Am 01.08.10 00:52, schrieb Lucas Nussbaum: > Source: monotone > Version: 0.48-1 > Severity: serious > Tags: squeeze sid > User: debian...@lists.debian.org > Usertags: qa-ftbfs-20100731 qa-ftbfs > Justification: FTBFS on amd64 > > Hi, > > During a rebuild of all packages in sid, your package failed to build on > amd64. > > Relevant part: [snip] >> 235 fail_cleanly_on_unreadable_db FAIL (line 51) 0:01, 0:00 >> on CPU [snip]
This is a weird failure. The test in question marks a directory unwritable via "chmod a-w" and then tries to perform a database migration which needs write access to the database. The migration is expected to fail because the database journal should not be writable to the directory, but from what I can see in the logs the migration runs through without erroring out anything. As the test works for me locally, I think there is something weird on the test machine. The chmod call seems to go through properly as well (exit status 0, nothing on stderr), but I suspect that the directory is still somehow writable for the current user, e.g. because of a dangling ACL which still allows write access. One could disable this test or place a guard around it, for example by creating a file in the directory and check for its existance after the chmod call. If it does exists, skip the test because the filesystem is still writable. Thomas. -- GPG-Key 0x160D1092 | tommyd3...@jabber.ccc.de | http://thomaskeller.biz Please note that according to the EU law on data retention, information on every electronic information exchange might be retained for a period of six months or longer: http://www.vorratsdatenspeicherung.de/?lang=en
signature.asc
Description: OpenPGP digital signature
