Skip to content

Pull requests: iree-org/iree

New pull request New
157 Open 17,409 Closed
157 Open 17,409 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

[Codegen][ROCm] Add WMMA intrinsics for gfx1250
#22516 opened Nov 1, 2025 by kuhar Loading…
1
[CPU][NFC] Trim IRs for lowering_config tests. (2/N)
#22512 opened Nov 1, 2025 by hanhanW Loading…
1
[WIP] Unrevert transfer_write dedup (0c1a5b4) and bump LLVM
#22509 opened Oct 31, 2025 by newling • Draft
Integrate torch-mlir at llvm/torch-mlir@288cd5e8adb
#22508 opened Oct 31, 2025 by IanWood1 Loading…
1
[LinalgExt] Decompose map_scatter with strided rank-reducing subviews
#22504 opened Oct 31, 2025 by Max191 Loading…
1
1
[Dispatch Creation] Drop unit dims from tensor.extract ops
#22503 opened Oct 31, 2025 by IanWood1 Loading…
[Stream] Fixing update order and improving the cache for ReplicateGlobalsPerAffinity pass.
#22499 opened Oct 31, 2025 by hanhanW Loading…
1
Medium ping pong scheduling improvement
#22496 opened Oct 31, 2025 by sebvince • Draft
[DispatchCreation] Enable splitting multiple reduction dimensions for weight backward convs
#22491 opened Oct 30, 2025 by yzhang93 Loading…
[Codegen] add type propagation for linalg.reduce
#22490 opened Oct 30, 2025 by bangtianliu Loading…
6
[Codegen][GPU] Allow intentional padding for non-K-major matmul layouts
#22486 opened Oct 30, 2025 by jerryyin Loading…
6
ScheduleExecution enhancements for timeline-aware scheduling and SCF. compiler/dialects Relating to the IREE compiler dialects (flow, hal, vm)
#22483 opened Oct 30, 2025 by benvanik • Draft
Fix CPU matmul transposition to only apply when mmt4d ukernel enabled
#22482 opened Oct 30, 2025 by Copilot AI • Draft
1
@hanhanW
2
[ROCM] Update bounds for large f16 data-tiling ukernel
#22481 opened Oct 30, 2025 by jtuyls Loading…
1
[Do not merge] Experiment with reverting
#22480 opened Oct 30, 2025 by lialan • Draft
[ci] Add CI for MI355x machines
#22475 opened Oct 30, 2025 by Groverkss Loading…
6
[DispatchCreation] Set split reduction size for ArgCompare
#22466 opened Oct 29, 2025 by bangtianliu Loading…
[Dispatch Creation] Don't add unfusable consumers to fusion group
#22461 opened Oct 29, 2025 by IanWood1 Loading…
1
Remove value bounds interface for ExpandShapeOp
#22460 opened Oct 29, 2025 by jtuyls Loading…
1
[ci] Switch to new RDNA3 runners
#22458 opened Oct 29, 2025 by Groverkss Loading…
1
[GPU] Make GPUApplyPaddingLevel propagate padding to producers
#22445 opened Oct 28, 2025 by Groverkss • Draft
[DispatchCreation] Enable fusion of encoding ops with multi-use producers
#22444 opened Oct 28, 2025 by Abhishek-Varma Loading…
16
[WIP]Initial support for aten::flex_attention and rewrite to linalgext.online_attention
#22441 opened Oct 28, 2025 by keshavvinayak01 • Draft
[DO NOT SUBMIT] Fixes for LLVM PR https://github.com/llvm/llvm-project/pull/162924
#22439 opened Oct 28, 2025 by MaheshRavishankar • Draft
[ROCM] Enable dispatch specialization by default
#22425 opened Oct 27, 2025 by jtuyls Loading…
ProTip! Mix and match filters to narrow down what you’re looking for.