<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 TRANSITIONAL//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; CHARSET=UTF-8">
<META NAME="GENERATOR" CONTENT="GtkHTML/3.32.2">
</HEAD>
<BODY>
Thanks Phil<BR>
<BR>
The crash occurs each time I just quit nvidia-setting, so even if I don't change anything in it (no need to change a specific thing).<BR>
In the attachment I provide you the export of the current nvidia settings (done via "save current configuration" function of the utility), but as I said, it seems that any configuration leads to segfault after quitting it.<BR>
By the way, one does not notice it through the system behavior, but it is always present in the log.<BR>
<BR>
Best regards<BR>
Bojan<BR>
<BR>
<BLOCKQUOTE TYPE=CITE>
<PRE>
----------------------------------------------------------------------
Message: 1
Date: Sun, 06 Jul 2014 08:42:18 +0100
From: Phil Perry <<A HREF="mailto:phil@elrepo.org">phil@elrepo.org</A>>
To: <A HREF="mailto:elrepo@lists.elrepo.org">elrepo@lists.elrepo.org</A>
Subject: Re: [elrepo] nvidia-setting segfault on CentOS 6.5 x86_64
Message-ID: <<A HREF="mailto:53B8FDDA.6060205@elrepo.org">53B8FDDA.6060205@elrepo.org</A>>
Content-Type: text/plain; charset=ISO-8859-1
Hi Bojan,
Welcome to the list.
On 05/07/14 11:56, Bojan Radovic wrote:
> Hello
>
> I am glad to join the ELRepo mailing list.
> I use a desktop PC with an nVidia GPU and CentOS-6.5 x86_64.
> The GPU model is : (output of lspci | grep VGA)
>
> 01:00.0 VGA compatible controller: NVIDIA Corporation G86 [Quadro NVS
> 290] (rev a1)
>
> So I have installed the NVIDIA driver kernel module from ELREPO, like
> explained here : <A HREF="http://elrepo.org/tiki/kmod-nvidia">http://elrepo.org/tiki/kmod-nvidia</A>
>
> Now the driver seems to work more or less well, but each time I use
> nvidia-setting utility, after quitting it there is a segfault series of
> error messages in /var/log/messages :
>
> Jul 5 12:47:38 Linux kernel: nvidia-settings[8876]: segfault at
> 3918201510 ip 0000003918201510 sp 00007fffb1200518 error 14
> Jul 5 12:47:39 Linux abrtd: Directory 'ccpp-2014-07-05-12:47:38-8876'
> creation detected
> Jul 5 12:47:39 Linux abrt[8878]: Saved core dump of pid 8876
> (/usr/bin/nvidia-settings)
> to /var/spool/abrt/ccpp-2014-07-05-12:47:38-8876 (49676288 bytes)
> Jul 5 12:47:39 Linux abrtd: Package 'nvidia-x11-drv' isn't signed with
> proper key
> Jul 5 12:47:39 Linux abrtd: 'post-create' on
> '/var/spool/abrt/ccpp-2014-07-05-12:47:38-8876' exited with 1
> Jul 5 12:47:39 Linux abrtd: Deleting problem directory
> '/var/spool/abrt/ccpp-2014-07-05-12:47:38-8876'
> Jul 5 12:47:48 Linux kernel: nvidia-settings[8882]: segfault at
> 3918201510 ip 0000003918201510 sp 00007fff04e7d9d8 error 14
> Jul 5 12:47:49 Linux abrt[8892]: Not saving repeating crash in
> '/usr/bin/nvidia-settings'
>
>
I am unable to replicate this issue. Please could you describe exactly
what you changed in nvidia-settings to trigger the crash. Does it crash
every time, only if you change something, only if you change a specific
thing?
> To install the ELRepo for CentOS-6 I had followed these instructions :
> <A HREF="http://elrepo.org/tiki/tiki-index.php">http://elrepo.org/tiki/tiki-index.php</A>
> so I wonder when I see that the nvidia-x11-drv' isn't signed with proper
> key
>
ABRT is Red Hat's Automatic Bug Reporting Tool. On RHEL, ABRT would be
expecting the package to be signed by Red Hat but as nvidia-x11-drv
isn't part of the disto that is not the case.
I'm not sure why CentOS ship abrt, perhaps they have patched abrt so
automatic bug reports are sent to them, thus one might assume on CentOS
abrt expects to see a package signed by the CentOS key.
Either way, this warning is safe to ignore. What is of more concern is
the crash itself.
Regards,
Phil
</PRE>
</BLOCKQUOTE>
<BR>
</BODY>
</HTML>