Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

gst-plugins-ugly1: update to 1.24.9 #18342

Merged
merged 1 commit into from
Nov 10, 2024

Conversation

mtelka
Copy link
Contributor

@mtelka mtelka commented Jul 31, 2024

No description provided.

@AndWac
Copy link
Contributor

AndWac commented Jul 31, 2024

We cannot publish encumbered packages at the moment. The new build system for it is not finished yet.

@mtelka
Copy link
Contributor Author

mtelka commented Jul 31, 2024

Okay. @AndWac, please ping me once it works. I'll mark this as a Draft in the meantime.

@mtelka mtelka marked this pull request as draft July 31, 2024 21:02
@mtelka
Copy link
Contributor Author

mtelka commented Oct 27, 2024

@AndWac what is the current plan to make publishing of encumbered packages working again? Thank you.

@AndWac
Copy link
Contributor

AndWac commented Oct 27, 2024 via email

@mtelka
Copy link
Contributor Author

mtelka commented Oct 28, 2024

Can I help somehow with this task? It is concerning that we are not able to publish new packages for so long time.

@AndWac
Copy link
Contributor

AndWac commented Oct 28, 2024 via email

@mtelka
Copy link
Contributor Author

mtelka commented Oct 28, 2024

@Toasterson how can I help you to make the encumbered publication working again?

@Toasterson
Copy link
Contributor

@mtelka sorry for the late reply, I had a lot of private stress.
The Original problem can be seen here https://hipster.openindiana.org/jenkins/job/oi-encumbered/381/console the Makefiles somehow have confusion about their working directory for every encumbered package when trying to bulk build. Individual works but not bulk. Since I cannot fix makefiles I had to look into finishing my rust based tools for repology/archive handling and building. I got it as far as being able to build ffmpeg with a few changes to the current splitting. https://github.com/OpenIndiana/openindiana-encumbered/tree/main/components/ffmpeg has the metadata setup for that and the code for the pkgdev utility is here. https://github.com/Toasterson/forge/tree/develop/crates/pkgdev invoking the utility like this pkgdev --gate encumbered.kdl build --component ffmpeg builds the package. When you are with your current dir in the repo root of https://github.com/OpenIndiana/openindiana-encumbered you have the following options to help.

  1. Fix the Makefiles
  2. Help finish the CI integration of pkgdev

Hope this gives some orienteation.

@mtelka
Copy link
Contributor Author

mtelka commented Nov 9, 2024

@mtelka sorry for the late reply, I had a lot of private stress. The Original problem can be seen here https://hipster.openindiana.org/jenkins/job/oi-encumbered/381/console the Makefiles somehow have confusion about their working directory for every encumbered package when trying to bulk build.

How can I reproduce the problem? I tried this and it seems to work:

$ /usr/bin/time gmake -C components/encumbered publish -k
gmake: Entering directory '/data/builds/oi-userland/components/encumbered'
building tools...
make[1]: Entering directory '/data/builds/oi-userland/tools'
make[1]: Leaving directory '/data/builds/oi-userland/tools'
/bin/mkdir -p /data/builds/oi-userland/i386/pkglint-cache
publish SDL2_mixer
publish audacity

(still running)

@mtelka
Copy link
Contributor Author

mtelka commented Nov 9, 2024

Okay, I think I found a way how to reproduce: when I run env - gmake -C components/encumbered clobber then it puts the tree into the state when it then fails like this:

$ env - gmake -C components/encumbered publish -k
gmake: Entering directory '/data/builds/oi-userland/components/encumbered'
building tools...
make[1]: Entering directory '/data/builds/oi-userland/tools'
make[1]: Leaving directory '/data/builds/oi-userland/tools'
/bin/mkdir -p /data/builds/oi-userland/i386/pkglint-cache
/bin/bash: line 1: cd: encumbered/meta-packages/history: No such file or directory
gmake: *** [/data/builds/oi-userland/components/Makefile:159: encumbered/meta-packages/history] Error 1
gmake: Target 'publish' not remade because of errors.
gmake: Leaving directory '/data/builds/oi-userland/components/encumbered'
$

I'll investigate and try to fix.

@mtelka
Copy link
Contributor Author

mtelka commented Nov 9, 2024

No. To reproduce the problem reliable one needs to run gmake (any target) with BASS_O_MATIC_MODE=incremental and the problem is in the components/encumbered/components.mk file:

$ rm -f components/encumbered/components.mk 
$ env - BASS_O_MATIC_MODE=incremental gmake -C components/encumbered clobber
gmake: Entering directory '/data/builds/oi-userland/components/encumbered'
Generating component list...
/bin/bash: line 1: cd: encumbered/meta-packages/history: No such file or directory
gmake: *** [/data/builds/oi-userland/components/Makefile:183: encumbered/meta-packages/history.nosetup] Error 1
gmake: Leaving directory '/data/builds/oi-userland/components/encumbered'
$ cat components/encumbered/components.mk
COMPONENT_DIRS += encumbered/meta-packages/history
$

Without BASS_O_MATIC_MODE=incremental the gmake works:

$ rm -f components/encumbered/components.mk
$ env - gmake -C components/encumbered clobber
gmake: Entering directory '/data/builds/oi-userland/components/encumbered'
Generating component list...
clobber SDL2_mixer
gmake[1]: Entering directory '/data/builds/oi-userland/components/encumbered/SDL2_mixer'
gmake[1]: Leaving directory '/data/builds/oi-userland/components/encumbered/SDL2_mixer'
clobber audacity
gmake[1]: Entering directory '/data/builds/oi-userland/components/encumbered/audacity'
gmake[1]: Leaving directory '/data/builds/oi-userland/components/encumbered/audacity'
clobber dav1d

^C
$ grep history components/encumbered/components.mk
COMPONENT_DIRS += meta-packages/history
$

@mtelka
Copy link
Contributor Author

mtelka commented Nov 9, 2024

I believe #19511 should fix the problem. I tested gmake with BASS_O_MATIC_MODE=incremental and it works.

@mtelka
Copy link
Contributor Author

mtelka commented Nov 9, 2024

The oi-encumbered build still fails, but I'm not sure what is the reason now. Maybe this?

+ ssh root@oirepo /root/graceful-restart-hipster-encumbered.sh
ssh: Could not resolve hostname oirepo: node name or service name not known

@mtelka
Copy link
Contributor Author

mtelka commented Nov 10, 2024

@AndWac I noticed that the oi-encumbered builds started to succeed. Does it mean that we are ready to merge encumbered PRs?

@AndWac
Copy link
Contributor

AndWac commented Nov 10, 2024

I think so.

@mtelka mtelka changed the title gst-plugins-ugly1: update to 1.24.6 gst-plugins-ugly1: update to 1.24.9 Nov 10, 2024
@mtelka
Copy link
Contributor Author

mtelka commented Nov 10, 2024

I think so.

So let's try it.

BTW, where can I find build logs (something similar to https://hipster.openindiana.org/logs/oi-userland/)?

@mtelka mtelka marked this pull request as ready for review November 10, 2024 21:17
@AndWac AndWac merged commit b19e510 into OpenIndiana:oi/hipster Nov 10, 2024
1 check failed
@mtelka mtelka deleted the gst-plugins-ugly1 branch November 10, 2024 21:21
@mtelka
Copy link
Contributor Author

mtelka commented Nov 10, 2024

The build failed and it looks like the failure is this:

pkgrecv: Could not operate on /zdata/oirepo/hipster-encumbered/publisher/hipster-encumbered/tmp/tmpms6itpep/catalog.attrs
because of insufficient permissions. Please try the command again as a privileged user.


pkgrecv: Cached files were preserved in the following directory:
	/var/tmp/pkgrecv-2elagyac
Use pkgrecv -c to resume the interrupted download.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants