Re: kern/45854

[Nat Sloss <nathanialsloss%yahoo.com.au@localhost>]

The following reply was made to PR kern/45854; it has been noted by GNATS.

From: Nat Sloss <nathanialsloss%yahoo.com.au@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: kern/45854
Date: Sat, 7 Jul 2012 15:02:02 +1000

 Hi.
 
 I think I'm having the same trouble on 6.99.7 i386.
 
 It happens a random for me.  It happened whilst copying, sometimes after a 
 long while (copying 120GB), sometimes after a short while (copying 50MB).
 
 This is the back trace:
 
 uvm_fault(0xc0c297a0, 0, 1) -> 0xe
 fatal page fault in supervisor mode
 trap type 6 code 0 eip c0326575 cs 8 eflags 210213 cr2 2c ilevel 0
 db{2}> bt
 genfs_do_putpages(c6112f24,0,0,0,0,201b,0,c6112f24,0,201b) at 
 netbsd:genfs_do_pu
 tpages+0x43a
 genfs_putpages(dbaf3c18,0,0,0,1,0,c4214aa0,c4214aa0,0,c0ad059c) at 
 netbsd:genfs_
 putpages+0x3f
 VOP_PUTPAGES(c6112f24,0,0,0,0,201b,c0bf63c0,c852fec0,8,c4214aa0) at 
 netbsd:VOP_P
 UTPAGES+0x68
 vinvalbuf(c6112f24,1,ffffffff,c4214aa0,0,0,0,c421ef8c,0,0) at 
 netbsd:vinvalbuf+0
 x62
 
vclean(c6112f24,8,c4214aa0,c4214aa0,c4214aa0,c08552aa,dbaf3d2c,c085530d,c0bf63c0
 ,c0bf63c0) at netbsd:vclean+0x200
 
cleanvnode(c0bf63c0,c0bf63c0,64,c4214aa0,c08552aa,0,0,c0100321,c4214aa0,db7000) 
 a
 t netbsd:cleanvnode+0x9a
 vdrain_thread(c4214aa0,db7000,dbf000,0,c0100307,0,0,0,0,0) at 
 netbsd:vdrain_thre
 ad+0x63
 
 it seams to read a null vnode maybe.  This would occur on an ffs v2 
 filesystem.
 
 Just a thought I don't know where the null values are coming from but would it 
 be possible not to add null vnodes to the free list/ or remove null entries 
 from the free list.
 
 Regards,
 
 Nat.