libclc: increase fp16 support #98149
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
|
Those implementations have been tested using 1 Intel and 1 AMD devices (with clvk/clspv) |
|
@alan-baker @kpet , could you please review? |
|
✅ With the latest revision this PR passed the Python code formatter. |
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
rjodinchr
force-pushed
the
pr/libclc-fp16
branch
2 times, most recently
from
c03c6e3 to
a387a97
Compare
alan-baker
approved these changes
Jul 9, 2024
|
@frasercrmck Could you review it please? |
Sure, will do, thanks for this. We've done something similar downstream so I can hopefully verify against what we've done. |
Increase fp16 support to allow clspv to continue to be OpenCL compliant following the update of the OpenCL-CTS adding more testing on math functions and conversions with half. Math functions are implemented by upscaling to fp32 and using the fp32 implementation. It garantees the accuracy required for half-precision float-point by the CTS.
rjodinchr
force-pushed
the
pr/libclc-fp16
branch
from
a387a97 to
449321d
Compare
|
Also, just leave that thought here: A next step might be to use |
|
@frasercrmck could you submit that PR? I cannot do it. |
Yeah, it would definitely need to be opt-in. Ideally we'd also implement the |
|
@rjodinchr I was about to merge and noticed your email is hidden. I just wanted to check with you whether you want to make it public, as per the discussion https://discourse.llvm.org/t/hidden-emails-on-github-should-we-do-something-about-it/74223/67. AFAICT there's still no official policy, but thought I'd check. |
I was not aware my email was private. It should be public now. Thanks |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/153/builds/3385 Here is the relevant piece of the build log for the reference: |
|
I am able to reproduce it, working on it right now |
|
This PR should fix the build issue: #99481 |
|
This change seems to have caused test regression. This is Gentoo Linux amd64, tried LLVM 72d8c27 and 18.1.8. To reproduce, I've done: The result: LastTest.log: LastTest.log If I go back to 6838c7a, they all pass. |
yuxuanchen1997
pushed a commit
that referenced
this pull request
Jul 25, 2024
Summary: Increase fp16 support to allow clspv to continue to be OpenCL compliant following the update of the OpenCL-CTS adding more testing on math functions and conversions with half. Math functions are implemented by upscaling to fp32 and using the fp32 implementation. It garantees the accuracy required for half-precision float-point by the CTS. Test Plan: Reviewers: Subscribers: Tasks: Tags: Differential Revision: https://phabricator.intern.facebook.com/D60251019
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.
Increase fp16 support to allow clspv to continue to be OpenCL compliant following the update of the OpenCL-CTS adding more testing on math functions and conversions with half.
Math functions are implemented by upscaling to fp32 and using the fp32 implementation. It garantees the accuracy required for half-precision float-point by the CTS.