Re: Log for attempted build of gnupg2_2.0.21-2 on m68k (dist=unstable)
On Wed, 4 Sep 2013, Thorsten Glaser wrote:
> Eric Dorland dixit:
>
> >> testing message digests takes too long on this slow platform (I
> >> noticed that as I ran top on the buildd while upgrading its chroot
> >> basedir) and builds are killed after inactivity for a while
>
> >Well that's a shame. Any suggestions?
>
> Only the really unconstructive one:
>
> Exclude the tests if DEB_BUILD_ARCH (not DEB_HOST_ARCH because the
> slowness is a property of the system compiling the source) matches m68k.
>
> Cc'ing the porters list for additional feedback, if any?
How about a test for actual inactivity? (E.g. processes blocked for too
long, or process cumulative CPU time not accumulating.)
Or just make the timeout adjustable or optional. When was the last time
you needed it?
Finn
>
> bye,
> //mirabilos
>
Reply to: