math_brute_force: fix fdim to use device's rounding when converting result back to half.
#2223
Conversation
This file contains 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
… result back to half.
svenvh
reviewed
Jan 10, 2025
| @@ -312,9 +317,9 @@ cl_int TestHalf(cl_uint job_id, cl_uint thread_id, void *data) | |||
| s2[j] = cl_half_to_float(p2[j]); | |||
| if (isNextafter) | |||
| r[j] = cl_half_from_float(reference_nextafterh(s[j], s2[j]), | |||
There was a problem hiding this comment.
The changes look reasonable to me.
However this affects all binary fp16 operations as far as I can tell, not only fdim? It would be good to make that clear in the commit message and description; right now it suggests the change would only affect fdim which is not the case.
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Outdated suggestions cannot be applied.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.
In the half-precision
fdimtest, the original code usedCL_HALF_RTEto convert the float result back to half, causing a mismatch in computation results when the hardware uses RTZ. Some of the examples:Fixed this by using the hardware's default rounding mode when converting the result back to half.