mirror of
https://github.com/gentoo-mirror/gentoo.git
synced 2026-11-20 22:27:17 -08:00
See https://bugs.gentoo.org/952059#c4, quoting that: > The story here is that GCC 15 will support [[clang::musttail]], but it > turns out that [[clang::musttail]] in Clang (and as it's a Clang-namespaced > extension, they get to decide the semantics) takes a "UB on invalid input" > approach rather than diagnosing. > > The plan is to make [[clang::musttail]] relaxed (matching Clang) and to > have [[gnu::musttail]] with diagnostics. > > This shows up now in qtwebengine because __has_attribute sees musttail is > available, so it automagically uses it (which is fine), just a GCC bug > then gets in the way. > > I'm probably going to backport a patch to just temporarily disable musttail > again until next week, as various fixes are landing. I've also backported a fix to fix an e2fsprogs/sqlite miscompilation on x86. Closes: https://bugs.gentoo.org/951997 Closes: https://bugs.gentoo.org/951999 Closes: https://bugs.gentoo.org/952059 Closes: https://bugs.gentoo.org/952085 Signed-off-by: Sam James <sam@gentoo.org>