Bug 30167 - The i586 version of 0ad was not updated with the x86_64 version the last two times it was updated
Summary: The i586 version of 0ad was not updated with the x86_64 version the last two ...
Status: RESOLVED FIXED
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: 8
Hardware: i586 Linux
Priority: Normal normal
Target Milestone: ---
Assignee: QA Team
QA Contact:
URL:
Whiteboard: MGA8-64-OK MGA8-32-OK
Keywords: advisory, validated_update
Depends on:
Blocks:
 
Reported: 2022-03-13 20:54 CET by Thomas Andrews
Modified: 2022-03-21 21:19 CET (History)
3 users (show)

See Also:
Source RPM:
CVE:
Status comment:


Attachments
screenshot of 0ad message window (96.05 KB, image/jpeg)
2022-03-16 00:12 CET, Thomas Andrews
Details

Description Thomas Andrews 2022-03-13 20:54:20 CET
Description of problem: When the 0ad game was updated from version 0.0.23b to 0.0.25 (Bug 29091), the i586 arch was not included in the build. Partly my fault that it went through, as I only tested the 64-bit version. Had I tried the 32-bit version, I would have caught it. 

The situation continued when 0ad was rebuilt to version 0.0.25b for the recent wxgtk update (Bug 29848). 

0ad-data, a noarch rpm, was included both times, and is in the repositories of both arches.

0ad 0.0.25b needs to be rebuilt, this time including the 32-bit version.
Comment 1 Thomas Backlund 2022-03-13 21:17:05 CET
Seems akien disabled i586 here with ExclusiveArch:

http://svnweb.mageia.org/packages/updates/8/0ad/current/SPECS/0ad.spec?r1=1688338&r2=1730041

interestingly its still built on 32bit arm

Assignee: bugsquad => rverschelde

Rémi Verschelde 2022-03-13 22:34:52 CET

Status: NEW => ASSIGNED

Comment 2 Rémi Verschelde 2022-03-13 22:38:50 CET
That looks like a mistake, this was a partial sync with Fedora but the spec at that time does have `%{ix86}`: https://src.fedoraproject.org/rpms/0ad/blob/880799858dc1aa1c5784fc06d982ba8df77c9519/f/0ad.spec

I don't recall why I would have disabled it on i586, and I wouldn't have done so without a comment, so this must be a PEBKAC. Will fix it tomorrow.
Comment 3 Rémi Verschelde 2022-03-15 17:36:49 CET
Advisory:
=========

Updated 0ad package restores i586 support

  A previous update mistakenly disabled the i586 build for 0ad. This update
  reintroduces the i586 package.

SRPM in core/updates_testing:
=============================

0ad-0.0.25b-1.mga8

RPM in core/updates_testing:
============================

0ad-0.0.25b-1.mga8


Note that there's no update for 0ad-data which is noarch and should therefore already be available on i586 (which also means that current i586 0.0.23b package would fail to run most likely, as the updated data isn't compatible).

Assignee: rverschelde => qa-bugs
CC: (none) => rverschelde

Comment 4 Rémi Verschelde 2022-03-15 17:38:07 CET
Note for testing: since it's a 3D game, it's likely that it might not run well in a virtual machine, unless you manage to have one with 3D acceleration using qemu-kvm etc.

So don't be overly worried if you do test in a VM and it complains about e.g. not finding a suitable version of OpenGL.
Comment 5 Thomas Andrews 2022-03-15 23:30:00 CET
Not just in a VM, Remi. 

I tried testing it first on real 32-bit hardware: Foolishness, my Dell Inspiron 5100, P4, Radeon RV200 graphics. No installation issues, but unfortunately the video, only capable of OpenGL 1.6, is not up to the task. The video card can be replaced, but finding something that will work is not easy, if even still possible.

Trying a 32-bit install on a 64-bit machine now...
Comment 6 Thomas Andrews 2022-03-16 00:10:30 CET
Hmmm. AMD Phenom II X4 910, Radeon HD 8490 graphics, 32-bit Plasma system using the server kernel.

No installation issues. Ran 0ad from the Plasma menu and was greeted with a ")ad Message window. (See attached screen shot. It wouldn't let me copy-and-paste) I saw the same window on Foolishness, but I thought it was because of the video card.

I clicked on "continue, and that took me to a "modifications" screen. I couldn't figure out any way to load and start the game. Something is not right here. The 64-bit version has always been easier to start.
Comment 7 Thomas Andrews 2022-03-16 00:12:17 CET
Created attachment 13179 [details]
screenshot of 0ad message window
Comment 8 Rémi Verschelde 2022-03-16 09:27:31 CET
I can reproduce the crash from comment 7 in a Mageia 8 i586 VM, I'll have a look. Thanks!

Assignee: qa-bugs => rverschelde

Comment 9 Rémi Verschelde 2022-03-16 11:17:12 CET
This was indeed an upstream bug, the 32-bit version couldn't handle large files (over 2.0 GiB, like the main data pack in /usr/share/games/0ad/mods/public/public.zip) so it was crashing on start (and then showing the mod management tool instead of the main game it couldn't load).

I backported the upstream fix, and tested the Cauldron i586 version on my x86_64 system and it worked fine: https://code.wildfiregames.com/D4413

Advisory:
=========

Updated 0ad package restores i586 support

  A previous update mistakenly disabled the i586 build for 0ad. This update
  reintroduces the i586 package, and fixes a bug in loading large files like
  the default 'public' game mod.

SRPM in core/updates_testing:
=============================

0ad-0.0.25b-1.1.mga8

RPM in core/updates_testing:
============================

0ad-0.0.25b-1.1.mga8

Assignee: rverschelde => qa-bugs

Comment 10 Thomas Andrews 2022-03-16 16:33:46 CET
Foolishness still doesn't like it because of the GPU. It might be an idea to warn i586 users that a certain level of hardware is needed before they go through the lengthy process of downloading and installing it, but I don't know how that can be accomplished.

It works fine on the install from Comment 6. I also tried the new 64-bit version with a different install on the same hardware, and it too worked OK.

Giving this OKs for both arches. Validating. Advisory in Comment 9.

Whiteboard: (none) => MGA8-64-OK MGA8-32-OK
CC: (none) => sysadmin-bugs
Keywords: (none) => validated_update

Dave Hodgins 2022-03-21 02:20:16 CET

CC: (none) => davidwhodgins
Keywords: (none) => advisory

Comment 11 Mageia Robot 2022-03-21 21:19:40 CET
An update for this issue has been pushed to the Mageia Updates repository.

https://advisories.mageia.org/MGAA-2022-0040.html

Resolution: (none) => FIXED
Status: ASSIGNED => RESOLVED


Note You need to log in before you can comment on or make changes to this bug.