Integrate llvm/llvm-project@27ac46e6bea2 #17662
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
antiagainst
requested review from
MaheshRavishankar,
rsuderman,
hanhanW,
stellaraccident,
qedawkins,
kuhar,
Groverkss and
benvanik
as code owners
antiagainst
removed request for
benvanik,
stellaraccident,
kuhar,
hanhanW and
qedawkins
Abbreviated Benchmark Summary@ commit b918d15fd1fa968cb5c401910f67edd7cd702c58 (vs. base f4279657ef8da12d07f068a37cbd93986edb47d8) Data-Tiling Comparison TableClick to show
Regressed Latencies 🚩
[Top 3 out of 4 results showed] Improved Latencies 🎉
[Top 3 out of 21 results showed] Improved Total Dispatch Sizes 🎉
[Top 3 out of 6 results showed] Regressed Stream IR Dispatch Count (# of cmd.dispatch ops) 🚩
[Top 3 out of 10 results showed] Improved Stream IR Dispatch Count (# of cmd.dispatch ops) 🎉
[Top 3 out of 6 results showed] For more information: |
kuhar
approved these changes
Jun 13, 2024
MaheshRavishankar
requested changes
Jun 13, 2024
|
This seems like a fairly likely candidate for the source of dispatch count changes: iree-org/llvm-project@7ef83f5 Especially because the changes are being observed in data tiling enabled benchmarks. cc @Max191 |
@MaheshRavishankar are you blocking the integrate for this or would you look at it in a follow up since Quinn has explained the possible reason for the difference? |
Could you try reverting that locally to see if that is the issue. Then we can decide what to do next |
@MaheshRavishankar PTAL at the benchmark comment now, the bot has edited it and it seems the dispatch number regression is gone with the revert. |
LLITCHEV
pushed a commit
to LLITCHEV/iree
that referenced
this pull request
Jul 30, 2024
Updated to llvm/llvm-project@27ac46e6bea2 * Used LLVM `MathExtras.h` to replace MLIR one * Updated `applySignatureConversion` usage Updated to openxla/stablehlo@dd48ec5 * `chlo.minimum_broadcast_shapes` op was removed openxla/stablehlo#2287 * `chlo.dynamic_reshape` op was removed openxla/stablehlo#2286 * Added batching dims to scatter dims openxla/stablehlo#2259 Updated to llvm/torch-mlir@77d7f64 --------- Co-authored-by: hanhanW <hanhan0912@gmail.com> Co-authored-by: Rob Suderman <rob.suderman@gmail.com> Co-authored-by: Quinn Dawkins <quinn@nod-labs.com> Co-authored-by: Nirvedh Meshram <nirvedh@gmail.com> Signed-off-by: Lubo Litchev <lubol@google.com>
ScottTodd
reviewed
Aug 19, 2024
| @@ -464,6 +464,7 @@ def find_git_submodule_revision(submodule_path): | |||
| install_requires=[ | |||
| "numpy", | |||
| "PyYAML", | |||
| "sympy", | |||
There was a problem hiding this comment.
Looks like this new dep is for the fx_importer in torch-mlir. That should have at least been documented in the PR description.
This should probably go in extras_require for a iree-compiler[torch] package, not the general requirements here. The dep seems relatively small, but I want to keep the core project's dependencies minimal.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Updated to llvm/llvm-project@27ac46e6bea2
MathExtras.hto replace MLIR oneapplySignatureConversionusageUpdated to openxla/stablehlo@dd48ec5
chlo.minimum_broadcast_shapesop was removedDelete chlo.minimum_broadcast_shapes op openxla/stablehlo#2287
chlo.dynamic_reshapeop was removedRemove unused chlo.dynamic_reshape and downstream ops openxla/stablehlo#2286
[stablehlo] Add batching dims to
stablehlo.gatherandstablehlo.scatteropenxla/stablehlo#2259Updated to llvm/torch-mlir@77d7f64