Skip to contentSkip to conversation

2.4.26 intermittent kernel bug on boot.

DISCUSSION Michael Frank 3 participants 4 messages Apr 16, 2004 – Apr 19, 2004

lore
Share & tools
lore
#0 lore raw

2.4.26 patched with swsusp2.

P4, SIS5513 chipset about the 7th boot of 2.4.26.

Linux version 2.4.26-mhf197 (root@mhfl4) (gcc version 2.95.3 20010315 (release)) #3 Fri Apr 16 22:23:11 HKT 2004 BIOS-provided physical RAM map:
BIOS-e820: 0000000000000000 - 000000000009fc00 (usable)
BIOS-e820: 000000000009fc00 - 00000000000a0000 (reserved)
BIOS-e820: 00000000000f0000 - 0000000000100000 (reserved)
BIOS-e820: 0000000000100000 - 000000001f7f0000 (usable)
BIOS-e820: 000000001f7f0000 - 000000001f7f3000 (ACPI NVS)
BIOS-e820: 000000001f7f3000 - 000000001f800000 (ACPI data)
BIOS-e820: 00000000fec00000 - 0000000100000000 (reserved)
503MB LOWMEM available.
On node 0 totalpages: 129008
zone(0): 4096 pages.
zone(1): 124912 pages.
zone(2): 0 pages.
Kernel command line: vga=0xf07 root=/dev/hda4 resume2=swap:/dev/hda1 console=tty0 console=ttyS0,115200n8r devfs=nomount nousb acpi=off Initializing CPU#0
Detected 2399.777 MHz processor.
Console: colour VGA+ 80x60
kernel BUG at slab.c:1238!
invalid operand: 0000

CPU:    0
EIP:    1010:[<c013bf39>]    Not tainted
EFLAGS: 00010002

EIP is at kmem_cache_alloc+0x31/0xdc [kernel]
eax: 00000000 ebx: 00000008 ecx: 00000000 edx: 00000001
esi: 00000000 edi: 0000000b ebp: 00000020 esp: c0407f10
ds: 1018 es: 1018 ss: 1018
Process swapper (pid: 0, stackpage=c0407000)

Stack: 0000000b 00000001 0000000b c040656c c0121f8c 00000000 00000020 c0406000
        0000000b 0000000b 00000000 c012206d 0000000b 00000001 c040656c 00000282
        0000000b c0406000 c012211c 0000000b 00000001 c0406000 00000286 c0406000
Call Trace:
  [<c0121f8c>] send_signal+0x2c/0xf0 [kernel]
  [<c012206d>] deliver_signal+0x1d/0x54 [kernel]
  [<c012211c>] send_sig_info+0x78/0x88 [kernel]
  [<c01221a9>] force_sig_info+0x7d/0x88 [kernel]
  [<c010a5e4>] do_double_fault+0x0/0x64 [kernel]
  [<c01223d5>] force_sig+0x11/0x18 [kernel]
  [<c010a619>] do_double_fault+0x35/0x64 [kernel]
  [<c0109dc4>] error_code+0x34/0x40 [kernel]

Code: 0f 0b d6 04 e0 b7 2b c0 8d 5e 08 9c 5f fa 8b 4e 08 39 d9 75

Entering kdb (current=0xc0406000, pid 0) Oops: invalid operand due to oops @ 0xc013bf39
eax = 0x00000000 ebx = 0x00000008 ecx = 0x00000000 edx = 0x00000001 esi = 0x00000000 edi = 0x0000000b esp = 0xc0407f10 eip = 0xc013bf39 ebp = 0x00000020 xss = 0x00001018 xcs = 0x00001010 eflags = 0x00010002 xds = 0x00001018 xes = 0x00001018 origeax = 0xffffffff &regs = 0xc0407edc kdb>

Board did 50K+ boots testing swsusp with 2.4.2[012345]...

It's still sitting in kdb, if you want me lookup something let me know.

#1 lore raw

Hi.

Quoted 1 line · Michael Frank

kernel BUG at slab.c:1238!

That's a really strange oops to see. It's testing that the GFP flags match the slab's flags. To get an oops there, you'd have to have a non-dma slab (which makes sense), but you've called the kmem_cache_alloc routine with a DMA flag. Line 444 of kernel/signal.c clearly doesn't do that! Could the args be being corrupted while being passed? What does a backtrace look like?

Nigel

Quoted 42 lines

invalid operand: 0000

CPU:    0
EIP:    1010:[<c013bf39>]    Not tainted
EFLAGS: 00010002

EIP is at kmem_cache_alloc+0x31/0xdc [kernel]
eax: 00000000 ebx: 00000008 ecx: 00000000 edx: 00000001
esi: 00000000 edi: 0000000b ebp: 00000020 esp: c0407f10
ds: 1018 es: 1018 ss: 1018
Process swapper (pid: 0, stackpage=c0407000)

Stack: 0000000b 00000001 0000000b c040656c c0121f8c 00000000 00000020 c0406000
        0000000b 0000000b 00000000 c012206d 0000000b 00000001 c040656c 00000282
        0000000b c0406000 c012211c 0000000b 00000001 c0406000 00000286 c0406000
Call Trace:
  [<c0121f8c>] send_signal+0x2c/0xf0 [kernel]
  [<c012206d>] deliver_signal+0x1d/0x54 [kernel]
  [<c012211c>] send_sig_info+0x78/0x88 [kernel]
  [<c01221a9>] force_sig_info+0x7d/0x88 [kernel]
  [<c010a5e4>] do_double_fault+0x0/0x64 [kernel]
  [<c01223d5>] force_sig+0x11/0x18 [kernel]
  [<c010a619>] do_double_fault+0x35/0x64 [kernel]
  [<c0109dc4>] error_code+0x34/0x40 [kernel]

Code: 0f 0b d6 04 e0 b7 2b c0 8d 5e 08 9c 5f fa 8b 4e 08 39 d9 75

Entering kdb (current=0xc0406000, pid 0) Oops: invalid operand due to oops @ 0xc013bf39
eax = 0x00000000 ebx = 0x00000008 ecx = 0x00000000 edx = 0x00000001 esi = 0x00000000 edi = 0x0000000b esp = 0xc0407f10 eip = 0xc013bf39 ebp = 0x00000020 xss = 0x00001018 xcs = 0x00001010 eflags = 0x00010002 xds = 0x00001018 xes = 0x00001018 origeax = 0xffffffff &regs = 0xc0407edc kdb>

Board did 50K+ boots testing swsusp with 2.4.2[012345]...

It's still sitting in kdb, if you want me lookup something let me know.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Signature · 7 lines
Nigel & Michelle Cunningham
C/- Westminster Presbyterian Church Belconnen
61 Templeton Street, Cook, ACT 2614.
+61 (2) 6251 7727(wk); +61 (2) 6254 0216 (home)

Evolution (n): A hypothetical process whereby infinitely improbable events occur 
with alarming frequency, order arises from chaos, and no one is given credit.
#2 lore raw
Quoted 10 lines · Nigel Cunningham

Hi.

Quoted 1 line · Michael Frank

kernel BUG at slab.c:1238!

That's a really strange oops to see. It's testing that the GFP flags match the slab's flags. To get an oops there, you'd have to have a non-dma slab (which makes sense), but you've called the kmem_cache_alloc routine with a DMA flag. Line 444 of kernel/signal.c clearly doesn't do that! Could the args be being corrupted while being passed?

I had sometimes hangs while calibrating delay loop with earlier kernels, Alan Cox suggested SMI screwing things up. This bug is earlier and new.

Quoted 2 lines

What does a backtrace look like?

Too bad, kdb not fully init, modules not loaded, so no BT command.

Michael

#3 lore raw
In reply to Nigel Cunningham

Silly question: Can you reproduce it without swsusp2?

Quoted 70 lines · Nigel Cunningham

Hi.

Quoted 1 line · Michael Frank

kernel BUG at slab.c:1238!

That's a really strange oops to see. It's testing that the GFP flags match the slab's flags. To get an oops there, you'd have to have a non-dma slab (which makes sense), but you've called the kmem_cache_alloc routine with a DMA flag. Line 444 of kernel/signal.c clearly doesn't do that! Could the args be being corrupted while being passed? What does a backtrace look like?

Nigel

Quoted 42 lines

invalid operand: 0000

CPU:    0
EIP:    1010:[<c013bf39>]    Not tainted
EFLAGS: 00010002

EIP is at kmem_cache_alloc+0x31/0xdc [kernel]
eax: 00000000 ebx: 00000008 ecx: 00000000 edx: 00000001
esi: 00000000 edi: 0000000b ebp: 00000020 esp: c0407f10
ds: 1018 es: 1018 ss: 1018
Process swapper (pid: 0, stackpage=c0407000)

Stack: 0000000b 00000001 0000000b c040656c c0121f8c 00000000 00000020 c0406000
        0000000b 0000000b 00000000 c012206d 0000000b 00000001 c040656c 00000282
        0000000b c0406000 c012211c 0000000b 00000001 c0406000 00000286 c0406000
Call Trace:
  [<c0121f8c>] send_signal+0x2c/0xf0 [kernel]
  [<c012206d>] deliver_signal+0x1d/0x54 [kernel]
  [<c012211c>] send_sig_info+0x78/0x88 [kernel]
  [<c01221a9>] force_sig_info+0x7d/0x88 [kernel]
  [<c010a5e4>] do_double_fault+0x0/0x64 [kernel]
  [<c01223d5>] force_sig+0x11/0x18 [kernel]
  [<c010a619>] do_double_fault+0x35/0x64 [kernel]
  [<c0109dc4>] error_code+0x34/0x40 [kernel]

Code: 0f 0b d6 04 e0 b7 2b c0 8d 5e 08 9c 5f fa 8b 4e 08 39 d9 75

Entering kdb (current=0xc0406000, pid 0) Oops: invalid operand due to oops @ 0xc013bf39
eax = 0x00000000 ebx = 0x00000008 ecx = 0x00000000 edx = 0x00000001 esi = 0x00000000 edi = 0x0000000b esp = 0xc0407f10 eip = 0xc013bf39 ebp = 0x00000020 xss = 0x00001018 xcs = 0x00001010 eflags = 0x00010002 xds = 0x00001018 xes = 0x00001018 origeax = 0xffffffff &regs = 0xc0407edc kdb>

Board did 50K+ boots testing swsusp with 2.4.2[012345]...

It's still sitting in kdb, if you want me lookup something let me know.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Signature · 13 lines
Nigel & Michelle Cunningham
C/- Westminster Presbyterian Church Belconnen
61 Templeton Street, Cook, ACT 2614.
+61 (2) 6251 7727(wk); +61 (2) 6254 0216 (home)

Evolution (n): A hypothetical process whereby infinitely improbable events occur 
with alarming frequency, order arises from chaos, and no one is given credit.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/