Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 5 May 2016 17:18:54 -0700 (PDT)
From:      Don Lewis <truckman@FreeBSD.org>
To:        rkoberman@gmail.com
Cc:        lists@opsec.eu, freebsd-ports@freebsd.org, bradleythughes@fastmail.fm
Subject:   Re: PRs ready for commit
Message-ID:  <201605060018.u460Isuj055739@gw.catspoiler.org>
In-Reply-To: <CAN6yY1v68qp_kFvkCOkRWVBXroSj-uPpCJ_mPpJTULSFMZErQA@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On  5 May, Kevin Oberman wrote:
> On Thu, May 5, 2016 at 2:16 PM, Don Lewis <truckman@freebsd.org> wrote:
> 
>> On  5 May, Kurt Jaeger wrote:
>> > Hi!
>> >
>> > One thing: There are issues with buildlogs attached to PR due
>> > to them blowing up the database of bugzilla. If all is fine,
>> > it's OK to just list the architectures tested. Second best
>> > is putting broken buildlogs on some external site and put the link
>> > in the PR. Third is adding broken buildlogs as attachment 8-}
>> >
>> > Yes, I know, disk space is infinite, but this is real life 8-)
>>
>> For broken builds, I prefer to just include the part of the build log
>> showing the error in the comment.  If someone thinks that there is still
>> some info necessary for debug that is missing, I can extract it from the
>> log and add it as another comment.  That way the info is easy to access
>> and permanently part of the PR.
>>
>> _______________________________________________
>> freebsd-ports@freebsd.org mailing list
>> https://lists.freebsd.org/mailman/listinfo/freebsd-ports
>> To unsubscribe, send any mail to "freebsd-ports-unsubscribe@freebsd.org"
>>
> 
> I suggest that it should be the end of the build log with the build done
> with "MAKE_JOBS_UNSAFE" defined so the error is likely to actually be at
> the end of the log.

That's generally the easiest way.  If it is a parallel build, just
attaching the tail of the log file is generally pretty useless.
Sometimes I'll rerun a failing parallel build single threaded to so that
the point of failure is close to the end where I can easily grab it.
That doesn't work if the problem is actually caused by the parallel
build (and the build is clean when done single threaded).  In that case
I search backwards from the end to find exactly what caused the failure.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201605060018.u460Isuj055739>