutionState for CustomCallThunk.
#35104 opened Dec 10, 2025 by copybara-service bot Loading…
@khasanovaa
Implement HloSharding::dimensions() for NamedSharding
#35103 opened Dec 10, 2025 by copybara-service bot Loading…
@KanishAnand
[XLA:GPU] Update memory allocator API to support VA allocation, VA/PA remap
#35101 opened Dec 10, 2025 by shawnwang18 • Draft
1
[XLA] Fix ASSERT_OK_AND_ASSIGN macro in patch
#35099 opened Dec 10, 2025 by copybara-service bot Loading…
@dextero
[GPU] Wrap single instructions in fusions before autotuning.
#35098 opened Dec 10, 2025 by sergachev Loading…
Enable CublasLt via GemmRewriter configuration. Moves DebugOptions flag xla_gpu_enable_cublas_lt out of GemmRewriter pass.
#35097 opened Dec 10, 2025 by copybara-service bot Loading…
@derdrdirk
ProTip! Mix and match filters to narrow down what you’re looking for.