* Emba/Gitlab tests
@ 2020-08-26 22:34 Alan Third
2020-08-27 8:52 ` Daniel Martín
2020-08-27 9:13 ` Michael Albinus
0 siblings, 2 replies; 6+ messages in thread
From: Alan Third @ 2020-08-26 22:34 UTC (permalink / raw)
To: Emacs-Devel devel
I was looking at the tests running on emba.gnu.org and had the idea that it
might make sense to add a simple build job to build against GNUstep
when any of the NS files are modified.
I don't think it would be worth running the test suite against it, but
it would still be potentially useful to know that the GNUstep build
isn't completely broken.
Does that sound OK, or are there any concerns about running too many
jobs?
--
Alan Third
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Emba/Gitlab tests
2020-08-26 22:34 Emba/Gitlab tests Alan Third
@ 2020-08-27 8:52 ` Daniel Martín
2020-08-27 9:13 ` Michael Albinus
1 sibling, 0 replies; 6+ messages in thread
From: Daniel Martín @ 2020-08-27 8:52 UTC (permalink / raw)
To: Alan Third; +Cc: Emacs-Devel devel
Alan Third <alan@idiocy.org> writes:
> I was looking at the tests running on emba.gnu.org and had the idea that it
> might make sense to add a simple build job to build against GNUstep
> when any of the NS files are modified.
>
> I don't think it would be worth running the test suite against it, but
> it would still be potentially useful to know that the GNUstep build
> isn't completely broken.
>
> Does that sound OK, or are there any concerns about running too many
> jobs?
Makes sense. The earlier that compilation problems in the GNUstep build
are detected, the better. I suppose the number of patches changing
something in GNUstep is small compared to the rest of changes, so the
impact on the number of running jobs should be small, but I don't have
numbers to back this up.
--
Daniel Martín
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Emba/Gitlab tests
2020-08-26 22:34 Emba/Gitlab tests Alan Third
2020-08-27 8:52 ` Daniel Martín
@ 2020-08-27 9:13 ` Michael Albinus
2020-08-27 17:31 ` Alan Third
1 sibling, 1 reply; 6+ messages in thread
From: Michael Albinus @ 2020-08-27 9:13 UTC (permalink / raw)
To: Alan Third; +Cc: Emacs-Devel devel
Alan Third <alan@idiocy.org> writes:
Hi Alan,
> I was looking at the tests running on emba.gnu.org and had the idea that it
> might make sense to add a simple build job to build against GNUstep
> when any of the NS files are modified.
>
> I don't think it would be worth running the test suite against it, but
> it would still be potentially useful to know that the GNUstep build
> isn't completely broken.
>
> Does that sound OK, or are there any concerns about running too many
> jobs?
Fine with me. If the load on emba will be too high, we can still adjust.
Best regards, Michael.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Emba/Gitlab tests
2020-08-27 9:13 ` Michael Albinus
@ 2020-08-27 17:31 ` Alan Third
2020-08-27 18:00 ` Michael Albinus
0 siblings, 1 reply; 6+ messages in thread
From: Alan Third @ 2020-08-27 17:31 UTC (permalink / raw)
To: Michael Albinus; +Cc: Ted Zlatanov, Emacs-Devel devel
On Thu, Aug 27, 2020 at 11:13:44AM +0200, Michael Albinus wrote:
>
> Fine with me. If the load on emba will be too high, we can still adjust.
Thanks.
Does this look acceptable? I don't want to have to push a whole bunch
of commits to work out any kinks, but I don't know how else to test
the code.
modified .gitlab-ci.yml
@@ -65,3 +65,29 @@ test-filenotify-gio:
- ./configure --without-makeinfo --with-file-notification=gfile
- make bootstrap
- make -C test autorevert-tests filenotify-tests
+
+test-gnustep:
+ stage: test
+ # This tests the GNUstep build process
+ only:
+ changes:
+ - .gitlab-ci.yml
+ - configure.ac
+ - src/nsfns.m
+ - src/nsfont.m
+ - src/nsgui.h
+ - src/nsimage.m
+ - src/nsmenu.m
+ - src/nsselect.m
+ - src/nsterm.h
+ - src/nsterm.m
+ - src/nsxwidget.h
+ - src/nsxwidget.m
+ - src/macfont.h
+ - src/macfont.m
+ script:
+ - DEBIAN_FRONTEND=noninteractive apt install --no-install-recommends -y -qq -o=Dpkg::Use-Pty=0 gnustep-devel
+ - ./autogen.sh autoconf
+ - ./configure --without-makeinfo --with-ns
+ - make bootstrap
+ - make install
--
Alan Third
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Emba/Gitlab tests
2020-08-27 17:31 ` Alan Third
@ 2020-08-27 18:00 ` Michael Albinus
2020-08-27 19:14 ` Alan Third
0 siblings, 1 reply; 6+ messages in thread
From: Michael Albinus @ 2020-08-27 18:00 UTC (permalink / raw)
To: Alan Third; +Cc: Ted Zlatanov, Emacs-Devel devel
Alan Third <alan@idiocy.org> writes:
Hi ALan,
> Does this look acceptable? I don't want to have to push a whole bunch
> of commits to work out any kinks, but I don't know how else to test
> the code.
Well, finally you have to do until it works as expected. Create a new
branch, and test this.
> + - ./autogen.sh autoconf
> + - ./configure --without-makeinfo --with-ns
> + - make bootstrap
> + - make install
I have no idea what "make install" will do on emba. Do you need it? The
other test sections don't run it.
OTOH, if you have a test file specific for gnustep, you could run it by
- make -C test gnustep-tests
Best regards, Michael.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Emba/Gitlab tests
2020-08-27 18:00 ` Michael Albinus
@ 2020-08-27 19:14 ` Alan Third
0 siblings, 0 replies; 6+ messages in thread
From: Alan Third @ 2020-08-27 19:14 UTC (permalink / raw)
To: Michael Albinus; +Cc: Ted Zlatanov, Emacs-Devel devel
On Thu, Aug 27, 2020 at 08:00:59PM +0200, Michael Albinus wrote:
> Alan Third <alan@idiocy.org> writes:
>
> Hi ALan,
>
> > Does this look acceptable? I don't want to have to push a whole bunch
> > of commits to work out any kinks, but I don't know how else to test
> > the code.
>
> Well, finally you have to do until it works as expected. Create a new
> branch, and test this.
That makes sense.
> > + - ./autogen.sh autoconf
> > + - ./configure --without-makeinfo --with-ns
> > + - make bootstrap
> > + - make install
>
> I have no idea what "make install" will do on emba. Do you need it? The
> other test sections don't run it.
On NS builds it packages everything into nextstep/Emacs.app, and
doesn't actually "install" anything. It's probably not necessary, but
I think it makes sense to run it.
Unless this Gitlab instance is set up funny all the jobs should be run
in their own docker containers anyway, so you always get a clean
environment every time. If you want to retain anything between jobs
you have to use the "Artifacts" keyword, and that's only for a single
pipeline anyway, iirc.
> OTOH, if you have a test file specific for gnustep, you could run it by
>
> - make -C test gnustep-tests
Unfortunately we don't have any, as far as I'm aware.
--
Alan Third
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2020-08-27 19:14 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-08-26 22:34 Emba/Gitlab tests Alan Third
2020-08-27 8:52 ` Daniel Martín
2020-08-27 9:13 ` Michael Albinus
2020-08-27 17:31 ` Alan Third
2020-08-27 18:00 ` Michael Albinus
2020-08-27 19:14 ` Alan Third
Code repositories for project(s) associated with this public inbox
https://git.savannah.gnu.org/cgit/emacs.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).