[GPU] Use affine.linearize_index (and delinearize_index) where possible #19087
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
krzysz00
force-pushed
the
gpu-distribute-with-linearize
branch
2 times, most recently
from
819a5ae to
c67c36d
Compare
There have been issues with the composition of affine maps being too general and loosing important information, like the fact that affine_map<(s0 + s1 * 32 + ... - (s0 floorDiv 16) * 16)> realy should be affine_map<(s0 mod 16 + s1 * 32 + ...)>, and other issues with the ultimate IR that block low-level arithmetic optimizations. The affine.delinearize_index operation represents the div/mod chains needed to break a flat index into its component parts. A recently added affine.linearize_index operation is its inverse - combining multiple indices into a flat 1D value. Another advantage to linearize/delinearize is simpler upstream canonicalizations and lead to more streamlined generated code. This PR updates the vector distribution code and other GPU-related code that I could find to 1. Use affine.linearize_index to construct flat thread IDs 2. Use affine.delinearize_index in places where there was a floorDiv/mod chain.
krzysz00
force-pushed
the
gpu-distribute-with-linearize
branch
from
c67c36d to
8192c87
Compare
krzysz00
requested review from
MaheshRavishankar,
qedawkins,
kuhar,
Groverkss,
antiagainst and
hanhanW
as code owners
| @@ -143,6 +153,22 @@ LogicalResult resolveGPUMappedForallOp(RewriterBase &rewriter, | |||
| newBlockArgs); | |||
| rewriter.eraseOp(forallTerminator); | |||
| rewriter.eraseOp(forallOp); | |||
|
|
|||
| // Step 5. Create the post-loop code that only executes on some workitems. | |||
| if (hasPostLoopTail) { | |||
There was a problem hiding this comment.
Could we move this into a separate PR?
|
Closing in favor of #19122 so I can split the PR |
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.
There have been issues with the composition of affine maps being too general and loosing important information, like the fact that affine_map<(s0 + s1 * 32 + ... - (s0 floorDiv 16) * 16)> really should be affine_map<(s0 mod 16 + s1 * 32 + ...)>, and other issues with the ultimate IR that block low-level arithmetic optimizations.
The affine.delinearize_index operation represents the div/mod chains needed to break a flat index into its component parts. A recently added affine.linearize_index operation is its inverse - combining multiple indices into a flat 1D value.
Another advantage to linearize/delinearize is simpler upstream canonicalizations and lead to more streamlined generated code.
This PR updates the vector distribution code and other GPU-related code that I could find to
Additionally
scf.forwith a thread ID as initial input approach to non-uniform execution for anscf.ifto better reflect the intended control flow and enable a linearize