https://github.com/wenju-he approved this pull request.
LGTM. I reproduce the issue on intel gpu:
```
ldexp...
1: ldexp fp32
ERROR: ldexp: 0.500000 ulp error at {-0x1.fffffcp-127 (0x807fffff), -1}:
*-0x1p-127 (0x80400000) vs. -0x1.fffff8p-128 (0x803fffff) at index: 1116
2: ldexp fp64
ERROR: ldexp: inf ulp error at {-nan, 1}: *-nan vs. -inf
ldexp FAILED
```
fp32 passes after this PR.
The fp64 fail is an existing issue that is not related to this PR. So I can't
verify the fp64 fix.
https://github.com/llvm/llvm-project/pull/233791
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
