public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [OOPS] with 2.4.18-pre4+linux-2.4.18-NFS_ALL
@ 2002-01-20 16:41 Hans-Peter Jansen
  2002-01-20 18:03 ` Trond Myklebust
  0 siblings, 1 reply; 6+ messages in thread
From: Hans-Peter Jansen @ 2002-01-20 16:41 UTC (permalink / raw)
  To: Trond Myklebust; +Cc: linux-kernel

Hi Trond et al.,

I can reliably reproduce this oops on my diskless with NFS_ALL applied,
but not with plain-pre4, just by quitting one of {StarOffice,VMware}.

Unable to handle kernel NULL pointer dereference at virtual address 0000005c
c0158fbf
*pde = 00000000
Oops: 0000
CPU:    0
EIP:    0010:[nfs_update_request+447/768]    Not tainted
EFLAGS: 00010246
eax: 00000000   ebx: 00000000   ecx: 00000000   edx: de25eab8
esi: de25e900   edi: 00000000   ebp: de25eab8   esp: e21d1ef0
ds: 0018   es: 0018   ss: 0018
Process soffice.bin (pid: 2030, stackpage=e21d1000)
Stack: de25e900 c17508c0 00000000 00001000 de25e9b4 de25e900 00000000 
de25e900 
       00001000 c01588a2 00000000 de25e900 c17508c0 00000000 00001000 
de25e9b4 
       de25e900 00000000 c17508c0 c0158a2b 00000000 de25e900 c17508c0 
00000000 
Call Trace: [nfs_writepage_async+34/224] [nfs_writepage+203/272] 
[nfs_writepage+0/272] [filemap_fdatasync+83/144] [msync_interval+105/208] 
Code: 8b 58 5c 85 db 74 1d 81 7b 18 f0 a4 4a 0f 74 14 68 b2 00 00 
Using defaults from ksymoops -t elf32-i386 -a i386

Code;  00000000 Before first symbol
00000000 <_EIP>:
Code;  00000000 Before first symbol
   0:   8b 58 5c                  mov    0x5c(%eax),%ebx
Code;  00000002 Before first symbol
   3:   85 db                     test   %ebx,%ebx
Code;  00000004 Before first symbol
   5:   74 1d                     je     24 <_EIP+0x24> 00000024 Before first 
symbol
Code;  00000006 Before first symbol
   7:   81 7b 18 f0 a4 4a 0f      cmpl   $0xf4aa4f0,0x18(%ebx)
Code;  0000000e Before first symbol
   e:   74 14                     je     24 <_EIP+0x24> 00000024 Before first 
symbol
Code;  00000010 Before first symbol
  10:   68 b2 00 00 00            push   $0xb2


1 warning and 1 error issued.  Results may not be reliable.

Haven't used any funky nfs options:

CONFIG_NFS_FS=y
CONFIG_NFS_V3=y
# CONFIG_NFS_DIRECTIO is not set
CONFIG_ROOT_NFS=y
CONFIG_NFSD=m
CONFIG_NFSD_V3=y
# CONFIG_NFSD_TCP is not set
CONFIG_SUNRPC=y
CONFIG_LOCKD=y
CONFIG_LOCKD_V4=y

It is not related to the symlink fix, I reverted it, just to be sure.
After reverting NFS_ALL, oopsing went away. Any ideas other then trying
all patches one by one :-(

Cheers,
  Hans-Peter

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [OOPS] with 2.4.18-pre4+linux-2.4.18-NFS_ALL
  2002-01-20 16:41 [OOPS] with 2.4.18-pre4+linux-2.4.18-NFS_ALL Hans-Peter Jansen
@ 2002-01-20 18:03 ` Trond Myklebust
  2002-01-20 22:27   ` Hans-Peter Jansen
  0 siblings, 1 reply; 6+ messages in thread
From: Trond Myklebust @ 2002-01-20 18:03 UTC (permalink / raw)
  To: Hans-Peter Jansen; +Cc: linux-kernel

>>>>> " " == Hans-Peter Jansen <hpj@urpla.net> writes:

     > Hi Trond et al., I can reliably reproduce this oops on my
     > diskless with NFS_ALL applied, but not with plain-pre4, just by
     > quitting one of {StarOffice,VMware}.

The new version should be rid of it. It was a call to get_file() which
was missing a test for a NULL argument.

Cheers,
  Trond

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [OOPS] with 2.4.18-pre4+linux-2.4.18-NFS_ALL
  2002-01-20 18:03 ` Trond Myklebust
@ 2002-01-20 22:27   ` Hans-Peter Jansen
  2002-01-20 23:25     ` Trond Myklebust
  0 siblings, 1 reply; 6+ messages in thread
From: Hans-Peter Jansen @ 2002-01-20 22:27 UTC (permalink / raw)
  To: Trond Myklebust; +Cc: linux-kernel

On Sunday, 20. January 2002 19:03, Trond Myklebust wrote:
> >>>>> " " == Hans-Peter Jansen <hpj@urpla.net> writes:
>      > Hi Trond et al., I can reliably reproduce this oops on my
>      > diskless with NFS_ALL applied, but not with plain-pre4, just by
>      > quitting one of {StarOffice,VMware}.
>
> The new version should be rid of it. It was a call to get_file() which
> was missing a test for a NULL argument.

Are you sure?

Unable to handle kernel NULL pointer dereference at virtual address 0000005c
c0157a2f
*pde = 00000000
Oops: 0000
CPU:    0
EIP:    0010:[nfs_update_request+447/752]    Not tainted
EFLAGS: 00013246
eax: e75a7340   ebx: 00000000   ecx: 00000005   edx: 00000000
esi: e75a7340   edi: 00000000   ebp: e75a74f8   esp: e9243e28
ds: 0018   es: 0018   ss: 0018
Process vmware (pid: 1269, stackpage=e9243000)
Stack: e75a7340 c1988a40 00000000 00001000 c1988a40 e75a7340 00000000 
e75a7340 
       00001000 c0157362 00000000 e75a7340 c1988a40 00000000 00001000 
c1988a40 
       e75a7340 00000000 c1988a40 c01574a4 00000000 e75a7340 c1988a40 
00000000 
Call Trace: [nfs_writepage_async+34/208] [nfs_writepage+148/208] 
[nfs_writepage+0/208] [filemap_fdatasync+83/144] [nfs_notify_change+145/432] 
   [<f298e820>] [<f298c7a5>] [unix_stream_sendmsg+502/688] [<f298b51f>] 
[set_page_dirty+67/80] [notify_change+76/320] 
Code: 8b 42 5c 85 c0 74 0b 81 78 18 f0 a4 4a 0f 74 02 0f 0b 50 e8 
Using defaults from ksymoops -t elf32-i386 -a i386

Trace; f298e820 <[vmmon].rodata.start+5a0/c1e>
Trace; f298c7a4 <[vmmon]Vmx86_ReleaseVM+7c/88>
Code;  00000000 Before first symbol
00000000 <_EIP>:
Code;  00000000 Before first symbol
   0:   8b 42 5c                  mov    0x5c(%edx),%eax
Code;  00000002 Before first symbol
   3:   85 c0                     test   %eax,%eax
Code;  00000004 Before first symbol
   5:   74 0b                     je     12 <_EIP+0x12> 00000012 Before first 
symbol
Code;  00000006 Before first symbol
   7:   81 78 18 f0 a4 4a 0f      cmpl   $0xf4aa4f0,0x18(%eax)
Code;  0000000e Before first symbol
   e:   74 02                     je     12 <_EIP+0x12> 00000012 Before first 
symbol
Code;  00000010 Before first symbol
  10:   0f 0b                     ud2a   
Code;  00000012 Before first symbol
  12:   50                        push   %eax
Code;  00000012 Before first symbol
  13:   e8 00 00 00 00            call   18 <_EIP+0x18> 00000018 Before first 
symbol

Just verified two times. Will revert it again :(

-rw-rw-r--    1 hp       lisa       141885 Jan 20 20:08 
linux-2.4.18-NFS_ALL-1.dif

Would you mind to number up your NFS_ALL patch versions?

> Cheers,
>   Trond

Cheers,
  Hans-Peter

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [OOPS] with 2.4.18-pre4+linux-2.4.18-NFS_ALL
  2002-01-20 22:27   ` Hans-Peter Jansen
@ 2002-01-20 23:25     ` Trond Myklebust
  2002-01-23  0:50       ` Hans-Peter Jansen
  0 siblings, 1 reply; 6+ messages in thread
From: Trond Myklebust @ 2002-01-20 23:25 UTC (permalink / raw)
  To: Hans-Peter Jansen; +Cc: Trond Myklebust, linux-kernel

>>>>> " " == Hans-Peter Jansen <hpj@urpla.net> writes:

     > On Sunday, 20. January 2002 19:03, Trond Myklebust wrote:
    >> >>>>> " " == Hans-Peter Jansen <hpj@urpla.net> writes:
    >> > Hi Trond et al., I can reliably reproduce this oops on my
    >> > diskless with NFS_ALL applied, but not with plain-pre4, just
    >> > by quitting one of {StarOffice,VMware}.
    >>
    >> The new version should be rid of it. It was a call to
    >> get_file() which was missing a test for a NULL argument.

     > Are you sure?

I forgot the nfs_cred_file() in the line above. That too is fixed now,
and so fsx is running fine again...

Cheers,
   Trond

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [OOPS] with 2.4.18-pre4+linux-2.4.18-NFS_ALL
  2002-01-20 23:25     ` Trond Myklebust
@ 2002-01-23  0:50       ` Hans-Peter Jansen
  2002-01-23  1:18         ` Dave Jones
  0 siblings, 1 reply; 6+ messages in thread
From: Hans-Peter Jansen @ 2002-01-23  0:50 UTC (permalink / raw)
  To: Trond Myklebust; +Cc: linux-kernel

On Monday, 21. January 2002 00:25, Trond Myklebust wrote:
> >>>>> " " == Hans-Peter Jansen <hpj@urpla.net> writes:
>      > On Sunday, 20. January 2002 19:03, Trond Myklebust wrote:
>     >> >>>>> " " == Hans-Peter Jansen <hpj@urpla.net> writes:
>     >> >
>     >> > Hi Trond et al., I can reliably reproduce this oops on my
>     >> > diskless with NFS_ALL applied, but not with plain-pre4, just
>     >> > by quitting one of {StarOffice,VMware}.
>     >>
>     >> The new version should be rid of it. It was a call to
>     >> get_file() which was missing a test for a NULL argument.
>     >>
>      > Are you sure?
>
> I forgot the nfs_cred_file() in the line above. That too is fixed now,
> and so fsx is running fine again...

Would you like to give me a pointer to the famous fsx?

Just for the record, your actual NFS_ALL patch finally solved the problem(s), 
it introduced ;-)

Many thanks.

> Cheers,
>    Trond

Happy-NFS-ing-ly y'rs,
  Hans-Peter

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [OOPS] with 2.4.18-pre4+linux-2.4.18-NFS_ALL
  2002-01-23  0:50       ` Hans-Peter Jansen
@ 2002-01-23  1:18         ` Dave Jones
  0 siblings, 0 replies; 6+ messages in thread
From: Dave Jones @ 2002-01-23  1:18 UTC (permalink / raw)
  To: Hans-Peter Jansen; +Cc: Trond Myklebust, linux-kernel

On Wed, Jan 23, 2002 at 01:50:09AM +0100, Hans-Peter Jansen wrote:
 > > I forgot the nfs_cred_file() in the line above. That too is fixed now,
 > > and so fsx is running fine again...
 > Would you like to give me a pointer to the famous fsx?

 It's on my website (url below) in the cruft/ dir, but there
 seem to be 'issues' at host right now. I've put a copy at
 ftp://ftp.kernel.org/pub/linux/kernel/people/davej/tools/
 
-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2002-01-23  1:19 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-01-20 16:41 [OOPS] with 2.4.18-pre4+linux-2.4.18-NFS_ALL Hans-Peter Jansen
2002-01-20 18:03 ` Trond Myklebust
2002-01-20 22:27   ` Hans-Peter Jansen
2002-01-20 23:25     ` Trond Myklebust
2002-01-23  0:50       ` Hans-Peter Jansen
2002-01-23  1:18         ` Dave Jones

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox