Hi Emilio, On 2025-12-03 13:38, Emilio Pozuelo Monfort wrote: > Control: tags -1 confirmed > > Hi Aurelien, > > On 02/12/2025 06:44, Aurelien Jarno wrote: > > Package: release.debian.org > > Severity: normal > > X-Debbugs-Cc: [email protected], [email protected] > > User: [email protected] > > Usertags: transition > > Control: affects -1 + src:glibc > > > > Dear release team, > > > > I would like to request a transition slot for glibc 2.42. It has been > > available since the beginning of August and appears to be in good shape. > > It has built successfully on all release architectures and on most ports > > architectures. > > > > One important change in this version is the removal of the obsolete > > termio interface in favor of the termios one. This causes a few packages > > to FTBFS. The addition of ISO C23 and ISO CY math functions also creates > > a few conflicts with names used in some source code, resulting in a few > > FTBFS. There are also a couple of additional FTBFS due to some other > > small changes. An archive rebuild was done (although 2 months ago) and > > all of the FTBFS found on amd64 have been reported [1], many have > > already been fixed. > > > > On the autopkgtest side, the experimental pseudo-excuses look good > > overall, most of the issues looks like false positive. The only real > > ones that I have identified are fuse-zip (due to valgrind), openmsx and > > zvbi (amd64 only). They also have been reported [1]. > > > > As glibc is using symbol versioning, there is no soname change. That > > said a few packages use internal libc symbols and have to be > > rebuilt for this transition. Here is the corresponding ben file: > > > > title = "glibc"; > > is_affected = .depends ~ /libc[0-9.]* \(<</; > > is_good = .depends ~ /libc[0-9.]* \(<< 2.43\)/; > > is_bad = .depends ~ /libc[0-9.]* \(<< 2.42\)/; > > > > In addition there have been some termios interface changes, but > > compatibility for older binaries is preserved through symbol versioning. > > Unfortunately this doesn't prevent the changes of the gnat-XXX Provides > > from the gnat-$GCVER-$DEB_TARGET_MULTIARCH packages once they are > > rebuilt [2]. This means that gcc-13 to gcc-15 also need to get rebuilt, > > and then all gnat packages depending on the old gnat-14 provides. Not > > sure how you want to deal with that, but I have the impression that > > while it has to be done closely after the glibc 2.42 upload, the two > > transitions are not completely coupled, and glibc 2.42 can migrate to > > testing before gnat-14. > > > > Speaking about migration, this glibc version provides a GLIBC_2.42 > > version of the existing termios symbols on a few architectures that have > > a testing suite. It also adds GLIBC_2.42 version for the __inet_ntop_chk > > and __inet_pton_chk symbols, that are used with _FORTIFY_SOURCE=2. > > Additionally there are new symbols for new ISO C23 and ISO CY math > > functions on all architectures and for vectorized math functions on > > arm64. Therefore there is probably more risks than for previous glibc > > transitions that one of these news symbols will be picked up by some > > packages, preventing them to migrate to testing before glibc 2.42. > > > > Thanks for considering. > > Please go ahead.
Thanks! I have just uploaded it. Cheers Aurelien -- Aurelien Jarno GPG: 4096R/1DDD8C9B [email protected] http://aurel32.net
