[Pkg-gmagick-im-team] Bug#536811: imagemagick: FTBFS with parallel in DEB_BUILD_OPTIONS

[Daniel Schepler]

clone 536811 -1
tags -1 - unreproducible moreinfo
severity -1 serious
retitle -1 imagemagick: FTBFS: Missing Build-Depends on libgomp1
severity 536811 normal
retitle 536811 imagemagick: Race condition building perlmagick
thanks

On Monday 13 July 2009 11:32:35 Nelson A. de Oliveira wrote:
> tags 536811 + unreproducible, moreinfo
> thanks
>
> On Mon, Jul 13, 2009 at 3:04 PM, Daniel Schepler<dschepler at gmail.com> wrote:
> > From my pbuilder build log, using a configuration which sets
> > DEB_BUILD_OPTIONS="parallel=3":
>
> I build my packages with -j4 (that also adds parallel=4 to
> DEB_BUILD_OPTIONS). Even with 3 threads I can't reproduce this.
>
> Maybe it's something wrong here (gcc or binutils maybe? could also be
> arch specific, as my environment is i386 and yours seems to be amd64):
>
> /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/4.3.3/libgomp.a(critical.o):
> relocation R_X86_64_32 against `.bss' can not be used when making a
> shared object; recompile with -fPIC
> /usr/lib/gcc/x86_64-linux-gnu/4.3.3/libgomp.a: could not read symbols: Bad
> value collect2: ld returned 1 exit status
>
> Is the package properly built if not using parallel?

OK, it turned out that it also doesn't build if not using parallel, with the 
same libgomp error.  I just thought it was a parallel build issue because it 
showed PerlMagick trying to find libMagickCore before it was built -- which 
still seems like a potential race condition to me.

The cause turned out to be that my recently created pbuilder chroot doesn't 
have libgomp1 installed.  If I installed libgomp1 in the chroot, the package 
built fine.
-- 
Daniel Schepler