On Sun, Jul 08, 2012 at 06:19:19PM -0500, Bruce Dubbs wrote:
> Ken Moffat wrote:
> > On Sun, Jul 08, 2012 at 05:11:30PM -0500, Bruce Dubbs wrote:
> >> I'm building gegl as a dependency for gimp and ran into a test problem.
> >>    There are 8 failures, but they are all the same.  The are doing:
> >>
> >> GEGL_PATH=../operations ./img_cmp  file1.png  file2.png
> >>
> >> The program is seg faulting.  When I test manually, I get:
> >>
> >> file1.png and file2.png are identical
> >> Segmentation fault
> >>
> >> According to the source, img_cmp "is a simple image difference detection
> >> tool for use in regression testing"
> >>
> >> Has anyone else seen this problem?
> 
> >   I think I've seen failures in the past (the files looked similar in
> > 'display' but were actually different), but not the segfault.
> > However, looking at my notes for 0.2.0 I don't see any mention of
> > problems.  Let me try on a recent system (I'm currently on LFS-7.0
> > which is too old to be useful).
> 
> I went ahead and built gimp and it seems to run OK for the little 
> testing I've done.  I haven't tried to build the help though.
> 
>    -- Bruce
 On my most recent system (LFS from 20120610, but parts of BLFS are
probably now "old") on x86_64 the testsuite from gegl all passes
without problems.  That is on a *complete* system, no idea what
would happen at the point where I usually build gegl without running
hte tests.  Gimp-help is unrelated to gegl (there is a switch to
build it without gimp) - but you probably only want 'en'.

 I'm moderately surprised that you seem to be running tests on
everything (I thought you had taken a similar stance to myself and
Andy for *blfs* tests, and I suspect that LFS-7.2 will turn out to
be sufficiently different from current -svn to make any results
meaningless, e.g. because of glibc-2.16), but it's your system and
your CPU cycles :)

ĸen
-- 
das eine Mal als Tragödie, das andere Mal als Farce
-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Reply via email to