Skip to contentSkip to conversation

2.6.17 oops, possibly ntfs/mmap related

DISCUSSION Anton Altaparmakov 5 participants 10 messages Sep 21, 2006 – Sep 24, 2006 1 message not in the archive

lore
Share & tools
lore
#1 lore raw
In reply to a message not in the archive

Hi,

On Tue, 2006-09-12 at 20:56 -0700, Andrew Morton wrote:

Andrew, thanks for forwarding me the message...

Quoted 20 lines

Begin forwarded message:

Date: Wed, 13 Sep 2006 10:05:42 +0930 (CST)
From: Jonathan Woithe <[email protected]>
To: [email protected]
Cc: [email protected] (Jonathan Woithe)
Subject: 2.6.17 oops, possibly ntfs/mmap related

We have a machine which is currently making heavy use of a usb hard disc formatted with ntfs. There have been two occasions where the kernel has oopsed while this disc was being accessed heavily. Before adding this HDD the machine in question was rock solid which leads me to think that it might be related to ntfs. USB drives formatted with other filesystems do not appear to suffer from this problem.

Unfortunately bogofilter considers the oops reports as spam so I cannot post them to the list. I have instead put the full text of my original post regarding this topic on the web at

Quoted 5 lines

http://www.atrad.com.au/~jwoithe/kernel/oopses-20060913.txt

I'm happy to try things to narrow down the cause if it will help.

Please CC me on reply.

These were the oopses from above text url:

Quoted 56 lines

The first oops caused the machine to totally lock up:

  BUG: unable to handle kernel paging request at virtual address e4004de0
   printing eip:
  c012de0c
  *pde = 00000000
  Oops: 0000 [#1]
  Modules linked in: ntfs 8139too via_agp agpgart usb_storage ehci_hcd uhci_hcd usbcore
  CPU:    0
  EIP:    0060:[<c012de0c>]    Not tainted VLI
  EFLAGS: 00010082   (2.6.17 #2) 
  EIP is at find_get_page+0x11/0x22
  eax: e4004de0   ebx: c02f01a8   ecx: e4004de0   edx: e4004de0
  esi: 00000000   edi: 00000066   ebp: cfc20574   esp: c770bee8
  ds: 007b   es: 007b   ss: 0068
  Process sh (pid: 10467, threadinfo=c770a000 task=cfa495c0)
  Stack: c012ea09 00000002 00000000 cfc204d8 cff882a4 cff88260 c770bf30 c5962544
         c02f01a8 00000000 ceec10a0 080aef10 c01385d1 00000000 cfc20574 080aef10
         c5962544 ceec10a0 00000002 c7aeb080 080aef10 ceec10a0 080aef10 c013886a
  Call Trace:
   <c012ea09> filemap_nopage+0x98/0x2b2  <c01385d1> do_no_page+0x6d/0x1e1
   <c013886a> __handle_mm_fault+0xc4/0x162  <c0112190> do_page_fault+0x23e/0x56b
   <c01c43c1> copy_to_user+0x41/0x49  <c0111f52> do_page_fault+0x0/0x56b
   <c010342f> error_code+0x4f/0x54 
  Code: a0 fe ff ff 89 ea b9 e2 d7 12 c0 6a 02 e8 5f ec 15 00 83 c4 44 5b 5e 5f 5d c3 fa 83 c0 04 e8 2c 3f 09 00 85 c0 89 c1 74 0f 89 c2 <8b> 00 f6 c4 40 74 03 8b 51 0c ff 42 04 fb 89 c8 c3 fa 83 c0 04 

In the case of the second oops the machine was still partially usable and a clean shutdown was possible. However, services such as sshd were no longer responding.

  BUG: unable to handle kernel paging request at virtual address 0010c744
   printing eip:
  c013be50
  *pde = 00000000
  Oops: 0002 [#1]
  Modules linked in: ntfs 8139too via_agp agpgart usb_storage ehci_hcd uhci_hcd usbcore
  CPU:    0
  EIP:    0060:[<c013be50>]    Tainted: G   M  VLI
  EFLAGS: 00010282   (2.6.17 #2) 
  EIP is at anon_vma_unlink+0x16/0x3c
  eax: 0010c740   ebx: cf1070cc   ecx: cf107104   edx: cf8bc740
  esi: cf8bc740   edi: b7e82000   ebp: 00000000   esp: cdad7f58
  ds: 007b   es: 007b   ss: 0068
  Process sh (pid: 20272, threadinfo=cdad6000 task=c0d8d580)
  Stack: cf1070cc cf61f3e4 c0136b5f cdad7f80 c4084b74 cf8b5860 00000001 00000000
         c013ab92 00000000 c0371b7c 000000b9 cf8b5860 c0d8d580 c01145dd cdad6000
         c0118187 cdad6000 00000000 00000000 cdad6000 c0118380 00000000 b7f9968c
  Call Trace:
   <c0136b5f> free_pgtables+0x41/0x82  <c013ab92> exit_mmap+0x6a/0xb8
   <c01145dd> mmput+0x1b/0x5e  <c0118187> do_exit+0x14e/0x2d1
   <c0118380> sys_exit_group+0x0/0xd  <c010299b> syscall_call+0x7/0xb
  Code: c9 74 10 8b 11 8d 40 38 89 42 04 89 53 38 89 48 04 89 01 5b c3 56 53 8b 70 40 89 c3 85 f6 74 2e 8d 48 38 8b 40 38 8b 51 04 89 02 <89> 50 04 c7 43 38 00 01 10 00 39 36 c7 41 04 00 02 20 00 75 0e 
  EIP: [<c013be50>] anon_vma_unlink+0x16/0x3c SS:ESP 0068:cdad7f58
   <1>Fixing recursive fault but reboot is needed!
Quoted 4 lines

I'm not entirely sure why the kernel considered itself tainted in the second oops and not in the first - the setup hadn't changed and precisely the same kernel modules were loaded. This machine does not have any external (ie:
out-of-tree) modules installed.

Weird. The traces do not include NTFS at all in them so I have no idea if NTFS has anything to do with this or not...

Anyone have any idea what these oopses mean?!?

Best regards,

Anton

Signature · 4 lines
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer / IRC: #ntfs on irc.freenode.net
WWW: http://www.linux-ntfs.org/ & http://www-stu.christs.cam.ac.uk/~aia21/
#2 lore raw

Hi,

Quoted 10 lines · Anton Altaparmakov

On Tue, 2006-09-12 at 20:56 -0700, Andrew Morton wrote:
Andrew, thanks for forwarding me the message...

Quoted 8 lines

Begin forwarded message:

We have a machine which is currently making heavy use of a usb hard disc formatted with ntfs. There have been two occasions where the kernel has oopsed while this disc was being accessed heavily. Before adding this HDD the machine in question was rock solid which leads me to think that it might be related to ntfs. USB drives formatted with other filesystems do not appear to suffer from this problem.

I have now seen such an oops too with 2.6.18 kernel. Note no NTFS file systems were mounted at the time (but I had an NTFS file system mounted earlier in the day).

The oops is caused by kswapd0 kernel thread, the stack trace is:

Call Trace:
 [<c10470a3>] shrink_inactive_list+0x46b/0x790
 [<c104747c>] shrink_zone+0xb4/0xd3
 [<c104797d>] kswapd+0x2de/0x3cf
 [<c102c18e>] kthread+0xc2/0xf0
 [<c1000bf1>] kernel_thread_helper+0x5/0xb
DWARF2 unwinder stuck at kernel_thread_helper+0x5/0xb
Leftover inexact backtrace:
 [<c1003e6c>] show_stack_log_lvl+0x8c/0x97
 [<c1003fc8>] show_registers+0x151/0x1c6
 [<c10041af>] die+0x172/0x27b
 [<c145f22c>] do_page_fault+0x42c/0x4f9
 [<c10037dd>] error_code+0x39/0x40
 [<c10470a3>] shrink_inactive_list+0x46b/0x790
 [<c104747c>] shrink_zone+0xb4/0xd3
 [<c104797d>] kswapd+0x2de/0x3cf
 [<c102c18e>] kthread+0xc2/0xf0
 [<c1000bf1>] kernel_thread_helper+0x5/0xb

And the EIP is at fs/buffer.c::try_to_release_page() the code of which is here:

int try_to_release_page(struct page *page, gfp_t gfp_mask)

{
        struct address_space * const mapping = page->mapping;

        BUG_ON(!PageLocked(page));
        if (PageWriteback(page))
                return 0;

        if (mapping && mapping->a_ops->releasepage)

^^^ bug happens here when the value of mapping->a_ops is used to obtain mapping->a_ops->releasepage

                return mapping->a_ops->releasepage(page, gfp_mask);
        return try_to_free_buffers(page);
}

This bug seems to suggest that there is a page which the kernel is trying to release private data which has page->mapping set to a valid value and page->mapping->a_ops apparently set to an invalid value and when page->mapping->a_ops->releasepage is dereferenced it causes an oops with the kernel saying:

BUG: unable to handle kernel paging request at virtual address 020030d2

The values of the relevant variables from the oops are:

page = 0xc2248fa0
page->mapping = 0xe3a79eac
page->mapping->a_ops = 0x020030aa

Note that 0x020030aa+0x28 = 020030d2 which is the oops causing address and 0x28 is the offset of the releasepage function pointer in the address space operations structure...

This oops is not identical to the oopses pointed out by Jonathan at:

http://www.atrad.com.au/~jwoithe/kernel/oopses-20060913.txt

But those oopses have to do with pages also so could be related...

Anyone have any ideas how a page can end up in such a weird state?

Best regards,

Anton

Signature · 4 lines
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer / IRC: #ntfs on irc.freenode.net
WWW: http://www.linux-ntfs.org/ & http://www-stu.christs.cam.ac.uk/~aia21/
#3 lore raw
Quoted 7 lines · On Thu, 21 Sep 2006 15:41:36 +0100 Anton Altaparmakov <[email protected]> wrote:

Hi,

Quoted 4 lines · Anton Altaparmakov

On Tue, 2006-09-12 at 20:56 -0700, Andrew Morton wrote:
Andrew, thanks for forwarding me the message...

Quoted 2 lines

Begin forwarded message:

Quoted 8 lines
Quoted 6 lines
Quoted 6 lines

We have a machine which is currently making heavy use of a usb hard disc formatted with ntfs. There have been two occasions where the kernel has oopsed while this disc was being accessed heavily. Before adding this HDD the machine in question was rock solid which leads me to think that it might be related to ntfs. USB drives formatted with other filesystems do not appear to suffer from this problem.

I have now seen such an oops too with 2.6.18 kernel.

I assume it is a once-off?

Quoted 51 lines

Note no NTFS file
systems were mounted at the time (but I had an NTFS file system mounted earlier in the day).

The oops is caused by kswapd0 kernel thread, the stack trace is:

Call Trace:
 [<c10470a3>] shrink_inactive_list+0x46b/0x790
 [<c104747c>] shrink_zone+0xb4/0xd3
 [<c104797d>] kswapd+0x2de/0x3cf
 [<c102c18e>] kthread+0xc2/0xf0
 [<c1000bf1>] kernel_thread_helper+0x5/0xb
DWARF2 unwinder stuck at kernel_thread_helper+0x5/0xb
Leftover inexact backtrace:
 [<c1003e6c>] show_stack_log_lvl+0x8c/0x97
 [<c1003fc8>] show_registers+0x151/0x1c6
 [<c10041af>] die+0x172/0x27b
 [<c145f22c>] do_page_fault+0x42c/0x4f9
 [<c10037dd>] error_code+0x39/0x40
 [<c10470a3>] shrink_inactive_list+0x46b/0x790
 [<c104747c>] shrink_zone+0xb4/0xd3
 [<c104797d>] kswapd+0x2de/0x3cf
 [<c102c18e>] kthread+0xc2/0xf0
 [<c1000bf1>] kernel_thread_helper+0x5/0xb

And the EIP is at fs/buffer.c::try_to_release_page() the code of which is here:

int try_to_release_page(struct page *page, gfp_t gfp_mask)

{
        struct address_space * const mapping = page->mapping;

        BUG_ON(!PageLocked(page));
        if (PageWriteback(page))
                return 0;

        if (mapping && mapping->a_ops->releasepage)

^^^ bug happens here when the value of mapping->a_ops is used to obtain mapping->a_ops->releasepage

                return mapping->a_ops->releasepage(page, gfp_mask);
        return try_to_free_buffers(page);
}

This bug seems to suggest that there is a page which the kernel is trying to release private data which has page->mapping set to a valid value and page->mapping->a_ops apparently set to an invalid value and when page->mapping->a_ops->releasepage is dereferenced it causes an oops with the kernel saying:

Quoted 7 lines
BUG: unable to handle kernel paging request at virtual address 020030d2

The values of the relevant variables from the oops are:

page = 0xc2248fa0
page->mapping = 0xe3a79eac
page->mapping->a_ops = 0x020030aa

I wonder if page->mapping really wanted to be 0xc3a79eac, only something set bit 29.

Quoted 9 lines

Note that 0x020030aa+0x28 = 020030d2 which is the oops causing address and 0x28 is the offset of the releasepage function pointer in the address space operations structure...

This oops is not identical to the oopses pointed out by Jonathan at:

http://www.atrad.com.au/~jwoithe/kernel/oopses-20060913.txt

But those oopses have to do with pages also so could be related...

Looks a bit different - Jonathan appears to have pulled a bad page* out of the radix tree whereas you got your page off the LRU.

Quoted 1 line

Anyone have any ideas how a page can end up in such a weird state?

Nope.

#4 lore raw
Quoted 5 lines · Andrew Morton
Quoted 3 lines · On Thu, 21 Sep 2006 15:41:36 +0100 Anton Altaparmakov <[email protected]> wrote:
BUG: unable to handle kernel paging request at virtual address 020030d2
Quoted 8 lines
Quoted 5 lines

The values of the relevant variables from the oops are:

page = 0xc2248fa0
page->mapping = 0xe3a79eac
page->mapping->a_ops = 0x020030aa

I wonder if page->mapping really wanted to be 0xc3a79eac, only something set bit 29.

Perhaps, but I'm more suspicious of that 0x0200 top half of the a_ops ptr.

Quoted 5 lines
Quoted 5 lines

Note that 0x020030aa+0x28 = 020030d2 which is the oops causing address and 0x28 is the offset of the releasepage function pointer in the address space operations structure...

Quoted 8 lines
Quoted 5 lines

This oops is not identical to the oopses pointed out by Jonathan at:

http://www.atrad.com.au/~jwoithe/kernel/oopses-20060913.txt

But those oopses have to do with pages also so could be related...

Looks a bit different - Jonathan appears to have pulled a bad page* out of the radix tree whereas you got your page off the LRU.

Jonathan does show a second oops, from a later boot:

  BUG: unable to handle kernel paging request at virtual address 0010c744
   printing eip:
  c013be50
  *pde = 00000000
  Oops: 0002 [#1]
  Modules linked in: ntfs 8139too via_agp agpgart usb_storage ehci_hcd uhci_hcd usbcore
  CPU:    0
  EIP:    0060:[<c013be50>]    Tainted: G   M  VLI
  EFLAGS: 00010282   (2.6.17 #2) 
  EIP is at anon_vma_unlink+0x16/0x3c
  eax: 0010c740   ebx: cf1070cc   ecx: cf107104   edx: cf8bc740
  esi: cf8bc740   edi: b7e82000   ebp: 00000000   esp: cdad7f58

I haven't worked out the disassembly in detail to support the idea (though certainly anon_vma_unlink would be trying to list_del around here), but that eax and esi do suggest a corrupted list: somehow the top half of a pointer overwritten by the top half of LIST_POISON1.

And in Anton's case, the top half of a pointer overwritten by the bottom half of LIST_POISON2.

Maybe just coincidence, and I've nothing more illuminating to add; but just a hint of a list_del going very wrong somewhere?

Hugh

#5 lore raw
Quoted 18 lines · Hugh Dickins
  BUG: unable to handle kernel paging request at virtual address 0010c744
   printing eip:
  c013be50
  *pde = 00000000
  Oops: 0002 [#1]
  Modules linked in: ntfs 8139too via_agp agpgart usb_storage ehci_hcd uhci_hcd usbcore
  CPU:    0
  EIP:    0060:[<c013be50>]    Tainted: G   M  VLI
  EFLAGS: 00010282   (2.6.17 #2) 
  EIP is at anon_vma_unlink+0x16/0x3c
  eax: 0010c740   ebx: cf1070cc   ecx: cf107104   edx: cf8bc740
  esi: cf8bc740   edi: b7e82000   ebp: 00000000   esp: cdad7f58

I haven't worked out the disassembly in detail to support the idea (though certainly anon_vma_unlink would be trying to list_del around here), but that eax and esi do suggest a corrupted list: somehow the top half of a pointer overwritten by the top half of LIST_POISON1.

Quoted 5 lines

And in Anton's case, the top half of a pointer overwritten by the bottom half of LIST_POISON2.

Maybe just coincidence, and I've nothing more illuminating to add; but just a hint of a list_del going very wrong somewhere?

Given a machine check happened, the state of the machine in general is questionable. I'd recommend a run of memtest86+

Dave

#9 lore raw
In reply to Andrew Morton
Quoted 14 lines · Andrew Morton
Quoted 12 lines · On Thu, 21 Sep 2006 15:41:36 +0100 Anton Altaparmakov <[email protected]> wrote:
Quoted 10 lines · Anton Altaparmakov

On Tue, 2006-09-12 at 20:56 -0700, Andrew Morton wrote:
Andrew, thanks for forwarding me the message...

Quoted 8 lines

Begin forwarded message:

We have a machine which is currently making heavy use of a usb hard disc formatted with ntfs. There have been two occasions where the kernel has oopsed while this disc was being accessed heavily. Before adding this HDD the machine in question was rock solid which leads me to think that it might be related to ntfs. USB drives formatted with other filesystems do not appear to suffer from this problem.

Quoted 3 lines
Quoted 1 line

I have now seen such an oops too with 2.6.18 kernel.

I assume it is a once-off?

So far yes. I now have seen a recursive locking thing reported by the new lock analyzer but that looks like it has to do with NFS (my home directory is on NFS) so I don't think it is in any way related.

Quoted 53 lines
Quoted 53 lines

Note no NTFS file
systems were mounted at the time (but I had an NTFS file system mounted earlier in the day).

The oops is caused by kswapd0 kernel thread, the stack trace is:

Call Trace:
 [<c10470a3>] shrink_inactive_list+0x46b/0x790
 [<c104747c>] shrink_zone+0xb4/0xd3
 [<c104797d>] kswapd+0x2de/0x3cf
 [<c102c18e>] kthread+0xc2/0xf0
 [<c1000bf1>] kernel_thread_helper+0x5/0xb
DWARF2 unwinder stuck at kernel_thread_helper+0x5/0xb
Leftover inexact backtrace:
 [<c1003e6c>] show_stack_log_lvl+0x8c/0x97
 [<c1003fc8>] show_registers+0x151/0x1c6
 [<c10041af>] die+0x172/0x27b
 [<c145f22c>] do_page_fault+0x42c/0x4f9
 [<c10037dd>] error_code+0x39/0x40
 [<c10470a3>] shrink_inactive_list+0x46b/0x790
 [<c104747c>] shrink_zone+0xb4/0xd3
 [<c104797d>] kswapd+0x2de/0x3cf
 [<c102c18e>] kthread+0xc2/0xf0
 [<c1000bf1>] kernel_thread_helper+0x5/0xb

And the EIP is at fs/buffer.c::try_to_release_page() the code of which is here:

int try_to_release_page(struct page *page, gfp_t gfp_mask)

{
        struct address_space * const mapping = page->mapping;

        BUG_ON(!PageLocked(page));
        if (PageWriteback(page))
                return 0;

        if (mapping && mapping->a_ops->releasepage)

^^^ bug happens here when the value of mapping->a_ops is used to obtain mapping->a_ops->releasepage

                return mapping->a_ops->releasepage(page, gfp_mask);
        return try_to_free_buffers(page);
}

This bug seems to suggest that there is a page which the kernel is trying to release private data which has page->mapping set to a valid value and page->mapping->a_ops apparently set to an invalid value and when page->mapping->a_ops->releasepage is dereferenced it causes an oops with the kernel saying:

BUG: unable to handle kernel paging request at virtual address 020030d2
Quoted 8 lines
Quoted 5 lines

The values of the relevant variables from the oops are:

page = 0xc2248fa0
page->mapping = 0xe3a79eac
page->mapping->a_ops = 0x020030aa

I wonder if page->mapping really wanted to be 0xc3a79eac, only something set bit 29.

I don't know, it could be but the machine is totally stable so I would be surprised if it is bad ram...

Quoted 8 lines
Quoted 8 lines

Note that 0x020030aa+0x28 = 020030d2 which is the oops causing address and 0x28 is the offset of the releasepage function pointer in the address space operations structure...

This oops is not identical to the oopses pointed out by Jonathan at:

http://www.atrad.com.au/~jwoithe/kernel/oopses-20060913.txt

Quoted 8 lines
Quoted 1 line

But those oopses have to do with pages also so could be related...

Looks a bit different - Jonathan appears to have pulled a bad page* out of the radix tree whereas you got your page off the LRU.

Quoted 1 line

Anyone have any ideas how a page can end up in such a weird state?

Nope.

)-:

Best regards,

Anton

Signature · 4 lines
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer / IRC: #ntfs on irc.freenode.net
WWW: http://www.linux-ntfs.org/ & http://www-stu.christs.cam.ac.uk/~aia21/
#6 lore raw
In reply to Dave Jones
Quoted 20 lines · Hugh Dickins
Quoted 19 lines
  BUG: unable to handle kernel paging request at virtual address 0010c744
   printing eip:
  c013be50
  *pde = 00000000
  Oops: 0002 [#1]
  Modules linked in: ntfs 8139too via_agp agpgart usb_storage ehci_hcd uhci_hcd usbcore
  CPU:    0
  EIP:    0060:[<c013be50>]    Tainted: G   M  VLI
  EFLAGS: 00010282   (2.6.17 #2) 
  EIP is at anon_vma_unlink+0x16/0x3c
  eax: 0010c740   ebx: cf1070cc   ecx: cf107104   edx: cf8bc740
  esi: cf8bc740   edi: b7e82000   ebp: 00000000   esp: cdad7f58

I haven't worked out the disassembly in detail to support the idea (though certainly anon_vma_unlink would be trying to list_del around here), but that eax and esi do suggest a corrupted list: somehow the top half of a pointer overwritten by the top half of LIST_POISON1.

Quoted 8 lines
Quoted 5 lines

And in Anton's case, the top half of a pointer overwritten by the bottom half of LIST_POISON2.

Maybe just coincidence, and I've nothing more illuminating to add; but just a hint of a list_del going very wrong somewhere?

Given a machine check happened, the state of the machine in general is questionable. I'd recommend a run of memtest86+

That was already done. No memory errors were reported over 10 passes.

Secondly, the machine check indication was only present on one of the two oopses we saw. Furthermore, there was no indication in any log files that a machine check had occurred in the case of the second oops. Then again, perhaps machine checks don't get logged which would make this observation irrelevant.

Could we be looking at a dying CPU?

Regards
jonathan

#10 lore raw
In reply to Anton Altaparmakov
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 6 lines
Quoted 6 lines

We have a machine which is currently making heavy use of a usb hard disc formatted with ntfs. There have been two occasions where the kernel has oopsed while this disc was being accessed heavily. Before adding this HDD the machine in question was rock solid which leads me to think that it might be related to ntfs. USB drives formatted with other filesystems do not appear to suffer from this problem.

Quoted 7 lines
Quoted 3 lines
Quoted 1 line

I have now seen such an oops too with 2.6.18 kernel.

I assume it is a once-off?

So far yes. I now have seen a recursive locking thing reported by the new lock analyzer but that looks like it has to do with NFS (my home directory is on NFS) so I don't think it is in any way related.

Our setup also has user home directories on NFS, so that much at least is common to our configurations. I don't know if the USB/NTFS user was writing to their home directory as part of their work at the time of the oops though.

Regards
jonathan

#7 lore raw
In reply to Jonathan Woithe
Quoted 5 lines · Jonathan Woithe
Quoted 2 lines

Given a machine check happened, the state of the machine in general is questionable. I'd recommend a run of memtest86+

That was already done. No memory errors were reported over 10 passes.

Quoted 7 lines

Secondly, the machine check indication was only present on one of the two oopses we saw. Furthermore, there was no indication in any log files that a machine check had occurred in the case of the second oops. Then again, perhaps machine checks don't get logged which would make this observation irrelevant.

Could we be looking at a dying CPU?

Maybe. Or some other hardware problem. Insufficient cooling/power for eg.

Dave

#8 lore raw
Quoted 11 lines
Quoted 11 lines
Quoted 2 lines

Given a machine check happened, the state of the machine in general is questionable. I'd recommend a run of memtest86+

That was already done. No memory errors were reported over 10 passes.

Secondly, the machine check indication was only present on one of the two oopses we saw. Furthermore, there was no indication in any log files that a machine check had occurred in the case of the second oops. Then again, perhaps machine checks don't get logged which would make this observation irrelevant.

Quoted 3 lines
Quoted 1 line

Could we be looking at a dying CPU?

Maybe. Or some other hardware problem. Insufficient cooling/power for eg.

Power and cooling should be fine, and I've checked fans etc for correct functioning - all is ok. The other thing worth noting is that the problems with this machine only started once the USB/NTFS HDD started being used. Before this the machine has been rock solid for 2+ years, and the usage patterns haven't changed. Anyway, I'll keep an eye on it and post any subsequent information as it becomes available.

Regards
jonathan