[SYCL] Implement sycl_ext_oneapi_address_cast #12382
Merged
Conversation
This file contains hidden or 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
Contributor
Pennycook
commented
Jan 12, 2024
Pennycook
commented
- Adds support for OpGenericCastToPtr SPIR-V intrinsic.
- Implements static_address_cast using OpGenericCastToPtr.
- Implements dynamic_address_cast using OpGenericCastToPtrExplicit.
- Adds tests for both new forms of address_cast.
- Adds support for OpGenericCastToPtr SPIR-V intrinsic. - Implements static_address_cast using OpGenericCastToPtr. - Implements dynamic_address_cast using OpGenericCastToPtrExplicit. - Adds tests for both new forms of address_cast. Signed-off-by: John Pennycook <[email protected]>
Pennycook
changed the title
Pennycook
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
Neither form of OpGenericCastToPtr is implemented for AMD. Signed-off-by: John Pennycook <[email protected]>
Pennycook
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
Neither form of OpGenericCastToPtr is implemented for NVIDIA. Signed-off-by: John Pennycook <[email protected]>
Pennycook
force-pushed
the
address_cast
branch
from
c145219 to
152e355
Compare
Pennycook
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
Pennycook
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
Naghasan
approved these changes
Jan 15, 2024
| template <typename dataT> | ||
| extern __attribute__((opencl_global)) dataT * | ||
| __SYCL_GenericCastToPtr_ToGlobal(void *Ptr) noexcept { | ||
| return (__attribute__((opencl_global)) dataT *) |
There was a problem hiding this comment.
minor: no need for cast, the returned type should be already correctly typed
There was a problem hiding this comment.
Good point! I just copied the existing definitions of __SYCL_GenericCastToPtrExplicit here.
I agree it makes sense to change this, but I'd prefer for it to be done as part of a follow-up PR. I don't really want to touch the existing definitions, in case something breaks unexpectedly.
Naghasan
reviewed
Jan 15, 2024
Comment on lines
+53
to
+54
| Success &= LocalPointer.get_raw() == nullptr; | ||
| Success &= PrivatePointer.get_raw() == nullptr; |
There was a problem hiding this comment.
@mmoadeli for your awareness, your patch is going to be required to enable this test on cuda and hip
sergey-semenov
approved these changes
Jan 16, 2024
smanna12
approved these changes
Jan 16, 2024
|
@intel/llvm-gatekeepers, I think this can be merged now. |
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.