afonso360 opened PR #4373 from interp-fminmax
to main
:
:wave: Hey,
This PR fixes a bug with the interpreter where it is returning the wrong value when comparing
0.0
and-0.0
infmin
andfmax
.This happens because
fmin
/fmax
are defined as returning -0.0 as smaller than 0.0.However this is not how these values are specified in IEEE754 and the interpreter was
returning the wrong value in these operations since it was just using the standard IEEE754 comparisons.This also tries to preserve NaN information by avoiding passing NaN's
through any operation that could canonicalize it.
afonso360 updated PR #4373 from interp-fminmax
to main
.
cfallin submitted PR review.
afonso360 updated PR #4373 from interp-fminmax
to main
.
afonso360 updated PR #4373 from interp-fminmax
to main
.
afonso360 updated PR #4373 from interp-fminmax
to main
.
cfallin merged PR #4373.
Last updated: Nov 22 2024 at 17:03 UTC