On 07.01.26 14:10, Jonathan Wakely wrote: > On Wed, 7 Jan 2026 at 12:58, Frank Scheiner wrote: >> >> === libstdc++ Summary === >> >> # of expected passes 18162 19243 >> # of unexpected failures 9 245 >> # of expected failures 131 140 >> # of unresolved testcases 2 >> # of unsupported tests 740 786 >> ``` >> >> Overall this looks like an improvement to me. Although the number of >> unexpected failures for libstdc++ increased dramatically. But the number >> of expected passes - where they increased - increased more than the >> number of failures. > > Is there a pattern to the unexpected failures? Are they all in a > particular part of the library? Are they compilation failures? ICEs? > failed assertions in the tests at runtime?
No, seems to be spread all over the place. While for GCC 15.1.0 we only had fails for "22_locale/time_get/get" and "std/format/ranges", for gcc-16-20251214 we now have additional fails in: * 18_support * 19_diagnostics * 20_util * 21_strings * 23_containers * 25_algorithms * 26_numerics * 27_io * 28_regex * 30_threads * experimental/any/assign/exception.cc * ext * std/format/functions * std/format/ranges (here some extra fails compared to GCC 15.1.0) * std/format * std/time * tr1 For details, see attached diff.gz. On a side note, could [1] be related? [1]: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121623#c33 Cheers, Frank
gcc-15.1.0-gcc-16-20251214-libstdcplusplus-fails.diff.gz
Description: application/gzip
