Re: fixing ci failure of 'pu' with the es/bugreport topic

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Emily,

On Wed, Apr 15, 2020 at 7:01 PM Emily Shaffer <emilyshaffer@xxxxxxxxxx> wrote:
>
> On Wed, Apr 15, 2020 at 6:28 PM Junio C Hamano <gitster@xxxxxxxxx> wrote:
> >
> > Junio C Hamano <gitster@xxxxxxxxx> writes:
> >
> > > Danh Doan <congdanhqx@xxxxxxxxx> writes:
> > >
> > >> Excerpt from build log:
> > >>
> > >>> fatal error C1083: Cannot open include file: 'config-list.h'
> > >>
> > >> It's from bugreport topic.
> > >> I've seen this failure in the past (when testing with pu),
> > >> then I saw it disappear.
> > >>
> > >> I thought it was fixed during my testing for v4.
> > >
> > > Is the issue something similar to 976aaedc (msvc: add a Makefile
> > > target to pre-generate the Visual Studio solution, 2019-07-29)?
> > >
> > > If that is the case, perhaps something like this would help?  I'll
> > > tentatively queue it on top of es/bugreport and merge the result to
> > > 'pu' to see what happens.
> >
> > The build just passed: https://github.com/git/git/runs/590781044
> >
> > Emily, you may need to squash in something along the line of this
> > change to the commit in your series that starts building and using
> > the config-list.h file (was it the first one?).  I've queued mine
> > as a follow-up "oops, it was wrong" patch, but that would not be
> > kosher from bisectability's point of view.
>
> Hm, ok. I'll send a reroll squashing this in verbatim tomorrow unless
> I hear otherwise from Dscho? Looks like it's indeed the first one
> (dd763e).
> I'm curious to know how I can check this build method for myself for next time.

Create a fork of github.com/git/git and open a pull request against
it.  (I believe you could also fork github.com/gitgitgadget/git and do
a pull request against it, but I switched over to /git/git a while
ago.) Immediately upon opening the pull request, a bunch of linux,
mac, windows, and freebsd builds will be triggered with various runs
of the testsuite.  Has been very useful for catching issues for me
before I sent them off to the list.

You can also make use of Dscho's gitgitgadget magic at that point to
take care of the git send-email step for you too by commenting
'/submit' in the PR, though you don't have to do that.  It's perfectly
fine to just open a PR for the automated testing and then close the PR
and do the rest yourself.  But if you leave it open and have it submit
the patch emails for you, it'll track their status.  I kinda like
being able to go to https://github.com/git/git/pulls/newren and have
it track the state of where all my open pull requests are.  (You might
even be able to click through some of those to see example build
results)

Dscho has done some great work with his gitgitgadget work and azure
builds.  SZEDER Gábor also has a few builds triggered via Travis that
check a few more things off the same PRs (static anlysis and whatnot).
I've been very happily using them all by just opening PRs, and have
appreciated the heads up of potential issues I would've otherwise
caused on various platforms otherwise.


Hope that helps,
Elijah




[Index of Archives]     [Linux Kernel Development]     [Gcc Help]     [IETF Annouce]     [DCCP]     [Netdev]     [Networking]     [Security]     [V4L]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [Fedora Users]

  Powered by Linux