Skip to contentSkip to conversation

2.6.9-rc1-mm5: double fault on resume on Athlon64 w/ NForce 3

DISCUSSION Rafael J. Wysocki 2 participants 4 messages Sep 13, 2004 – Sep 18, 2004

lore
Share & tools
lore
#0 lore raw

Pavel,

JFYI, I get a double fault on resume on an Athlon64-based box:

Relocating
pagedir .....................................................................................................................| Reading image data (44091 pages): 100% 44091 done.
Stopping tasks: ===|
Freeing memory... done (0 pages freed)
PM: Restoring saved image.
<0>double fault: 0000 [1]
CPU 0

Modules linked in: ohci_hcd usbserial parport_pc lp parport joydev sg sd_mod 
sr_mod scsi_mod cpufreq_userspace powernow_k8 thermal proced
Pid: 18616, comm: hibernate.sh Not tainted 2.6.9-rc1-mm5
RIP: 0010:[<ffffffff80124aa7>] <ffffffff80124aa7>{do_page_fault+55}
RSP: 0000:000001001fdfff48  EFLAGS: 00010016
RAX: ffffffff80124a70 RBX: 0000000000000001 RCX: 000ffffffffff000
RDX: 000000001fce5000 RSI: 0000000000000000 RDI: 000001001fe00068
RBP: 0000000080111213 R08: 0000000000000000 R09: 000001001fe27e54
R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000
R13: 000001001f999148 R14: 000001001f8b92b0 R15: 0000000000000000
FS:  0000000000000000(0000) GS:ffffffff80551b40(0000) knlGS:0000000000000000
CS:  0010 DS: 0018 ES: 0018 CR0: 0000000080050033
CR2: 000001001fdfff38 CR3: 0000000000101000 CR4: 00000000000006e0

Process hibernate.sh (pid: 18616, threadinfo 000001000aad8000, task 000001001f8b92b0)

(that's all I get, unfortunately). It does not occur every time, but fairly often (every 3-4 suspends or so).

The .config is available at:
http://www.sisk.pl/kernel/040913/2.6.9-rc1-mm5.config

The output of dmesg is available at:
http://www.sisk.pl/kernel/040913/2.6.9-rc1-mm5-dmesg.log

Greets,
RJW

Signature · 3 lines
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
		-- Lewis Carroll "Alice's Adventures in Wonderland"
#1 lore raw

Hi!

Quoted 1 line

JFYI, I get a double fault on resume on an Athlon64-based box:

Try the patch I sent to lkml few minutes ago.
Pavel

Signature · 2 lines
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!
#2 lore raw
Quoted 5 lines · Pavel Machek

Hi!

Quoted 1 line

JFYI, I get a double fault on resume on an Athlon64-based box:

Try the patch I sent to lkml few minutes ago.

Unfortunately, on 2.6.9-rc2-mm1 I'm still getting things like this:

Relocating pagedir .....:::::|
Reading image data (10973 pages): 100% 10973 done. Stopping tasks: ===|
Freeing memory... done (0 pages freed)
PM: Restoring saved image.
<0>double fault: 0000 [1]
CPU 0

Modules linked in: usbserial parport_pc lp parport joydev sg sd_mod sr_mod 
scsi_mod cpufreq_userspace powernow_k8 thermal processor fd
Pid: 12135, comm: bash Not tainted 2.6.9-rc2-mm1
RIP: 0010:[<ffffffff8012501d>] <ffffffff8012501d>{do_page_fault+1453}
RSP: 0000:000001001fe00008  EFLAGS: 00010006
RAX: 000000001fce9748 RBX: 0000000000000001 RCX: 000ffffffffff000
RDX: 000000001fce9000 RSI: 0000010000000000 RDI: 000001001fe000e8
RBP: 000001001fce9748 R08: 0000000000000000 R09: 000001001fe27e54
R10: 000000000000279d R11: 0000000000000000 R12: 0000000000000000
R13: 0000000000000000 R14: 000001001c0266e0 R15: 000001001fe000e8
FS:  0000000000000000(0000) GS:ffffffff80552b40(0000) knlGS:0000000000000000
CS:  0010 DS: 0018 ES: 0018 CR0: 0000000080050033
CR2: 000001001fdffff8 CR3: 0000000000101000 CR4: 00000000000006e0
Process bash (pid: 12135, threadinfo 0000010014062000, task 000001001a517250)
Stack: 000001001fe01000 0000000000000060 000001001fe00060 000001001a517250
       fffefffefffefffe fffefffe00030001 fffefffefffefffe fffefffefffefffe
       fffefffefffefffe fffefffefffefffe
Call Trace:<ffffffff8011154d>{error_exit+0} 
<ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} <ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} 
<ffffffff8027f5d7>{acpi_os_allocate+12}
       <ffffffff8011154d>{error_exit+0} <ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8027f5d7>{acpi_os_allocate+12}
       <ffffffff8011154d>{error_exit+0} <ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} 
<ffffffff8027f5d7>{acpi_os_allocate+12}
       <ffffffff8011154d>{error_exit+0} <ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} 
<ffffffff8027f5d7>{acpi_os_allocate+12}
       <ffffffff8011154d>{error_exit+0} <ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} 
<ffffffff8027f5d7>{acpi_os_allocate+12}
       <ffffffff8011154d>{error_exit+0} <ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} 
<ffffffff8027f5d7>{acpi_os_allocate+12}
       <ffffffff8011154d>{error_exit+0} <ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8011154d>{error_exit+0} <ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} 
<ffffffff8027f5d7>{acpi_os_allocate+12}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} 
<ffffffff8027f5d7>{acpi_os_allocate+12}
       <ffffffff8011154d>{error_exit+0} <ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} 
<ffffffff8027f5d7>{acpi_os_allocate+12}
       <ffffffff8011154d>{error_exit+0} <ffffffff8012501d>{do_page_fault+1453}
       <ffffffff8027f5d7>{acpi_os_allocate+12} 
<ffffffff8011154d>{error_exit+0}
       <ffffffff8012501d>{do_page_fault+1453} 

<ffffffff8027f5d7>{acpi_os_allocate+12}

Code: f6 04 06 81 0f 84 52 fc ff ff 48 8b 05 52 ab 3a 00 48 89 c2

RIP <ffffffff8012501d>{do_page_fault+1453} RSP <000001001fe00008>

every second suspend on the average. I don't even start X, I just suspend the box from a console.

Greets,
RJW

Signature · 3 lines
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
		-- Lewis Carroll "Alice's Adventures in Wonderland"
#3 lore raw
Quoted 8 lines · Rafael J. Wysocki
Quoted 5 lines · Pavel Machek

Hi!

Quoted 1 line

JFYI, I get a double fault on resume on an Athlon64-based box:

Try the patch I sent to lkml few minutes ago.

Unfortunately, on 2.6.9-rc2-mm1 I'm still getting things like this:

[- snip -]

I should have said it's with your patch applied. Sorry,

RJW

Signature · 3 lines
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
		-- Lewis Carroll "Alice's Adventures in Wonderland"