Skip to content

Pull requests: iree-org/iree

New pull request New
158 Open 15,615 Closed
158 Open 15,615 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

[stream] Dedeplicate stream.yield values from stream.async.execute
#20056 opened Feb 21, 2025 by rsuderman Loading…
Changes to dispatch formation to allow for fusion with trunc in case of gemm + transpose gemm fusion
#20055 opened Feb 21, 2025 by MaheshRavishankar Loading…
10
[Stream] Do not specialize encodings that is either unrecognized or already serialized
#20054 opened Feb 21, 2025 by hanhanW Loading…
[Dispatch Creation] Don't propagate through attention k1 dims
#20052 opened Feb 21, 2025 by IanWood1 Loading…
1
3
[DT] Don't materialize encodings in case of missing information
#20051 opened Feb 21, 2025 by jtuyls Loading…
1
[GPU] fix bug in expand permutation calculation when concretizing MMA shapes
#20049 opened Feb 20, 2025 by nirvedhmeshram Loading…
3
[AMDGPU] Add RDNA4 targets
#20048 opened Feb 20, 2025 by krzysz00 • Draft
[WIP] Add optimization flags
#20047 opened Feb 20, 2025 by IanWood1 • Draft
2
New flag: iree-codegen-math-transform-tweaks
#20042 opened Feb 20, 2025 by bjacob Loading…
[Codegen][Tuner] add attention op into default tuning spec
#20039 opened Feb 20, 2025 by bangtianliu Loading…
7
Clarify error message for missing module dependencies. runtime/tools IREE's runtime tooling (iree-run-module, iree-benchmark-module, etc) runtime/vm IREE bytecode VM interpreter
#20036 opened Feb 19, 2025 by ScottTodd Loading…
1
Allow preferCloneToConsumer ops to ignore affinities.
#20033 opened Feb 19, 2025 by benvanik • Draft
Support cloning operand-less dispatches
#20025 opened Feb 18, 2025 by rsuderman Loading…
[DT] Add an option to not hoist infusible encoding ops out of dispatch.
#20017 opened Feb 18, 2025 by hanhanW Loading…
Allow expand shape propagation scross linalg ops with dynamic shapes.
#20014 opened Feb 18, 2025 by MaheshRavishankar • Draft
1
[Dispatch] Remove k2 constraint from BubbleUpExpandShapes
#20006 opened Feb 17, 2025 by IanWood1 Loading…
Let ConvertToROCDL use MathToROCDL, not MathToLLVM
#19994 opened Feb 14, 2025 by bjacob • Draft
1
[LLVMGPUVectorDistribute] Add general support for statically tiled codegen on dynamic shapes
#19992 opened Feb 14, 2025 by manupak • Draft
Typo Fix in hip-hal-driver.md
#19982 opened Feb 13, 2025 by mehmetcanfarsak Loading…
1
[codegen][gpu] Adding conv filter layout fhwc to preprocessing pipeline
#19974 opened Feb 12, 2025 by jerryyin Loading…
4
[AMDGPU] Do not rewrite or approximate math functions on ROCm
#19970 opened Feb 12, 2025 by bjacob • Draft
4
[Flow] Add dispatch annotation for horizontal contract fusion
#19966 opened Feb 11, 2025 by Max191 Loading…
Add mi300 runner for toy_llama tests
#19961 opened Feb 11, 2025 by stbaione Loading…
9
Revert "Fixing execution region result placement. (#19872)"
#19958 opened Feb 11, 2025 by stbaione Loading…
3
[LLVMGPU] Prefer column major MFMA to vectorize stores
#19919 opened Feb 5, 2025 by qedawkins Loading…
1
ProTip! What’s not been updated in a month: updated:<2025-01-21.