Summary: | dev-cpp/rapidfuzz-cpp-1.10.4: failed (test phase) | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Joonas Niilola <juippis> |
Component: | Current packages | Assignee: | Michał Górny <mgorny> |
Status: | CONFIRMED --- | ||
Severity: | normal | ||
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
See Also: | https://github.com/maxbachmann/rapidfuzz-cpp/issues/100 | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
rapidfuzz-cpp-1.10.4:20230108-074550.log
LastTest.log |
Description
Joonas Niilola
2023-01-08 08:13:24 UTC
Created attachment 847884 [details]
rapidfuzz-cpp-1.10.4:20230108-074550.log
Created attachment 847886 [details]
LastTest.log
At a first glance, it looks like a float precision problem. Could you experiment a bit with CFLAGS to see if you can get them to pass, to confirm? -mfpmath=sse may be relevant. FAIL: -march=native -O2 -pipe -frecord-gcc-switches FAIL: -march=i686 -mfpmath=sse -O2 -pipe -frecord-gcc-switches SUCCESS: -march=native -mfpmath=sse -O2 -pipe -frecord-gcc-switches |