l...@gnu.org (Ludovic Courtès) writes:

> civodul pushed a commit to branch core-updates
> in repository guix.
> commit 90ea83ee95ad467fc4a49f4c42d0df94d2952330
> Author: Ludovic Courtès <l...@gnu.org>
> Date:   Sat Aug 5 00:02:55 2017 +0200
>     gnu: bluez: Work around test failure on ARM.
>     This aims to skip the segfaulting test at
>     <https://hydra.gnu.org/build/2195551>.

I also noticed that a great many failures on ARM are due to a bluez
failure, but this seems to be a case where Hydra is misbehaving again.

The bluez build on ARM, as found by searching for "bluez" in the
evaluation, actually succeeded:


However, this is not the same bluez derivation that Hydra is using as
inputs to many other packages.  Looking at the source code, I see only
one bluez package.  This reminds me of similar problems I had on
'core-updates' some time ago, when I resorted to making a new
'core-updates-2' jobset to work around this problem.

Any idea what's going on here?


Reply via email to