libclc: frexp: fix implementation regarding denormals #134823
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
Devices not supporting denormals can compare them true against zero.
It leads to result not matching the CTS expectation when either
supporting or not denormals.
For example for 0x1.008p-140 we get {0x1.008p-140, 0} while the CTS
expects {0x1.008p-1, -139} when supporting denormals, or {0, 0} when
not supporting denormals (flushed to zero).
Ref llvm#129871
|
@frasercrmck could you please review? |
frasercrmck
reviewed
Apr 8, 2025
| @@ -26,7 +26,7 @@ __clc_frexp(__CLC_GENTYPE x, __CLC_ADDRESS_SPACE __CLC_INTN *ep) { | |||
| (ai & (__CLC_INTN)MANTBITS_SP32); | |||
|
|
|||
| __CLC_INTN is_inf_nan_or_zero = | |||
| x == __CLC_FP_LIT(0.0) || __clc_isinf(x) || __clc_isnan(x); | |||
| ai == (__CLC_INTN)0 || __clc_isinf(x) || __clc_isnan(x); | |||
There was a problem hiding this comment.
I know this was in the original code that we're essentially reverting to, but do we know this is correct for -0.0?
There was a problem hiding this comment.
Yes, this is correct for -0.0, ai is the bitcast of the absolute value of x.
There was a problem hiding this comment.
This is passing the CTS on all platforms I could put my hand on.
There was a problem hiding this comment.
Yes, this is correct for
-0.0,aiis the bitcast of the absolute value ofx.
Oh I'm silly, yes of course. Not sure how I thought the opposite was going on!
frasercrmck
approved these changes
Apr 8, 2025
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Outdated suggestions cannot be applied.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.
Devices not supporting denormals can compare them true against zero. It leads to result not matching the CTS expectation when either supporting or not denormals.
For example for 0x1.008p-140 we get {0x1.008p-140, 0} while the CTS expects {0x1.008p-1, -139} when supporting denormals, or {0, 0} when not supporting denormals (flushed to zero).
Ref #129871