Skip to content

Pull requests: openxla/xla

New pull request New
3,444 Open 21,327 Closed
3,444 Open 21,327 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Propagate loop metadata in double_buffer_loop_unrolling.
#25275 opened Apr 15, 2025 by jreiffers Loading…
1
NFC: Use the free function variants for dyn_cast/cast/isa/....
#25274 opened Apr 15, 2025 by copybara-service bot Loading…
@chsigg
Move fusion-dispatch to RunBackend.
#25273 opened Apr 15, 2025 by jreiffers Loading…
7
[xla:gpu] Allow S4 in generic triton emitter.
#25272 opened Apr 15, 2025 by copybara-service bot Loading…
@chsigg
Add DebugOptions to gpu backend base class.
#25268 opened Apr 15, 2025 by copybara-service bot Loading…
[XLA:GPU] Add simple triton support test for tuple ops
#25266 opened Apr 15, 2025 by copybara-service bot Loading…
[XLA:GPU] Move cuda_collectives static methods to the cuda_executor.
#25265 opened Apr 15, 2025 by copybara-service bot Loading…
[XLA:GPU] Move collectives registration from collective_thunk to stream_executor.
#25263 opened Apr 15, 2025 by copybara-service bot Loading…
PR #23315: [ROCM][NFC] GpuBlaslt matmul thunk cache refactoring part II
#25262 opened Apr 15, 2025 by copybara-service bot Loading…
[XLA:GPU] Remove the xla_gpu_triton_gemm_disable_reduced_precision_reduction flag.
#25260 opened Apr 15, 2025 by copybara-service bot Loading…
@thomasjoerg
Automated Code Change
#25259 opened Apr 15, 2025 by copybara-service bot Loading…
Change unique_ptr to shared_ptr for latency_estimator and async_tracker in latency_hiding_scheduler, so that we can potentially reuse them across different users without needing to re-construct new ones.
#25258 opened Apr 15, 2025 by copybara-service bot Loading…
PR #18410: [XLA:CPU][oneDNN] Absorb Transpose into matmul whenever possible
#25257 opened Apr 15, 2025 by copybara-service bot Loading…
Automated Code Change
#25256 opened Apr 15, 2025 by copybara-service bot Loading…
Automated Code Change
#25254 opened Apr 15, 2025 by copybara-service bot Loading…
Port Client tests to HloTestBase where possible.
#25253 opened Apr 15, 2025 by copybara-service bot Loading…
@nvgrw
Implement ComputeAndCompareR1U8 in ClientLibraryTestRunnerMixin.
#25252 opened Apr 15, 2025 by copybara-service bot Loading…
@nvgrw
Integrate LLVM at llvm/llvm-project@179d30f8c3fd
#25250 opened Apr 15, 2025 by copybara-service bot Loading…
PR #18410: [XLA:CPU][oneDNN] Absorb Transpose into matmul whenever possible
#25249 opened Apr 15, 2025 by copybara-service bot Loading…
Run build_cleaner on BUILD file(s) located in /xla/stream_executor/integrations.
#25248 opened Apr 15, 2025 by copybara-service bot Loading…
@ghpvnist
[code cleanup] deprecate the xla_gpu_enable_pipelined_collectives flag in XLA GPU
#25247 opened Apr 15, 2025 by copybara-service bot Loading…
1. Add a new Printer implementation, based on HighwayHash, that incrementally computes fingerprints.
#25246 opened Apr 14, 2025 by copybara-service bot Loading…
Will not be needed.
#25245 opened Apr 14, 2025 by copybara-service bot Loading…
@sparc1998
Removed deprecated CoordinationService methods and types.
#25242 opened Apr 14, 2025 by copybara-service bot Loading…
Replace string, char* etc with absl::string_view.
#25241 opened Apr 14, 2025 by copybara-service bot Loading…
ProTip! Filter pull requests by the default branch with base:main.