Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!
Bug 43149 - Kernel Panic when executing vmware with $DISPLAY redirected to another machine.
Summary: Kernel Panic when executing vmware with $DISPLAY redirected to another machine.
Status: RESOLVED NEEDINFO
Alias: None
Product: Gentoo Linux
Classification: Unclassified
Component: [OLD] Core system (show other bugs)
Hardware: x86 Linux
: High critical (vote)
Assignee: Chris Gianelloni (RETIRED)
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2004-02-27 18:20 UTC by Vali "Blue" Dragnuta
Modified: 2004-04-09 11:43 UTC (History)
1 user (show)

See Also:
Package list:
Runtime testing required: ---


Attachments
kernel,hardware and my config. (bugdetails.txt,27.89 KB, text/plain)
2004-02-27 18:24 UTC, Vali "Blue" Dragnuta
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Vali "Blue" Dragnuta 2004-02-27 18:20:19 UTC
I am trying to run vmware with it's display redirected to another workstation in the network. (I do not have access to the server's console, and besides that it's 1mb videocard isn't very confortable).
After emerging vmware and compiling it's modules, I export DISPLAY=x.x.x.x:0 and
launch vmware. I am able to create a virtual machine, to create it's disks, but
when I start the virtual machine , the kernel on the server panics.
It does not matter if I launch vmware as an ordinary user - it still crashes the kernel.
My kernel is : 2.4.22-gentoo-r7
The crash also happens with  2.4.22-gentoo-r5


Reproducible: Always
Steps to Reproduce:
1.export DISPLAY=x.x.x.x:0
2.xterm (everything ok, xterm works just fine)
3.vmware
4.create/modify a virtual machine
5.start the virtual machine
6.wait 3..10 seconds.

Actual Results:  
Unable to handle kernel NULL pointer dereference at virtual address 000005e8
 printing eip:
e0aca92f
*pde = 00000000
Oops: 0002
CPU:    0
EIP:    0010:[<e0aca92f>]    Tainted: PF
EFLAGS: 00010246
eax: 00000000   ebx: 00000001   ecx: 00000000   edx: 00000000
esi: 00000000   edi: c2276400   ebp: c3ee3e70   esp: c3ee3e58
ds: 0018   es: 0018   ss: 0018
Process vmware-vmx (pid: 1179, stackpage=c3ee3000)
Stack: 00000000 c024da4e 42614000 c34ba660 c331d2e0 ffffffea c3ee3f90 e0ac8703
       c2276400 00000000 00000000 00000000 c30e6560 00000007 c30e6560 df9d5400
       c01ff59d c30e6560 00000000 c3ee3f3c c30e6560 c0201916 c30e6560 00000007
Call Trace:    [<c024da4e>] [<e0ac8703>] [<c01ff59d>] [<c0201916>] [<c0248a8c>]
  [<c035765f>] [<c01cd396>] [<c01cd9c8>] [<c01f7ed5>] [<c01a1783>]

Code: 89 9e e8 05 00 00 50 50 8b 45 0c 50 57 e8 33 0d 00 00 83 c4


Expected Results:  
Even if vmware does not work over the network, It is not normal for a userland
process (and even a non privileged one, too) to crash the entire system

# uname -a
Linux tleilax.localnet 2.4.22-gentoo-r7A #1 SMP Sat Feb 28 03:15:41 EET 2004
i686 Pentium III (Coppermine) GenuineIntel GNU/Linux
Comment 1 Vali "Blue" Dragnuta 2004-02-27 18:24:12 UTC
Created attachment 26494 [details]
kernel,hardware and my config.

Details about my system : kernel,hardware and my config.
Comment 2 Chris Gianelloni (RETIRED) gentoo-dev 2004-02-28 08:33:19 UTC
Unfortunately, VMWare isn't really a userland process.  It has kernel modules that can cause serious headaches.  I'll see what I can come up with.

Is there any way you can test with a 2.6 kernel?  maybe the vanilla or gentoo-dev-sources?

I just want to make sure it isn't kernel specific.  The 2.4.22-gentoo-r5(-r6,-r7) kernels all use the same patchset with minor differences, so I need to rule that out.  Possibly even 2.4.25 vanilla, too.
Comment 3 Vali "Blue" Dragnuta 2004-02-28 09:09:48 UTC
Well, I know that vmware is not an entirely userland process due to its modules
loaded into the kernel. However, the exception occurs in userland.
Anyway, I will test on vanilla-sources to see if the  problem apears again.
I will also try to use another video card.  I will also try to change the video card and see if the problem persists. I'll be back with new details in a few hours.
Comment 4 Vali "Blue" Dragnuta 2004-02-28 10:59:26 UTC
Just repeated the experiment using vanilla 2.4.25 kernel.
The system does not crash any more. 
So, there must be something in the gentoo patches. 
I heared people complaining about the preempt patch with vmware, however I am not
quite sure about that being the problem in my case. I will try to further investigate the problem.
If I find something else, I will post again. I still want to run the gentoo
kernel because I need a few of it's features (ipsec is one of them).
Comment 5 Chris Gianelloni (RETIRED) gentoo-dev 2004-03-18 16:35:15 UTC
Can you see if there is still a problem using the new 4.5.1 ebuild in portage and report back to me?
Comment 6 Chris Gianelloni (RETIRED) gentoo-dev 2004-04-09 11:43:09 UTC
Well, it's been a few weeks and no word, so I am resolving this bug.  If you arestill having problems with the new version of VMWare, then reopen this bug.