Skip to content
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

[LLVMGPU] Disable scf.forall distribution for matmulSimt #19302

Merged
merged 1 commit into from
Nov 26, 2024

Conversation

pashu123
Copy link
Contributor

We are moving towards tile&fuse pipeline. matmulSimt pipeline is to be deprecated. Meanwhile, all existing tests that depend on matmulsimt pipeline shouldn't break and hence disabling the forall distribution for the pipeline.

We are moving towards tile&fuse pipeline. `matmulSimt` pipeline is to be
deprecated. Meanwhile, all existing tests that depend on matmulsimt
pipeline shouldn't break and hence disabling the forall distribution for
the pipeline.
@pashu123 pashu123 merged commit a4c6f35 into iree-org:main Nov 26, 2024
37 checks passed
@kuhar
Copy link
Member

kuhar commented Nov 26, 2024

@pashu123 seems like there may be issues with SDXL perf after merging this PR

@kuhar
Copy link
Member

kuhar commented Nov 26, 2024

cc: @qedawkins for visibility :^

@qedawkins
Copy link
Contributor

That's surprising because this is just reverting to the behavior we've had for a long time, are we sure it's this PR?

@kuhar
Copy link
Member

kuhar commented Nov 26, 2024

It started failing just after merging this one, but I can't say for sure because I haven't tried reproducing this locally

@pashu123
Copy link
Contributor Author

@kuhar It might be because of 2602a2a (this is just reverting to the old behaviour) that @qedawkins mentioned.

@qedawkins
Copy link
Contributor

qedawkins commented Nov 26, 2024

I see a similar huge regression right before this PR too: https://github.com/iree-org/iree/actions/runs/12023092109/job/33516722915

maybe an unstable runner?

@kuhar
Copy link
Member

kuhar commented Nov 26, 2024

cc: @saienduri

@pashu123
Copy link
Contributor Author

It started here: 77ff99c

@kuhar
Copy link
Member

kuhar commented Nov 26, 2024

@pashu123 seems like it recovered on its own here: #19122, unstable runner most likely then

Groverkss pushed a commit to Groverkss/iree that referenced this pull request Dec 1, 2024
)

We are moving towards tile&fuse pipeline. `matmulSimt` pipeline is to be
deprecated. Meanwhile, all existing tests that depend on matmulsimt
pipeline shouldn't break and hence disabling the forall distribution for
the pipeline.
giacs-epic pushed a commit to giacs-epic/iree that referenced this pull request Dec 4, 2024
)

We are moving towards tile&fuse pipeline. `matmulSimt` pipeline is to be
deprecated. Meanwhile, all existing tests that depend on matmulsimt
pipeline shouldn't break and hence disabling the forall distribution for
the pipeline.

Signed-off-by: Giacomo Serafini <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants