[Angstrom-devel] clean systemd-image build failing with non-zero exit ...

[Denys Dmytriyenko]

On Wed, Mar 28, 2012 at 01:45:06PM -0400, Denys Dmytriyenko wrote:
> On Wed, Mar 28, 2012 at 09:48:56AM -0600, Gary Thomas wrote:
> > On 2012-03-28 08:27, Martin Jansa wrote:
> > >On Tue, Mar 27, 2012 at 10:07:54PM -0700, Russell Senior wrote:
> > >>
> > >>... and no effing clue what exited, except it was part of do_rootfs.
> > >>The error message is emitted in the middle of another message relating
> > >>to some pam modules, but it is not at all obvious that the error is
> > >>related to pam at all.  The error message points at the log file, but
> > >>the log file gives no clue what went wrong.  The messages that get
> > >>interrupted seem to be related to the .dot file creation, but the .dot
> > >>file appears complete in the aftermath.
> > >>
> > >>http://www.tdb.com/~russell/log.do_rootfs.txt
> > >>
> > >>wtf?
> > >>
> > >>This was on an up-to-date angstrom checkout and update as of last
> > >>night, compiling for beaglebone.
> > >>
> > >>MACHINE=beaglebone bitbake systemd-image
> 
> Yep, been seeing this issue past couple days on Cloud9 image...
> 
> > >Can you retry with buildhistory disabled?
> > 
> > What's the best way to disable buildhistory?
> 
> In meta-angstrom:
> 
> 
> diff --git a/conf/distro/include/angstrom.inc 
> b/conf/distro/include/angstrom.inc
> index 1290352..81d7e7d 100644
> --- a/conf/distro/include/angstrom.inc
> +++ b/conf/distro/include/angstrom.inc
> @@ -101,7 +101,7 @@ QA_LOG = "1"
>  
>  # Log information on images and packages
>  # https://github.com/Angstrom-distribution/buildhistory/
> -INHERIT += "buildhistory"
> +#INHERIT += "buildhistory"
>  BUILDHISTORY_COMMIT = "1"
>  
>  #run QA tests on recipes
> 
> 
> Trying it now...

Ok, I can confirm that disabling buildhistory with the patch above allows 
systemd-based images to successfully finish do_rootfs.

-- 
Denys