[Codegen][Common] Add a pass to linearize memrefs #19332
+639
−0
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
Abhishek-Varma
force-pushed
the
avarma_linearize_memrefs_pass
branch
2 times, most recently
from
0fb0475 to
85bdf08
Compare
qedawkins
requested changes
Dec 2, 2024
MaheshRavishankar
requested changes
Dec 3, 2024
|
Hi @MaheshRavishankar @qedawkins - I've addressed the comments. This is ready for re-reviewing. Thanks! |
Abhishek-Varma
force-pushed
the
avarma_linearize_memrefs_pass
branch
2 times, most recently
from
c9dc955 to
20747ec
Compare
|
Hi @krzysz00 @qedawkins @MaheshRavishankar - I've addressed all comments and the CI seems to be happy now - you may re-review/approve. Thanks! |
qedawkins
reviewed
Jan 28, 2025
| } | ||
|
|
||
| static FailureOr<SmallVector<OpFoldResult>> | ||
| getMixedOrigSize(Location loc, PatternRewriter &rewriter, Value sourceValue) { |
There was a problem hiding this comment.
This method could be simpler if you use memref::getMixedSizes. Something like
MemRefType sourceType = llvm::cast<MemRefType>(sourceValue.getType());
Operation *sourceOp = sourceValue.getDefiningOp();
if (!isa<memref::AllocOp, memref::AllocaOp>(sourceOp) && !sourceType.hasStaticShape()) {
return failure();
}
return memref::getMixedSizes(rewriter, sourceValue);
There was a problem hiding this comment.
Similar reasoning as previous thread.
There was a problem hiding this comment.
Quinn is right. We should be able to just use getMixedSizes. I think you dont even need the if condition that he has there.
There was a problem hiding this comment.
Correct. I've responded here.
|
To link what I think we might want to do, #19851 |
Abhishek-Varma
force-pushed
the
avarma_linearize_memrefs_pass
branch
from
eb1957f to
a37f8a4
Compare
-- This commit creates a pass to linearize memrefs. -- The pass `iree-linearize-memrefs` will be iteratively worked upon to make it an inter-procedural pass. -- Currently it supports limited operations. Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Abhishek-Varma
force-pushed
the
avarma_linearize_memrefs_pass
branch
from
a37f8a4 to
67a3439
Compare
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.
-- This commit creates a pass to linearize memrefs.
-- The pass
iree-linearize-memrefswill be iteratively worked upon to make it an inter-procedural pass.-- Currently it supports limited operations.
Signed-off-by: Abhishek Varma avarma094@gmail.com