| === pushkarnk1 is now known as pushkarnk | ||
| === pushkarnk1 is now known as pushkarnk | ||
| liushuyu | ubuntu-archive: Can any archive admin take a look at https://code.launchpad.net/~liushuyu-011/ubuntu-seeds/+git/i386/+merge/471489? Thanks! | 08:35 |
|---|---|---|
| === pushkarnk1 is now known as pushkarnk | ||
| tillkamppeter | ghostscript 10.03.1 is not migrating into Oracular release. It is hanging on the tests of the Universe package ocrmypdf, See | 09:56 |
| tillkamppeter | https://ubuntu-archive-team.ubuntu.com/proposed-migration/oracular/update_excuses.html | 09:56 |
| tillkamppeter | The tests only fail on armhf and s390x. The latter seems nothing to have to do with Ghostscript. | 09:56 |
| tillkamppeter | The one on armhf Has the following Ghostscript command line failing: | 09:56 |
| tillkamppeter | gs -dBATCH -dNOPAUSE -dSAFER -dCompatibilityLevel=1.6 -sDEVICE=pdfwrite -dAutoRotatePages=/None -sColorConversionStrategy=LeaveColorUnchanged -dPDFSTOPONERROR -dAutoFilterColorImages=true -dAutoFilterGrayImages=true -dJPEGQ=95 -dPDFA=2 -dPDFACompatibilityPolicy=1 -o - -sstdout=%stderr tests/resources/blank.pdf > out.pdf | 09:57 |
| tillkamppeter | and this one works perfectly well on amd64. | 09:57 |
| tillkamppeter | armhf gives following in log: | 09:57 |
| tillkamppeter | WARNING ocrmypdf._validation:_validation.py:39 You are running OCRmyPDF in a 32-bit (x86) Python interpreter. This is not supported. 32-bit does not have enough address space to process large files. Please use a 64-bit (x86-64) version of Python. | 09:57 |
| tillkamppeter | 744s ERROR ocrmypdf._exec.ghostscript:ghostscript.py:297 GPL Ghostscript 10.03.1 (2024-05-02) | 09:57 |
| tillkamppeter | 744s Copyright (C) 2024 Artifex Software, Inc. All rights reserved. | 09:57 |
| tillkamppeter | 744s This software is supplied under the GNU AGPLv3 and comes with NO WARRANTY: | 09:57 |
| tillkamppeter | 744s see the file COPYING for details. | 09:57 |
| tillkamppeter | 744s Processing pages 1 through 1. | 09:58 |
| tillkamppeter | 744s Page 1 | 09:58 |
| tillkamppeter | 744s Error: /typecheck in --runpdf-- | 09:58 |
| tillkamppeter | 744s Operand stack: --nostringval-- --nostringval-- 0 | 09:58 |
| tillkamppeter | 744s Execution stack: %interp_exit .runexec2 --nostringval-- runpdf --nostringval-- 2 %stopped_push --nostringval-- runpdf runpdf false 1 %stopped_push 1949 1 3 %oparray_pop 1948 1 3 %oparray_pop 1933 1 3 %oparray_pop 1934 1 3 %oparray_pop runpdf runpdf runpdf 2 1 1 runpdf %for_pos_int_continue runpdf | 09:58 |
| tillkamppeter | 744s Dictionary stack: --dict:756/1123(ro)(G)-- --dict:0/20(G)-- --dict:86/200(L)-- --dict:7/10(L)-- | 09:58 |
| tillkamppeter | 744s Current allocation mode is local | 09:58 |
| tillkamppeter | 744s GPL Ghostscript 10.03.1: Unrecoverable error, exit code 1 | 09:58 |
| tillkamppeter | 744s | 09:58 |
| tillkamppeter | 744s ERROR ocrmypdf._pipelines._common:_common.py:261 SubprocessOutputError: Ghostscript PDF/A rendering failed | 09:58 |
| tillkamppeter | 744s ____________________________ test_deskew_blank_page ____________________________ | 09:58 |
| tillkamppeter | 744s | 09:58 |
| tillkamppeter | Anyone has any idea here or could help me out with testing on an armhf system? | 09:58 |
| ginggs | tillkamppeter: if the failures have nothing to do with ghostscript, then trigger migration-reference/0 tests | 09:59 |
| ginggs | if they fail, then ghostscript will migrate | 09:59 |
| tillkamppeter | Thanks, @ginggs, what does "trigger migration-reference/0 tests" mean? What do I exactly have to do for that? | 10:22 |
| Skia | tillkamppeter: https://wiki.ubuntu.com/ProposedMigration#How_to_run_autopkgtests_of_a_package_against_the_version_in_the_release_pocket | 10:24 |
| Skia | I've already triggered those tests myself, you can see them running here: https://autopkgtest.ubuntu.com/packages/ocrmypdf | 10:28 |
| tillkamppeter | Skia, thanks for the link to the appropriate docs, and that you have already kicked off the needed tests explains that for me the URL (I hope this is the correct one) https://autopkgtest.ubuntu.com/request.cgi?release=oracular&arch=armhf&package=ocrmypdf&trigger=migration-reference/0 gave "Test already runnuing". | 10:33 |
| Skia | yes exactly, the URL is correct and has a mechanism to avoid running duplicate tests, that you hit | 10:34 |
| -queuebot:#ubuntu-release- Unapproved: accepted dotnet8 [source] (noble-proposed) [8.0.108-8.0.8-0ubuntu1~24.04.2] | 10:52 | |
| tillkamppeter | @Skia Now all tests in the "Oracular" column are marked as passed and no logs shown any more. Does it mean that ghostscript will migrate now? | 10:52 |
| jbicha | tillkamppeter: the migration-reference/0 test passes which suggests that something did change in the new ghostscript version that is causing the ocrmypdf autopkgtest to fail on s390x | 10:56 |
| jbicha | https://autopkgtest.ubuntu.com/packages/ocrmypdf/oracular/s390x | 10:56 |
| -queuebot:#ubuntu-release- New: accepted mrcal [source] (noble-proposed) [2.4.1-1ubuntu0.24.04.1] | 11:20 | |
| -queuebot:#ubuntu-release- New binary: mrcal [amd64] (noble-proposed/none) [2.4.1-1ubuntu0.24.04.1] (no packageset) | 11:25 | |
| -queuebot:#ubuntu-release- New binary: mrcal [armhf] (noble-proposed/universe) [2.4.1-1ubuntu0.24.04.1] (no packageset) | 11:27 | |
| -queuebot:#ubuntu-release- New binary: mrcal [ppc64el] (noble-proposed/universe) [2.4.1-1ubuntu0.24.04.1] (no packageset) | 11:28 | |
| -queuebot:#ubuntu-release- New binary: mrcal [s390x] (noble-proposed/universe) [2.4.1-1ubuntu0.24.04.1] (no packageset) | 11:28 | |
| -queuebot:#ubuntu-release- New binary: mrcal [arm64] (noble-proposed/universe) [2.4.1-1ubuntu0.24.04.1] (no packageset) | 11:32 | |
| -queuebot:#ubuntu-release- Unapproved: landscape-client (jammy-proposed/main) [23.02-0ubuntu1~22.04.3 => 23.02-0ubuntu1~22.04.4] (ubuntu-server) | 11:58 | |
| -queuebot:#ubuntu-release- Unapproved: landscape-client (focal-proposed/main) [23.02-0ubuntu1~20.04.3 => 23.02-0ubuntu1~20.04.4] (ubuntu-server) | 11:59 | |
| -queuebot:#ubuntu-release- Unapproved: landscape-client (noble-proposed/main) [24.02-0ubuntu5.1 => 24.02-0ubuntu5.2] (ubuntu-server) | 11:59 | |
| tillkamppeter | @jbicha, @Skia what does this test exactly mean? Does the "migration-reference/0" one take Ghostscript from the release pocket or ocrmypdf from the release pocket? | 12:13 |
| Skia | everything: it's just running the autopkgtest suite against the release pocket to kind of "reset the baseline" | 12:14 |
| -queuebot:#ubuntu-release- New binary: mrcal [riscv64] (noble-proposed/universe) [2.4.1-1ubuntu0.24.04.1] (no packageset) | 12:15 | |
| tillkamppeter | @Skia @jbicha I think we should somehow skip this test. First, the upstream developers of ocrmypdf explicitly warn to not use their software on 32-bit systems: | 12:28 |
| tillkamppeter | WARNING ocrmypdf._validation:_validation.py:39 You are running OCRmyPDF in a 32-bit (x86) Python interpreter. This is not supported. 32-bit does not have enough address space to process large files. Please use a 64-bit (x86-64) version of Python. | 12:28 |
| tillkamppeter | So then why should we force such a program into 32-bit systems? | 12:29 |
| jbicha | s390x is a 64-bit architecture https://wiki.debian.org/ArchitectureSpecificsMemo | 12:29 |
| tillkamppeter | Also, the Ghostscript upstream developers told me that the Ghostscript command line seems to be for creating PDF/A, but actually the command line does not produce actual PDF/A. And it even streams out the resulting PDF file to stdout while the output device could be seeking when writing the file. | 12:33 |
| -queuebot:#ubuntu-release- Unapproved: accepted shim [sync] (jammy-proposed) [15.8-0ubuntu1] | 12:36 | |
| tillkamppeter | For me ocrmypdf makes the impression of that we should not stop the migration of Ghostscript on it. | 12:36 |
| -queuebot:#ubuntu-release- Unapproved: accepted shim-signed [sync] (jammy-proposed) [1.51.4] | 12:52 | |
| -queuebot:#ubuntu-release- Unapproved: accepted shim [sync] (focal-proposed) [15.8-0ubuntu1] | 13:15 | |
| -queuebot:#ubuntu-release- Unapproved: accepted shim-signed [sync] (focal-proposed) [1.40.10] | 13:16 | |
| tillkamppeter | @Skia, @jbicha WDYT? Should we introduce some exception to stop ghostscript being stopped migrating by ocrmypdf? Also, the log for s390x makes for me the impression that the regression is not caused by Ghostscript, perhaps by some other dependency of ocrmypdf, some Python library or so. | 13:40 |
| jbicha | ocrmypdf has no reverse dependencies. I think it would be less work to hint ghostscript through than to try to exclude ocrmypdf on s390x, and not really worth the effort to fix it properly | 13:42 |
| jbicha | but you need the Ubuntu Release team for that hinting | 13:42 |
| Skia | the hints are here, if you want to have a look and prepare an MP: https://git.launchpad.net/~ubuntu-release/britney/+git/hints-ubuntu/tree/ubuntu-release | 13:43 |
| tillkamppeter | @jbicha, I assume that the release team is on this channel. | 13:43 |
| Skia | yes. Probably pinging g_inggs will help you there :-) | 13:43 |
| tillkamppeter | To the release team, especially @ginggs, WDYT about making an exception for ghostscript so that it migrates despite the current ocrmypdf test failures? | 13:45 |
| ginggs | tillkamppeter: i'm not keen. according to the failing s390x log, the only packages coming from -proposed at from src:ghostscript | 13:47 |
| ginggs | so i'm not sure how you reach the conclusion that it's some other dependency of ocrmypdf | 13:48 |
| ginggs | as for armhf, you could request removal of ocrmypdf on the architecture, based on "You are running OCRmyPDF in a 32-bit (x86) Python interpreter" | 13:49 |
| ginggs | but as far as I can see, that is only a warning | 13:50 |
| ginggs | and ocrmypdf is not failing in oracular with the old ghostscript | 13:50 |
| tillkamppeter | @ginggs, so you prefer to give priority to a working ocrmypdf and if we get the 10.03.1 Ghostscript not working with it to stay with the old version in Oracular? | 13:55 |
| tillkamppeter | @ginggs, I am discussing this problem also with the upstream developers of Ghostscript and they tell that the Ghostscript command line built by ocrmypdf is not correct, not producing the intended PDF/A and also tell about that message discouraging the use of ocrmypdf on 32-bit. | 13:58 |
| ginggs | tillkamppeter: i think the failures should be investigated. i don't see that you've provided a reason why we should hint | 13:58 |
| ginggs | other then jbicha saying it's less work | 13:59 |
| === pushkarnk1 is now known as pushkarnk | ||
| -queuebot:#ubuntu-release- Unapproved: ubuntustudio-default-settings (noble-proposed/universe) [24.04.26.2 => 24.04.26.3] (ubuntustudio) | 14:28 | |
| * Eickmeyer can never find sil2100 | 14:28 | |
| Eickmeyer | ubuntu-sru: Fix for noble build failures ^ | 14:28 |
| Eickmeyer | ubuntu-sru: In other words, rather urgent | 14:32 |
| vorlon | jbicha: what's the chain that uses rustc-1.80 on i386? | 15:59 |
| Eickmeyer | ubuntu-sru: Urgent fix for ubuntustudio-default-settings for noble in queue to fix ISO build failures. Please accept ASAP for verification. Thanks! | 16:29 |
| jbicha | vorlon: gcc-14, linux, librsvg, & gstreamer1.0 all have Build-Depends: rustc. rust-defaults now provides rustc (previously it was source: rustc). rust-defaults now depends on rustc-1.80 | 16:34 |
| -queuebot:#ubuntu-release- Unapproved: ceph (noble-proposed/main) [19.2.0~git20240301.4c76c50-0ubuntu6.1 => 19.2.0~is.really.19.1.0-0ubuntu0.24.04.1] (ubuntu-server) | 16:34 | |
| -queuebot:#ubuntu-release- Unapproved: rabbitmq-server (noble-proposed/main) [3.12.1-1ubuntu1 => 3.12.1-1ubuntu1.1] (ubuntu-server) | 16:34 | |
| -queuebot:#ubuntu-release- Unapproved: open-vm-tools (noble-proposed/main) [2:12.3.5-5build3 => 2:12.4.5-1ubuntu0.24.04.1] (ubuntu-cloud, ubuntu-desktop, ubuntu-server) | 16:34 | |
| jbicha | bootstrapping may take a few steps. ordinarily the previous version of rustc is used to compile rustc (so rustc 1.79 compiles rustc 1.80) | 16:35 |
| -queuebot:#ubuntu-release- Unapproved: gnome-initial-setup (noble-proposed/main) [46.2-1ubuntu0.24.04.1 => 46.3-1ubuntu3~24.04.1] (ubuntu-desktop) | 16:35 | |
| -queuebot:#ubuntu-release- Unapproved: svn2git (noble-proposed/universe) [2.4.0-3 => 2.4.0-3ubuntu0.1] (no packageset) | 16:35 | |
| jbicha | so one option is to enable i386 builds for rustc-1.76, .77, .78, .79 and then use a ppa to build those up to .80, copy to Ubuntu, and then rebuild .80 using .80 | 16:36 |
| enr0n | ubuntu-sru: please consider releasing systemd to noble-updates. Note that bug 2076340 (which shows as not verified on the pending-sru report) is just a no-change rebuild tracking bug. | 18:01 |
| -ubottu:#ubuntu-release- Bug 2076340 in s390-tools-signed (Ubuntu Noble) "SRU: no-change rebuild to pick up changed build flags on ppc64el and s390x" [Undecided, New] https://launchpad.net/bugs/2076340 | 18:02 | |
| mfo | Eickmeyer, ack for your sru vanguard ping above (just found it via irclogs). | 18:07 |
| rbasak | enr0n: I suggest you mark that bug verification-done-noble then with an explanation. But, don't you need to follow the plan "We will validate picking up the changed build flags by inspecting the log files on ppc64el and s390x" as documented in that bug? | 18:30 |
| rbasak | Maybe you shouldn't flip the tag without performing and reporting on those steps for all the packages in Noble, but if we need to get systemd landed now, presumably it should at least be documented for Noble? | 18:31 |
| enr0n | rbasak: yeah it didn't seem right to me to set that tag given all the other packages associated with that bug. I did check but I forgot to add my comment on the bug, but I have done that now | 18:38 |
| mfo | Eickmeyer, hey o/ there's 2 minor things to address in that upload; i provided feedback and rejected it. Please just ping me if/once you re-upload, and I'm happy to resume on it. | 18:39 |
| -queuebot:#ubuntu-release- Unapproved: rejected ubuntustudio-default-settings [source] (noble-proposed) [24.04.26.3] | 18:39 | |
| -queuebot:#ubuntu-release- New binary: uwsgi [amd64] (oracular-proposed/universe) [2.0.26-2ubuntu2] (no packageset) | 18:44 | |
| -queuebot:#ubuntu-release- New binary: uwsgi [s390x] (oracular-proposed/universe) [2.0.26-2ubuntu2] (no packageset) | 18:44 | |
| -queuebot:#ubuntu-release- New binary: uwsgi [arm64] (oracular-proposed/universe) [2.0.26-2ubuntu2] (no packageset) | 18:45 | |
| -queuebot:#ubuntu-release- New binary: uwsgi [armhf] (oracular-proposed/universe) [2.0.26-2ubuntu2] (no packageset) | 18:45 | |
| -queuebot:#ubuntu-release- New binary: uwsgi [ppc64el] (oracular-proposed/universe) [2.0.26-2ubuntu2] (no packageset) | 18:46 | |
| vorlon | jbicha: ack thanks | 19:09 |
| vorlon | jbicha: (though I don't see a build-dep from gcc-14 to rustc, which sounded unusual to me) | 19:11 |
| jbicha | vorlon: oh I guess it's cargo but still built from our unified rustc source package | 19:12 |
| jbicha | https://launchpad.net/ubuntu/+source/gcc-14/14.2.0-3ubuntu1 | 19:12 |
| vorlon | jbicha: ah interesting | 19:12 |
| jbicha | there is a gccrs thing as an alternative to rustc, I don't know much about it | 19:12 |
| -queuebot:#ubuntu-release- Unapproved: rejected mdadm [source] (noble-proposed) [4.3-1ubuntu3] | 20:35 | |
| -queuebot:#ubuntu-release- New binary: uwsgi [riscv64] (oracular-proposed/universe) [2.0.26-2ubuntu2] (no packageset) | 20:37 | |
| -queuebot:#ubuntu-release- Unapproved: rejected s390-tools-signed [source] (noble-proposed) [2.31.0-0ubuntu5.1] | 21:06 | |
| tillkamppeter | @vorlon, did you see my posts about ghostscript not migrating due to the autopkgtests on ocrmypdf on armhf and s390x failing? What is the best to do about that? | 21:13 |
| tillkamppeter | @vorlon I nver heard about ocrmypdf before and now it blocks migration of ghostscript, but there are regresions only on armhf and s390x, and the log for s390x makes the impression that it is not caused by Ghostscript. | 21:17 |
| === dbungert1 is now known as dbungert | ||
| -queuebot:#ubuntu-release- Unapproved: accepted rtorrent [source] (noble-proposed) [0.9.8-1ubuntu0.1] | 21:29 | |
| Eickmeyer | mfo: ack, reuploaded | 21:37 |
| -queuebot:#ubuntu-release- Unapproved: ubuntustudio-default-settings (noble-proposed/universe) [24.04.26.2 => 24.04.26.3] (ubuntustudio) | 21:38 | |
| Eickmeyer | (sorry, was out running errands, but yes, it's needed to fix iso builds) | 21:38 |
| mfo | Eickmeyer, ack, thanks! | 21:45 |
| -queuebot:#ubuntu-release- Unapproved: accepted ubuntustudio-default-settings [source] (noble-proposed) [24.04.26.3] | 22:18 | |
| Eickmeyer | ubuntu-release: How often are the cron jobs for the manual triggers for the iso rebuilds done? Because I did a rebuild request for Ubuntu Studio Oracular and got crickets chirping in return. | 22:47 |
| -queuebot:#ubuntu-release- Unapproved: rejected python-docker [source] (noble-proposed) [5.0.3-1ubuntu1.1] | 22:50 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!