Skip to content

[GEN] Change existing OCL lowering to the new interface#1224

Merged
whitneywhtsang merged 2 commits into
llvm-targetfrom
whitneywhtsang/2dblockread
Jun 10, 2024
Merged

[GEN] Change existing OCL lowering to the new interface#1224
whitneywhtsang merged 2 commits into
llvm-targetfrom
whitneywhtsang/2dblockread

Conversation

@whitneywhtsang
Copy link
Copy Markdown
Contributor

@whitneywhtsang whitneywhtsang commented Jun 1, 2024

Please do not squash and merge this PR.

@whitneywhtsang whitneywhtsang self-assigned this Jun 1, 2024
@whitneywhtsang whitneywhtsang force-pushed the whitneywhtsang/2dblockread branch from db8db94 to c24267e Compare June 1, 2024 19:23
@whitneywhtsang
Copy link
Copy Markdown
Contributor Author

Awaiting for a release driver that contains the support of the OCL builtin, but ready for review.

@whitneywhtsang whitneywhtsang marked this pull request as ready for review June 2, 2024 21:29
Copy link
Copy Markdown

@victor-eds victor-eds left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Are CI failures related to these changes?

Comment thread third_party/intel/lib/TritonGENToLLVM/TritonGENToLLVMPass.cpp
Comment thread test/TritonGEN/tritongen-to-llvm.mlir Outdated
Comment thread third_party/intel/lib/TritonGENToLLVM/TritonGENToLLVMPass.cpp
@whitneywhtsang
Copy link
Copy Markdown
Contributor Author

Are CI failures related to these changes?

Yes, will pass when we update to the next agama rolling driver.

Copy link
Copy Markdown
Contributor

@FMarno FMarno left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code seems fine to me

Comment thread test/TritonGEN/tritongen-to-llvm.mlir Outdated
Comment thread test/TritonGEN/tritongen-to-llvm.mlir Outdated
whitneywhtsang added a commit that referenced this pull request Jun 10, 2024
Signed-off-by: Whitney Tsang <whitney.tsang@intel.com>
@whitneywhtsang whitneywhtsang force-pushed the whitneywhtsang/2dblockread branch from 761892b to c99dc29 Compare June 10, 2024 17:16
@whitneywhtsang whitneywhtsang marked this pull request as draft June 10, 2024 17:31
Signed-off-by: Whitney Tsang <whitney.tsang@intel.com>
Signed-off-by: Whitney Tsang <whitney.tsang@intel.com>
@whitneywhtsang whitneywhtsang force-pushed the whitneywhtsang/2dblockread branch from c99dc29 to 75411ea Compare June 10, 2024 22:08
@whitneywhtsang whitneywhtsang marked this pull request as ready for review June 10, 2024 22:39
@whitneywhtsang whitneywhtsang merged commit c141986 into llvm-target Jun 10, 2024
@whitneywhtsang whitneywhtsang deleted the whitneywhtsang/2dblockread branch June 10, 2024 22:39
whitneywhtsang pushed a commit that referenced this pull request Jun 17, 2024
…1247)

#1224 converts `triton_gen.2Dblockload` to OCL builtin `2d_block_read`
calls. Add cache controls decoration capturing original `cache_control`
semantics.

---------

Signed-off-by: Victor Perez <victor.perez@codeplay.com>
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.

[GEN] Use OCL builtin for 2D block read available on driver 881.12

5 participants