buildbot failure in Mercurial on hg tests

[Brendan Cully]

On 12-11-16 09:20 AM, Kevin Bullock wrote:
> On 16 Nov 2012, at 11:12 AM, hgbuildbot at kublai.com wrote:
>
>> The Buildbot has detected a new failure on builder hg tests while building hg.
>> Full details are available at:
>> http://hgbuildbot.kublai.com/builders/hg%20tests/builds/336
>>
>> Buildbot URL: http://hgbuildbot.kublai.com/
>>
>> Buildslave for this Build: p64.osuosl.test
>>
>> Build Reason: scheduler
>> Build Source Stamp: [branch default] a1f94e2f5da2dee4e107eab035685285d9923999
>> Blamelist: Kevin Bullock <kbullock at ringworld.org>
>>
>> BUILD FAILED: failed run-tests.py (python2.4) run-tests.py (python2.5) run-tests.py (python2.6) pure http2
>
> --- /var/lib/buildbot/slaves/p64/hg_tests/build/tests/test-rebase-collapse.t
> +++ /var/lib/buildbot/slaves/p64/hg_tests/build/tests/test-rebase-collapse.t.err
> @@ -415,308 +415,9 @@
>   Create repo d:
>
>     $ hg init d
> +  abort: No space left on device: $TESTTMP/d
> +  [255]
>     $ cd d
> +  /tmp/tmpdYCyt7hg-tst: 216: cd: can't cd to d
>
> So that's probably not good. But probably unrelated to the changes I pushed.
>
> pacem in terris / мир / शान्ति / ‎‫سَلاَم‬ / 平和
> Kevin R. Bullock
>

/tmp was full of svnwrap_test and hgtest leftovers from I guess previous 
failed runs. I've cleaned that up, and made tests run in temporary 
tmpdirs, which should prevent this from repeating.