Bug#442418: sigcontext.h broken
Package: libc6-dev
Version: 2.6.1-3
Severity: grave
Hi,
as discussed on IRC:
13:10 < _rene_> MadCoder: ping?
13:14 < MadCoder> _rene_: small pont
13:14 -!- ema (Emanuele Rocca) [~ema@fw.galliera.it] has joined #debian-devel
13:14 < MadCoder> pong
13:14 -!- tdykstra [~tdykstra@baardmijt.demon.nl] has quit [Ping timeout: 480 seconds]
13:16 -!- Col_Kernel (Joachim Reichel) [~joachim@p54BB044F.dip0.t-ipconnect.de] has joined #debian-devel
13:18 -!- pabs (Paul Wise) [~pabs@CPE-58-166-79-22.nsw.bigpond.net.au] has joined #debian-devel
13:23 < _rene_> MadCoder: is there a b0rkage of sigcontext.h known? I have it since 2.6.1-3
13:24 -!- kaeso [~luca@151.16.11.125] has quit [Quit: E quindi uscimmo a riveder le stelle]
13:24 -!- Irssi: Pasting 9 lines to #debian-devel. Press Ctrl-K if you wish to do this or Ctrl-C to cancel.
13:24 < _rene_> In file included from /usr/include/asm/sigcontext.h:8, from system.h:86, from conditn.c:38:
13:24 < _rene_> /usr/include/asm-i386/sigcontext.h:19: error: redefinition of 'struct _fpreg'
13:24 < _rene_> /usr/include/asm-i386/sigcontext.h:24: error: redefinition of 'struct _fpxreg'
13:24 < _rene_> /usr/include/asm-i386/sigcontext.h:30: error: redefinition of 'struct _xmmreg'
13:24 < _rene_> /usr/include/asm-i386/sigcontext.h:34: error: redefinition of 'struct _fpstate'
13:24 < _rene_> /usr/include/asm-i386/sigcontext.h:58: error: redefinition of 'struct sigcontext'
13:24 < _rene_> dmake: Error code 1, while making '../../unxlngi6.pro/slo/conditn.obj'
[...]
14:10 < MadCoder> _rene_: I suppose there is a problem with linux-libc-dev again
14:11 < MadCoder> _rene_: you have to file a bug against it, it's not a libc problem
[...]
16:30 < MadCoder> waldi: it's bits/sigcontext.h
16:30 < MadCoder> not sys/
16:30 < MadCoder> sorry
16:31 < noshadow> mrvn: install has the disadvantage that one has to think about all permissions and ownerships, just making
the changes one likes after dh_fixperms makes sure only those are changed.
16:31 < mrvn> noshadow: true.
16:31 < dendrite> noshadow: In this case we have to think about all if it anyhow, so that's not a real downer.
16:31 < dendrite> noshadow: (for us.)
16:31 < waldi> MadCoder: which was mostly empty before the last update
16:31 < waldi> MadCoder: no bug in the kernel
16:31 < MadCoder> hmmmm
16:31 < mrvn> dendrite: I find chown -R actually quite efficient.
16:31 < dendrite> I was also mightily confused between DEBIAN and debian directories.
16:32 < MadCoder> and aurel32 is off for some days
16:32 < dendrite> mrvn: Yeah, it certainly works. :-)
16:32 < MadCoder> and he did the last upload
16:32 < MadCoder> hmmm
16:32 < MadCoder> waldi, _rene_: okay, I'll try to see what's the problem is then
16:32 < noshadow> dendrite: just make sure all those things are set in the binary (or rules invoked at that time like install)
of debian/rules. So that at package creation time fakeroot still knows what permissions you set them to.
16:32 < dendrite> mrvn: Thanks.
16:33 < dendrite> debian/rules...
16:33 < mrvn> dendrite: indeed it does :)
16:33 < dendrite> mrvn: =)
16:34 < waldi> MadCoder: the kernel exports its interface for more than glibc
16:34 -!- bandini (Michele Baldessari) [~bandini@host11-39-dynamic.4-87-r.retail.telecomitalia.it] has joined #debian-devel
16:34 < MadCoder> and ?
16:35 < _rene_> MadCoder: as said, it broke in -3. afair, -2 did work fine.
16:35 < _rene_> MadCoder: (fwiw)
16:35 -!- foka (Anthony Fok) [~foka@125.33.8.177] has joined #debian-devel
16:35 < MadCoder> yeah it seems to be related to the:
16:35 < MadCoder> * sysdeps/amd64.mk: uses x86_64 headers also for the i486 flavour now
16:35 < MadCoder> that they are compatible.
16:35 < MadCoder> * sysdeps/i386.mk: uses x86_64 headers also for the main flavour.
16:36 < MadCoder> the problematic header comes from the x86_64/bits/ directory of the glibc sources
16:36 < dendrite> So the debian/rules file is basically a makefile, and doesn't actually create explicit settings for the
packaged files? That really needs to be done before the package is created, using the usual commands?
16:38 < waldi> MadCoder: yes. sigcontext.h for xx86_64 is split into two
[...]
16:49 < MadCoder> _rene_: okay it's definitely a libc problem
16:49 < dendrite> Yoe: No, I can IRC @ work, I'm just too busy coding then. :-)
16:49 < MadCoder> I won't have tiem to deal with it right now, so please open a bug (RC :/) so that I think about
Doing :)
Regards,
Rene
Reply to: