Bug 9766

Summary: post script fails for firefox-gl during mga2 -> mga3 upgrade
Product: Mageia Reporter: claire robinson <eeeemail>
Component: RPM PackagesAssignee: Mageia Bug Squad <bugsquad>
Status: RESOLVED OLD QA Contact:
Severity: normal    
Priority: High CC: davidwhodgins, dmorganec, ennael1, mageia, thierry.vignaud
Version: 3   
Target Milestone: ---   
Hardware: i586   
OS: Linux   
Whiteboard: 3RC
Source RPM: mtink CVE:
Status comment:
Attachments: report.bug.xz

Description claire robinson 2013-04-16 20:46:46 CEST
See bug 9302, this appears to be the same.



Reproducible: 

Steps to Reproduce:
Comment 1 claire robinson 2013-04-16 20:48:05 CEST
Created attachment 3755 [details]
report.bug.xz
claire robinson 2013-04-16 20:48:45 CEST

CC: (none) => dmorganec, ennael1, mageia

Comment 2 claire robinson 2013-04-16 20:55:53 CEST
Release blocker as this can't be fixed on the isos after release.

Priority: Normal => release_blocker
Whiteboard: (none) => 3RC

Comment 3 claire robinson 2013-04-16 21:02:23 CEST
Basic details, as it's all repeating what is in bug 9302 but upgrading from mga2 to 3 gives an error.

ERROR: 'script' failed for firefox-gl-17.0.5-1.mga3.noarch.

When OK'd the upgrade completes. Possibly the same problem as
https://bugs.mageia.org/show_bug.cgi?id=9302#c19
Comment 4 Colin Guthrie 2013-04-16 21:09:56 CEST
Although the bug details look similar, these details only indicate a general packageing script error that can happen from time to time. In #9302, the cause in several packages was a common one and was solved with a common solution.

In this case, it seems like some kind of packaging bug in firefox-i10n, but as there are no scripts in these packages I do not know how the script could fail!

http://svnweb.mageia.org/packages/cauldron/firefox-l10n/current/SPECS/firefox-l10n.spec?revision=HEAD&view=markup

I seriously doubt this should be a release blocker. It is apparently only affecting one language and it's a non-fatal error it seems.
Comment 5 Thierry Vignaud 2013-04-18 18:08:16 CEST
Actually, the package name is off by one.
I didnât yet release a fixed urpmi for the off by one count when a package as %pretrans script (such as 'filesystem') (bug #9529)

CC: (none) => thierry.vignaud

Comment 6 Colin Guthrie 2013-04-18 18:26:39 CEST
Ha! Nice catch Thierry. Do we know what pkg it's actually complaining about?
Comment 7 Thierry Vignaud 2013-04-18 22:05:35 CEST
Obviously one of:
* trans: scheduling update of libXm2-0.95.2-8.mga3.i586
* trans: scheduling update of cups-drivers-pegg-0.23-10.mga3.i586
* trans: scheduling update of cups-drivers-2008-5.mga3.i586
* trans: scheduling update of cups-drivers-magicolor5440dl-1.2.1-14.mga3.i586 
* trans: scheduling update of libxp6-1.0.1-3.mga3.i586 
* trans: scheduling update of cups-drivers-magicolor5430dl-1.8.1-14.mga3.i586 
* trans: scheduling update of libgimp2.0_0-2.8.2-3.mga3.i586
* trans: scheduling update of mtink-1.0.16-9.mga3.i586

Probably one with &>/dev/null

And the only one with scripts is mtink...
The post script is covered by true but not the %postun.
I think it's the older one that caused the error.
Which will happen again on next upgrade.

BTW I think the error was not a one by off error.
We register id as data for callback but I think we got nothing for removed packages (rpmtsAddEraseElement() -> removePackage() -> rpmteNew()-> addTe() results in passing NULL as rpm key).

Thus undef id is interpreted as 0 hence the strange "firefox-gl" which doesn't appear anywhere else in those logs but must have been the first package of Core Release's synthesis when that error happened.
(rpm & yum register Headers as key)

BTW it also looks like I could get the type of script that failed.
Thierry Vignaud 2013-04-18 22:05:48 CEST

Source RPM: (none) => mtink

Comment 8 Dave Hodgins 2013-04-23 02:43:19 CEST
As per comment 3, that it doesn't block the upgrade, I'm
removing the release blocker tag.

While the pop-up will be annoying, I don't think it's
important enough to be a release blocker.

Priority: release_blocker => High
CC: (none) => davidwhodgins

Manuel Hiebel 2014-02-11 22:29:09 CET

Version: Cauldron => 3

Comment 9 Marja Van Waes 2015-03-31 16:04:40 CEST
Mageia 3 changed to end-of-life (EOL) status 4 months ago.
http://blog.mageia.org/en/2014/11/26/lets-say-goodbye-to-mageia-3/ 

Mageia 3 is no longer maintained, which means that it will not receive any further security or bug fix updates. As a result we are closing this bug.

If you can reproduce this bug against a currently maintained version of Mageia
please feel free to click on "Version" change it against that version of Mageia
and reopen this bug.

Thank you for reporting this bug and we are sorry it could not be fixed.

--
The Mageia Bugsquad

Status: NEW => RESOLVED
Resolution: (none) => OLD