Re: Bug#253030: libc6.1 2.3.2.ds1-13 ia64 is broken
At Mon, 7 Jun 2004 18:35:29 -0600,
LaMont Jones wrote:
> On Tue, Jun 08, 2004 at 12:39:10AM +0100, Colin Watson wrote:
> > On Mon, Jun 07, 2004 at 09:31:23PM +0900, GOTO Masanori wrote:
> > > At Sun, 6 Jun 2004 20:21:57 +0200,
> > > Ooh what has happened!?
> > According to the GPG signature on the .changes, it was LaMont ...
>
> And I uploaded what the autobuilder built:
>
> >From debian/control:
> Package: nscd
> Architecture: alpha arm i386 m68k mips mipsel powerpc sparc ia64 hppa s390 sh3 sh4 sh3eb sh4eb freebsd-i386
> ...
> Depends: libc6 (>= ${Source-Version})
>
> That would be a package bug, near as I can tell.
Duh. Of course the current problem is debian/control Depends field
shows libc6, not libc6.1.
Alpha needs to depends on libc6.1, and the build is correctly
succeeded and it does not break. I have no idea that it's really
package bug or not.
In past Jeff Bailey built glibc on ia64, and if it was correctly
built, then that version of nscd should depend on libc6.1. Jeff,
could you check it?
In debian/rules.d/control.mk generates:
sed -e 's%@libc@%$(libc)%g;s%@glibc@%glibc%g' \
-e 's%@threads_archs@%$(threads_archs)%g' < $@T > $@
($@ == debian/control)
from the input data:
Package: nscd
Architecture: @threads_archs@
Section: admin
Priority: optional
Depends: @libc@ (>= ${Source-Version})
Replaces: @libc@ (<< 2.1-4)
So this $(libc) is incorrectly set as "libc6", not "libc6.1".
We only know that the package Depends field is not correct; Lamont, do
you have build log or the environment to reappear this problem?
Apparenly this behavior is out of our intension.
Regards,
-- gotom
Reply to: