forked from llvm/llvm-project
-
Notifications
You must be signed in to change notification settings - Fork 3
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Tosa Changes Integration #524
Draft
jorickert
wants to merge
40
commits into
bump_to_e2402615
Choose a base branch
from
jrickert.bump_integration
base: bump_to_e2402615
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Draft
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
jorickert
commented
Mar 20, 2025
•
edited
Loading
edited
- Bump LLVM to e240261
- Test MLIR
- Sync Torch-MLIR with e240261 [AutoBump] Merge with fixes of 12250739 (Jan 28, needs LLVM bump) (161) torch-mlir#555
- Update and Test Torch-MLIR
- Sync ONNX-MLIR with e240261 Bump to e240261 onnx-mlir#328
- Update and Test ONNX-MLIR
- Test xten_nn
- Test downstream
- investigate mlir warning/error: : Caused by [mlir] Add support for MLIR_LINK_MLIR_DYLIB llvm/llvm-project#119408 -> Partially reverted for now
- Final update of all branches
- Final test of all components
…monotonic_slice_params"" This reverts commit 018230a.
`computeStaticLoopSizes()` is functionally identical to `getStaticLoopRanges()`. Replace all uses of `computeStaticLoopSizes()` by `getStaticLoopRanges()` and remove the former.
[AutoBump] Merge with fixes of 3494ee9 (Nov 01) (Needs downstream fixes) (8)
Fix problem where the shape of the insert shape was calculated incorrectly
… are executed before constraints are called, but this is not always the case, as operations can be reordered in PDLInterp if there is no dependency between them. For example: Pdll pattern: ``` let someOp = op<someDialect.SomeOp>(input: Value<inputTy: Type) {axis = inputAxis: Attr } -> (resTypes: TypeRange); let someResult = someConstraint(inputAxis); ``` If SomeOp requires axis to have a valid value, it is easy to (wrongly) assume that someConstraint always gets called with a not-null inputAxis. This is not correct. The linearized PDLInterp (pseudo-code) could be the following: ``` %0 = pdl_interp.get_attribute "axis" of %arg0 %1 = pdl_interp.apply_constraint “someConstraint”(%0) pdl_interp.is_not_null(%0) pdl_interp.check_operation_name of %arg0 is "someDialect.SomeOp" ``` Note that here someConstraint can be called with a null attribute. This commit changes the prioritization of predicates, so that constraints are run after other predicates. ``` %0 = pdl_interp.get_attribute "axis" of %arg0 pdl_interp.is_not_null(%0) pdl_interp.check_operation_name of %arg0 is "someDialect.SomeOp" %1 = pdl_interp.apply_constraint “someConstraint”(%0) ``` This ensures that null or operation name checks are run before constraints. This is closer to the mental model when writing PDLL patterns and should make it less likely to run into bugs caused by assuming implicit checks for not null.
pdl_interp: Sort constraints to end of predicate list.
Do not print unnecessary newlines if attributes are elided
[FXML-5704] Compute affine expression bounds
Ensure 0 <= x mod N < N semantics
Add tosa.cast folding for unsigned integers
[AutoBump] Merge with ceeb08b (Nov 18) (9)
[AutoBump] Merge with 258a5d4 (Nov 20) (11)
[AutoBump] Merge with fixes of 68a3908 (Nov 18) (10)
[AutoBump] Merge with fixes of beff2ba (Nov 22) (14)
[TOSA] Extend folding/canonicalization for concat, tile and slice.
[AutoBump] Merge with fixes of cbc7802 (Nov 20) (12) Needs downstream changes
[AutoBump] Merge with 56eb559 (Nov 26) (15)
[AutoBump] Merge with 776476c (Nov 22) (13)
….bump_integration Required various changes to tosa.tile canonicalization/folding
1 task
…tered more than once
…535ea8eb18 to LLVM
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.