On Thu, May 08, 2014 at 05:25:02PM +0200, Matthias Klose wrote:
>
> I would like to see some partial test rebuilds (like buildd or minimal chroot
> packages) for other architectures. Any possibility to setup such a test
> rebuild
> for some architectures by the porters? Afaics the results for the
This comes from you doing your directory->symlink migration in preinst,
rather than postinst (to see an example of it being done correctly, see
openssh-server, or php4-mysql as examples).
If you migrate in preinst, you remove the directory, create the symlink,
THEN dpkg removes your package's file
Package: g++-3.3
Version: 1:3.3.5-2
Severity: important
While building bochs on m68k, g++-3.3 fails with the following error:
m68k-linux-g++ -c -I.. -I./.. -I../instrument/stubs
-I./../instrument/stubs -I. -I./. -I./stubs -I././stubs -g -O2
-D_FILE_OFFSET_BITS=64 -D_LARGE_FILES `sdl-config
Package: gcc-snapshot
Version: 20041030-1
Severity: serious
gcc-snapshot ICEs during (or shortly after?) stage1 when building
on m68k. Please see the build log at buildd.debian.org[1] for
further info.
... Adam
[1]
http://buildd.debian.org/fetch.php?&pkg=gcc-snapshot&ver=20041030-1&arch=m68k&s
rst use in this
function)
../../src/gcc/calls.c:301: error: (Each undeclared identifier is reported only
once
../../src/gcc/calls.c:301: error: for each function it appears in.)
make[4]: *** [calls.o] Error 1
A full build log[1] is available at buildd.debian.org.
... Adam Conrad
Matthias Klose wrote:
>
> Any news?
>
> Thanks, Matthias
The rebuild failed as well. I just forgot to follow up to the bug.
... Adam
[EMAIL PROTECTED] wrote:
>
> So it appears linux/limits.h is not there. It seems to me that it
> ought to be there. Looking at libc6-dev_2.3.1-17_m68k.deb, I see that
> linux/limits.h is included. So if it is not on the build machine, it
> appears that the installation of the build machine is brok
Matijs van Zuijlen wrote:
> On Sun, May 04, 2003 at 10:06:33AM -0600, Adam Conrad wrote:
> >
> > It's my buildd. I can provide the config.log when I get back from
lunch.
> > :)
>
> That would be great. TIA.
Here you go:
http://people.debian.org/~adconrad/utah-glx/
... Adam
Matijs van Zuijlen wrote:
> On Sun, May 04, 2003 at 05:15:19PM +0200, Martin v. L?wis wrote:
> > Can you provide the relevant fragment of config.log, for
> > more details?
>
> Sorry, but I don't have access to that machine. It's one of the
> autobuilders.
It's my buildd. I can provide the confi
> -Original Message-
> From: Mathieu CUNY [mailto:[EMAIL PROTECTED]
>
> ii binutils2.13.90.0.18-1.6
Upgrade again today. The bug was in binutils, and has been fixed.
... Adam
> -Original Message-
> From: Adam Conrad [mailto:[EMAIL PROTECTED]
>
> Output of the three gcc -v --save-temps invocations follow,
> and the preprocessed source is attached as a tar.bz2
Well, it is now, at any rate.
... Adam
pack.tar.bz2
Description: Binary data
0 (nil) (set (reg/v:DF 311)
(const_double:DF 0.0 [0x0.0p+0])) -1 (nil)
(nil))
../functions/pack.c:807: internal compiler error: in extract_insn, at
recog.c:2175
Please submit a full bug report,
with preprocessed source if appropriate.
See http://www.gnu.org/software/gcc/bugs.html> fo
r to make it work.
The GCC maintainer(s) (of which I am not one) have made their position
pretty clear on this in the past. GCC alternatives are asking for
trouble, because maintainer machines may not always be running the
default compilers, and things get messy, from a distribution standpoint.
13 matches
Mail list logo
