[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#617299: dpkg-deb: should give a hint when it fails due to filling /tmp



(dropping debian-kernel from CC as we're back to discussing the dpkg
issue itself)
2012/7/14 Jonathan Nieder <jrnieder@gmail.com>:
> clone 617299 -1
> retitle -1 kernel-handbook: encourage disabling DEBUG_INFO and setting TMPDIR
> submitter -1 !
> reassign -1 debian-kernel-handbook 1.0.13
> quit
>
> Martin-Éric Racine wrote:
>
>> The solution to bug #617299 is therefore to either define TMPDIR to
>> some actual hard-disk with sufficient storage space via a commandline
>> environment e.g. 'TMPDIR=/var/tmp make deb-pkg' whenever using
>> commands that require an unusually large amount of /tmp space, or to
>> completely disable the usage of a tmpfs for /tmp via initscripts
>> defaults in /etc/defaults/rcS and /etc/defaults/tmpfs, as appropriate.
>
> Sort of.  It's still a dpkg (wishlist?) bug.  I would even think it
> might be possible for dpkg-deb to write its output directly to the
> expected target path or another file in the same directory.

Writing to the target path would indeed make sense. Just a wild guess,
but dpkg probably uses /tmp because, by definition, it's guaranteed to
be temporary and deleted upon reboot, whereas writing to target path
might leave an increasing amount of cruft whenever a build repeatedly
fails.

Still, as the original reporter complained, blindly trying to figure
out why a package repeatedly fails to build with "No space left" -
even though it compiled succesfully - on a hard-disk with several GB
of space, when it turns out that it's simply because dpkg uses as a
scratchpad a /tmp that happens to be mounted as a small tmpfs, is a
genuine problem. I am curious to hear what dpkg developers have to say
about solving this particular one.

Martin-Éric



Reply to: