Bug 23042 - Gigacage allocation problem causes drakconf and evolution to segfault on startup (using vm.overcommit_memory=2)
Summary: Gigacage allocation problem causes drakconf and evolution to segfault on star...
Status: RESOLVED OLD
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: 6
Hardware: All Linux
Priority: Normal major
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL:
Whiteboard:
Keywords: UPSTREAM
Depends on:
Blocks:
 
Reported: 2018-05-17 05:00 CEST by jim l
Modified: 2020-07-31 13:18 CEST (History)
6 users (show)

See Also:
Source RPM: webkit2
CVE:
Status comment:


Attachments

Description jim l 2018-05-17 05:00:30 CEST
Description of problem:
starting drakconf on my system leads to segfault and drakconf does not start.  

Version-Release number of selected component (if applicable):
13.16

How reproducible:

every time

This error is described on the forum:
https://forums.mageia.org/en/viewtopic.php?f=7&t=12354&sid=d1d4c8f4c6aa7cf894b8037c1abb89de

and appears to be related to the webkit bug described therein
https://bugs.webkit.org/show_bug.cgi?id=183329

My system is configured with  vm.overcommit_memory=2 which seems to trigger the bug.  The suggested workaround in the webkit bugzilla does not work.

I have also just discovered the identical behavior in evolution.
Comment 1 Marja Van Waes 2018-05-17 09:43:24 CEST
Thanks for the report, Jim

Do you know which of our webkit SRPMs this bug should be assigned to?

Keywords: (none) => UPSTREAM
Summary: drakconf segfaults on startup => drakconf and evolution segfault on startup (using vm.overcommit_memory=2)
See Also: (none) => https://bugs.webkit.org/show_bug.cgi?id=183329
CC: (none) => marja11

Comment 2 jim l 2018-05-17 19:06:59 CEST
(In reply to Marja Van Waes from comment #1)
> Thanks for the report, Jim
> 
> Do you know which of our webkit SRPMs this bug should be assigned to?

I have not the faintest idea.
Comment 3 Marja Van Waes 2018-05-19 09:21:01 CEST
I don't see any webkit in the drakconf spec file (nor in the drakxtools spec file), but in the evolution spec I do see:

BuildRequires:	pkgconfig(webkitgtk-3.0) >= 2.2.0
BuildRequires:	pkgconfig(webkit2gtk-4.0)
BuildRequires:	pkgconfig(webkit2gtk-web-extension-4.0)

But build requires aren't requires, are they?

I don't know whether this bug should be assigned to webkit or to webkit2 or both. Btw, why do both (in the specfiles) have webkitgtk.org as upstream? Or am I missreading bbecause the train I'm in (in real life) shakes too much?

CC'ing martinw and doktor5000

CC: (none) => doktor5000, mageia

Comment 4 Martin Whitaker 2018-05-19 11:55:50 CEST
(In reply to Marja Van Waes from comment #3)
> I don't know whether this bug should be assigned to webkit or to webkit2 or
> both.

'urpmq -d drakconf' lists (amongst many others)

lib64webkit2gtk4.0_37
webkit2

Same for evolution.

> Btw, why do both (in the specfiles) have webkitgtk.org as upstream? Or
> am I missreading bbecause the train I'm in (in real life) shakes too much?

webkitgtk supports both the original webkit and newer webkit2 APIs. Our webkit package uses an old version of webkitgtk (2.4.10), so I'm guessing there's some compatibility problem with newer versions.

Because our webkit package is pinned to an older version, I think we can safely assume this bug only applies to webkit2.
Comment 5 Marja Van Waes 2018-05-20 09:21:06 CEST
Thanks, Martin!

Assigning to webkit2, but not yet assigning to any packager, because this is an upstream issue.

CC'ing all packagers collectively, because there is no registered maintainer for webkit2

Also CC'ing ns80, because he touched this package a lot and might be one of the first to be aware when this gets fixed upstream.

This report should be assigned to a packager when an upstream fix is available.

Source RPM: drakconf-13.16-1.mga6.src.rpm => webkit2
CC: (none) => nicolas.salguero, pkg-bugs
Summary: drakconf and evolution segfault on startup (using vm.overcommit_memory=2) => Gigacage allocation problem causes drakconf and evolution to segfault on startup (using vm.overcommit_memory=2)

Comment 6 Lewis Smith 2020-01-30 16:39:22 CET
Jim
This bug was for Mageia 6. Can you say whether it still applies to Mageia 7?
Lewis Smith 2020-01-30 16:42:03 CET

CC: (none) => lewyssmith

Comment 7 Marja Van Waes 2020-07-31 13:18:13 CEST
(In reply to Lewis Smith from comment #6)
> Jim
> This bug was for Mageia 6. Can you say whether it still applies to Mageia 7?

No reply, so closing as OLD

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


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