[Codegen][GPU] Improve forall hoisting pattern for single trip loops #18418
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
Max191
approved these changes
Sep 5, 2024
Max191
reviewed
Sep 5, 2024
compiler/src/iree/compiler/Codegen/Dialect/GPU/Transforms/FuseAndHoistParallelLoops.cpp
Show resolved
Hide resolved
For single trip scf.forall loops the `tensor.extract_slice` on the output can be folded away, causing the forall loop hoisting pattern to fail. Single trip loops with processor ID mappings cannot be folded away because they can resolve to an `scf.if`. So this patch extends the loop hoisting pattern to support hoisting in the case of single trip loops where the `tensor.extract_slice` has been folded away.
qedawkins
force-pushed
the
single_trip_hoisting
branch
from
c52d852 to
7611c94
Compare
josemonsalve2
pushed a commit
to josemonsalve2/iree
that referenced
this pull request
Sep 14, 2024
…ree-org#18418) For single trip scf.forall loops the `tensor.extract_slice` on the output can be folded away, causing the forall loop hoisting pattern to fail. Single trip loops with processor ID mappings cannot be folded away because they can resolve to an `scf.if`. So this patch extends the loop hoisting pattern to support hoisting in the case of single trip loops where the `tensor.extract_slice` has been folded away.
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Outdated suggestions cannot be applied.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.
For single trip scf.forall loops the
tensor.extract_sliceon the output can be folded away, causing the forall loop hoisting pattern to fail. Single trip loops with processor ID mappings cannot be folded away because they can resolve to anscf.if. So this patch extends the loop hoisting pattern to support hoisting in the case of single trip loops where thetensor.extract_slicehas been folded away.