And now I see it on ubuntu 17.10, this time building a gst-plugins-good:

dh_shlibdeps -l/opt/blort/deps-64-12/lib -- --ignore-missing-info
semop(1): encountered an error: Invalid argument

Seems to be a persistent, low-probability flakiness.


On Mon, Apr 3, 2017 at 10:19 PM, Dan Kegel <d...@kegel.com> wrote:
> aha, it also failed two months ago (really, two builds ago, I haven't
> been running this often) at hour 5 with
>
> + fakeroot sh /tmp/maketar.sh.JQLvyH7VMj
> opt/blort/deps-32-11/
> opt/blort/deps-32-11/include/
> opt/blort/deps-32-11/include/x264_config.h
> opt/blort/deps-32-11/include/x264.h
> opt/blort/deps-32-11/bin/
> opt/blort/deps-32-11/bin/x264
> semop(1): encountered an error: Invalid argument
> /usr/bin/fakeroot: line 1: kill: (7959) - No such process
> Makefile:14: recipe for target 'all' failed
> make[1]: *** [all] Error 1
>
> so maybe I'll try running fakeroot tar in a loop.
>
> On Mon, Apr 3, 2017 at 10:15 PM, Dan Kegel <d...@kegel.com> wrote:
>> I may have seen something similar with ubuntu core today on raspberry pi3
>> (using the classic snap, which is at 16.04.2)
>> fakeroot is 1.20.2-1ubuntu1.
>>
>> Here's what I saw while packaging a version of graphicsmagick (about three 
>> hours
>> into a script that
>>
>>    dh_builddeb
>> dpkg-deb: building package 'blort-graphicsmagick' in
>> '../blort-graphicsmagick_XX.YY_armhf.deb'.
>> semop(1): encountered an error: Invalid argument
>> dpkg-deb: error: subprocess tar -cf returned error exit status 1
>> dh_builddeb: dpkg-deb --build debian/blort-graphicsmagick .. returned
>> exit code 2
>> semop(1): encountered an error: Identifier removed
>> dh_builddeb: dpkg-deb --build debian/blort-graphicsmagick-dbg ..
>> returned exit code 1
>> debian/rules:4: recipe for target 'binary' failed
>> make: *** [binary] Error 1
>> /usr/bin/fakeroot: line 1: kill: (8907) - No such process
>> dpkg-buildpackage: error: fakeroot debian/rules binary gave error exit 
>> status 2
>> debuild: fatal error at line 1376:
>> dpkg-buildpackage -rfakeroot -D -us -uc -b failed
>> program finished with exit code 29
>>
>> I have not tried your patches or turning on debugging output yet.
>> I'm rerunning debuild now, but it's bedtime, I'll have to check in the 
>> morning
>> (raspberry pi rather dreadfully slow).

Reply via email to