Re: kern/38782: Panic at uvm_swap_markbad when starting Xorg

[Mindaugas Rasiukevicius <rmind%NetBSD.org@localhost>]

The following reply was made to PR kern/38782; it has been noted by GNATS.

From: Mindaugas Rasiukevicius <rmind%NetBSD.org@localhost>
To: adam.hoka%gmail.com@localhost
Cc: gnats-bugs%NetBSD.org@localhost, netbsd-bugs%netbsd.org@localhost
Subject: Re: kern/38782: Panic at uvm_swap_markbad when starting Xorg
Date: Wed, 28 May 2008 20:47:35 +0100

 Hello,
 
 > >Number:         38782
 > >Category:       kern
 > >Synopsis:       Panic at uvm_swap_markbad when starting Xorg
 > ...
 
 > uvm_fault(0xcc3ebd08, 0, 1) -> 0xe
 > kernel: supervisor trap page fault, code=0
 > Stopped in pid 15917.1 (Xorg) at        netbsd:uvm_swap_markbad+0x30:   movl 
 >    
 > 0x24(%eax),%eax
 > ?
 > db{0}> syncing disks... 1 done
 
 - Can you reproduce this with debugging options enabled?
 - Are you using kern.ipc.shm_use_phys=1?
 - What is the memory and swap usage?
 
 It looks like swapdrum_getsdp() in uvm_swap_markbad() returns NULL, what
 should not happen. But I am not sure..
 
 -- 
 Best regards,
 Mindaugas
 www.NetBSD.org