Control: found -1 2.54.1-1
On Mon, 18 Dec 2017 at 18:18:41 +0000, Simon McVittie wrote:
> On Mon, 18 Dec 2017 at 09:17:46 +0000, Simon McVittie wrote:
> > The failure on the buildd is new, although I think I might have seen it
> > before in local testing (but never reproducibly). It's probably some
> > rarely-hit race condition?
>
> I can reproduce this by running the gdbus-peer test in a loop (this is
> with GLib git master, not Debian's GLib, but that's close enough).
I can also reproduce it by building 2.54.1-1 or 2.54.2-2 and re-running
that test in a loop, but not easily (running 100 times in a loop often
succeeds). Having looked at the history of the code involved, I think
this is a long-standing race condition; so this isn't a regression,
and shouldn't block testing migration.
smcv
