Nptl na sparcu

[Jakub Bogusz]

On Sun, Oct 17, 2004 at 02:43:42PM +0200, Paweł Sikora wrote:
> On Saturday 16 of October 2004 22:47, Mariusz Mazur wrote:
> > On sobota 16 październik 2004 18:57, Mariusz Mazur wrote:
> > > ranlib /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.a
> > > sparc-pld-linux-gcc    -nostdlib -nostartfiles -r
> > > -o /home/users/builder/rpm/BUILD/libc/builddir/elf/librtld.map.o
> > > '-Wl,-(' /home/users/builder/rpm/BUILD/libc/builddir/elf/dl-allobjs.os
> > > /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.a -lgcc '-Wl,-)'
> > > -Wl,-Map,/home/users/builder/rpm/BUILD/libc/builddir/elf/librtld.mapT
> > > /usr/bin/ld: __libc_errno: TLS definition
> > > in /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.a(errno.os)
> > > section .tbss mismatches non-TLS reference
> > > in /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.a(check_fds.os)
> > > /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.a: could not read
> > > symbols: Bad value
> > > collect2: ld returned 1 exit status
> > >
> > > Problem z głowy :) Na sparcu glibc bez nptla i bez tlsa.
> >
> > sparc-pld-linux-gcc    -nostdlib -nostartfiles -r
> > -o /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.os \
> >  -Wl,-d
> > -Wl,--whole-archive /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.a
> > /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.a(malloc.os)
> > (.bss[.__sparc32_atomic_locks]+0x0): multiple definition of
> > `__sparc32_atomic_locks'
> > /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.a(cxa_finalize.os)
> > (.bss[.__sparc32_atomic_locks]+0x0): first defined here
> > /home/users/builder/rpm/BUILD/libc/builddir/libc_pic.a(set-freeres.os)
> > (.bss[.__sparc32_atomic_locks]+0x0): multiple definition of
> > `__sparc32_atomic_locks'
> >
> > Właściwie rzecz biorąc, to w ogóle rezygnujemy ze sparca, bo glibca nie ma
> > :P
> 
> zgłosiłeś błąd na http://sources.redhat.com/bugzilla/enter_bug.cgi,
> a może wolisz tak stać jak cieć przy hałdzie na wysypisku i jęczeć?

http://sources.redhat.com/bugzilla/show_bug.cgi?id=333
taki jest efekt zgłaszania błędów budowania w bugzilli
(tego _jeszcze_ nie)


-- 
Jakub Bogusz    http://cyber.cs.net.pl/~qboosh/