FC5 Kernel Issues

[Kevin DeKorte]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On my FC5 machine I was recently upgraded to kernel 2.6.17-1.2187_FC5.
When running this kernel (with 3 tainted modules) I had 5 lockups of X
within about 2 hrs. (I could ssh to the box and reboot so the box was
not 100% dead). Rolling back to 2174 (with the same set of modules and
versions) has resulted in no lockups on the same machine over a longer
period of time(I have been running with this config for nearly two
weeks, and there are no messages in /var/log/messages about anything
crashing).

Modules:	fglrx, madwifi and vmware
		fglrx and madwifi are packaged by livna
		vmware is running with the any-any-104 patches and built 		locally.
Equipment:	Thinkpad T60p (core duo, 2GB RAM)

Has anyone else, had any issue with X becoming non-responsive with 2187.
I know I'm treading on thin ice due to the modules, but since 2174 works
well I would expect 2187 to work as well.

Thanks,

Kevin


- --
Get my public GnuPG key from
http://keyserver.veridis.com:11371/export?id=-1509369771935935023
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (GNU/Linux)
> Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org
>
> iD8DBQFFDw0K6w2kMH0L1dERAtpyAKCNpoDnGhi0PCEdasLMq690cce0OwCfcgnc
> QGGztQIIcMbIS2cy/V3OFTY=
> =9QqO
> -----END PGP SIGNATURE-----

[David Nielsen]

man, 18 09 2006 kl. 15:18 -0600, skrev Kevin DeKorte:
>  -----BEGIN PGP SIGNED MESSAGE-----
>  Hash: SHA1
>  
>  On my FC5 machine I was recently upgraded to kernel 2.6.17-1.2187_FC5.
>  When running this kernel (with 3 tainted modules) I had 5 lockups of X
>  within about 2 hrs. (I could ssh to the box and reboot so the box was
>  not 100% dead). Rolling back to 2174 (with the same set of modules and
>  versions) has resulted in no lockups on the same machine over a longer
>  period of time(I have been running with this config for nearly two
>  weeks, and there are no messages in /var/log/messages about anything
>  crashing).
>  
>  Modules:	fglrx, madwifi and vmware
>  		fglrx and madwifi are packaged by livna
>  		vmware is running with the any-any-104 patches and built 		locally.
>  Equipment:	Thinkpad T60p (core duo, 2GB RAM)
>  
>  Has anyone else, had any issue with X becoming non-responsive with 2187.
>  I know I'm treading on thin ice due to the modules, but since 2174 works
>  well I would expect 2187 to work as well. 
Does it work if you don't load the modules in question?

- David

[Adam Pribyl]

On Mon, 18 Sep 2006, Kevin DeKorte wrote:

>  -----BEGIN PGP SIGNED MESSAGE-----
>  Hash: SHA1
>
>  On my FC5 machine I was recently upgraded to kernel 2.6.17-1.2187_FC5.
>  When running this kernel (with 3 tainted modules) I had 5 lockups of X
>  within about 2 hrs. (I could ssh to the box and reboot so the box was
>  not 100% dead). Rolling back to 2174 (with the same set of modules and
>  versions) has resulted in no lockups on the same machine over a longer
>  period of time(I have been running with this config for nearly two
>  weeks, and there are no messages in /var/log/messages about anything
>  crashing).
>
>  Modules:	fglrx, madwifi and vmware
>  		fglrx and madwifi are packaged by livna
>  		vmware is running with the any-any-104 patches and built 		locally.
>  Equipment:	Thinkpad T60p (core duo, 2GB RAM)
>
>  Has anyone else, had any issue with X becoming non-responsive with 2187.
>  I know I'm treading on thin ice due to the modules, but since 2174 works
>  well I would expect 2187 to work as well.
>
>  Thanks,
>
>  Kevin 
I had similar problems alredy with 2174 but I identified that switching 
back to older udev solves the problem.
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=204073

I do not have any "tained" modules. May be different issue, but could you 
try to revert udev?

Adam Pribyl

[Kevin DeKorte]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Adam Pribyl wrote:

>  I had similar problems alredy with 2174 but I identified that
> switching
>  back to older udev solves the problem.
>  https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=204073
>  
>  I do not have any "tained" modules. May be different issue, but could
>  you try to revert udev?
>  
>  Adam Pribyl
>   
It is not a udev issue for me... all I have to do is switch kernels and
the issue goes away. I can remove the kernel modules, but then actually
being able to use my machine, goes nowhere. I guess fglrx is probably
the primary concern since it is an X lockup. Perhaps I'll try that.

Kevin


- --
Get my public GnuPG key from
http://keyserver.veridis.com:11371/export?id=-1509369771935935023
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (GNU/Linux)
> Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org
>
> iD8DBQFFEA/G6w2kMH0L1dERAkWPAKCENooS5Gb4Yuevt4EH6x0D0xpBfACeNglE
> MwR37+LRWcV6w/SBriQIGqA=
> =lxxR
> -----END PGP SIGNATURE-----

[Horst H. von Brand]

Kevin DeKorte <kdekorte(a)gmail.com&gt; wrote:
>  Adam Pribyl wrote:
>  
>  > I had similar problems alredy with 2174 but I identified that switching
>  > back to older udev solves the problem.
>  > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=204073
>  > 
>  > I do not have any "tained" modules. May be different issue, but could
>  > you try to revert udev?
>  > 
>  > Adam Pribyl 
>  It is not a udev issue for me... all I have to do is switch kernels
> and
>  the issue goes away. I can remove the kernel modules, but then actually
>  being able to use my machine, goes nowhere. I guess fglrx is probably
>  the primary concern since it is an X lockup. Perhaps I'll try that. 
Perhaps a brokem mkinitrd? The old kernel has a working initrd, the newer
one doesn't...
-- 
Dr. Horst H. von Brand                   User #22616 counter.li.org
Departamento de Informatica                     Fono: [phone]
Universidad Tecnica Federico Santa Maria              [phone]
Casilla 110-V, Valparaiso, Chile                Fax:  [phone]

[Paul Johnson]

On 9/19/06, Horst H. von Brand <vonbrand(a)inf.utfsm.cl&gt; wrote:
>  Kevin DeKorte <kdekorte(a)gmail.com&gt; wrote:
>  > Adam Pribyl wrote:
>  >
>  > > I had similar problems alredy with 2174 but I identified that switching
>  > > back to older udev solves the problem.
>  > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=204073
>  > >
>  > > I do not have any "tained" modules. May be different issue, but
> could
>  > > you try to revert udev?
>  > >
>  > > Adam Pribyl 

I've recently discovered these "freezes" are due to faulty hardware,
and the crash happens not particularly because the kernel is bad, but
because it exercises slightly different elements in the hardware.

On Dell GX270, a motherboard with bad capacitors will not always
crash, only when the power demand is over a threshold.  For me, the
crashes becan recently.

To check your motherboard, in case you have the same variety, look at
the capacitors (little columns on MOBO) and make sure the tops are
flat, not bulging or even swolen.

Also, run the memory test on the FC rescue disk.


-- 
Paul E. Johnson
Professor, Political Science
1541 Lilac Lane, Room 504
University of Kansas

[Kevin DeKorte]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Paul Johnson wrote:

>  
>  I've recently discovered these "freezes" are due to faulty hardware,
>  and the crash happens not particularly because the kernel is bad, but
>  because it exercises slightly different elements in the hardware.
>  
>  On Dell GX270, a motherboard with bad capacitors will not always
>  crash, only when the power demand is over a threshold.  For me, the
>  crashes becan recently.
>  
>  To check your motherboard, in case you have the same variety, look at
>  the capacitors (little columns on MOBO) and make sure the tops are
>  flat, not bulging or even swolen.
>  
>  Also, run the memory test on the FC rescue disk.
>  
>   
I don't think my problem is with the hardware as the machine is only two
weeks old. I orginally had problems with the memory, but IBM replaced
the modules with improved ones and I have not had a problem since then.
Also when the memory failed the entire machine would lock and I could
not ssh to it, which I can do with my recent lockups.

Kevin

- --
Get my public GnuPG key from
http://keyserver.veridis.com:11371/export?id=-1509369771935935023
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (GNU/Linux)
> Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org
>
> iD8DBQFFELeP6w2kMH0L1dERAuQ5AJ9ff5KzA2mgvFrt2eMdP1ANLyG+BwCeIb92
> IaojGjVTSnWs1RVqrI/9yaY=
> =CcYB
> -----END PGP SIGNATURE-----

[Genes MailLists]

On Monday 18 September 2006 17:18, Kevin DeKorte wrote:
>  On my FC5 machine I was recently upgraded to kernel
> 2.6.17-1.2187_FC5. 
>  Has anyone else, had any issue with X becoming non-responsive with
> 2187. 
   Similar troubles here.

   I had trouble immediately after installing 2187 - machine locked up within 
a few minutes (IBM T42 laptop, ATI M10 graphics). No tainted modules.

    No trouble with 2174 and no hangs at all with Dave Jones test kernel 3001 
(aside from a fail to lock message with test kernel - unrelated).

    This was repeatable - my session hung totally - a few minutes after 
starting to work on each of 3 tries before I gave up and reverted kernel. I 
could not change VC's either - power cycle was all I could do. 

  g/