[SYCL] atomic_memory_order_capabilities query for device and context#8517
Merged
AlexeySachkov merged 28 commits intointel:syclfrom Mar 23, 2023
Merged
Conversation
added 3 commits
February 28, 2023 07:07
This satsifies llvm-test-suite/SYCL/AtomicRef/atomic-memory-order.cpp if you remove the "UNSUPPORTED OpenCL" test line, but doesn't implement the other capabilities this could return. Further discussion is required for further work to continue.
Minimal implementaton for now that calls similar query for each device and combines them. Currently checks for memory_order::relaxed, others coming soon.
AlexeySachkov
previously requested changes
Mar 2, 2023
added 4 commits
March 8, 2023 08:45
Covering all current OpenCL versions
Return full set of them as in OCL 2.x
Instead of resolving it as its own plugin call with PI code, just call the related one for each of its devices and intersect the results together.
added 2 commits
March 9, 2023 05:13
Unsure whether ACQUIRE and RELEASE need removing as current pi.h expects 5 bits for memory order info but the OCL headers say only 3 are necessary in 3.0 as the values are RELAXED, ACQ_REL, and SEQ_CST.
ACQ_REL infers ACQUIRE and RELEASE, so adding them to the return value.
Author
|
This will need to wait for #8595 to be merged first to resolve potential merge conflicts. |
added 2 commits
March 10, 2023 08:00
Moved mask location, added bitwise operations to convert CL to PI enum values, masked off relevant bits, and improved handling of out parameters.
andylshort
force-pushed
the
alamzeds/atomic_mem_order_caps_fix
branch
from
40a96e4 to
58c53ad
Compare
added 3 commits
March 14, 2023 08:04
…om/andylshort/llvm into alamzeds/atomic_mem_order_caps_fix
added 6 commits
March 16, 2023 05:07
steffenlarsen
approved these changes
Mar 21, 2023
Contributor
steffenlarsen
left a comment
steffenlarsen
left a comment
There was a problem hiding this comment.
Sorry for the delay. LGTM!
Author
|
I'm just testing a patch that addresses this comment from #8586 (comment) as there now are UR equivalents of the PI flags I needed, so just need to hold fire on the merge for a short while. |
AlexeySachkov
dismissed
their stale review
My comments were resolved, dismissing that review
veselypeta
pushed a commit
to veselypeta/llvm
that referenced
this pull request
Sep 21, 2023
…ntel#8517) This patch implements the `atomic_memory_order_capabilities` query in the OpenCL and Level Zero backends/plugins for `device` and `context` Specifically: - OpenCL <2.0 returns the minimum required capability set (`relaxed`) defined in [Section 4.2 of the OpenCL 3.0 specification](https://registry.khronos.org/OpenCL/specs/3.0-unified/html/OpenCL_API.html#CL_DEVICE_ATOMIC_MEMORY_CAPABILITIES). - OpenCL <3.0 and Level Zero backends return all memory order capabilities. - OpenCL >=3.0 queries the actual device to get the supported memory order capabilities. E2E test have also been updated to reflect these changes: intel/llvm-test-suite#1627
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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 patch implements the
atomic_memory_order_capabilitiesquery in the OpenCL and Level Zero backends/plugins fordeviceandcontextSpecifically:
relaxed) defined in Section 4.2 of the OpenCL 3.0 specification.E2E test have also been updated to reflect these changes: intel/llvm-test-suite#1627