ata_exec_internal crash at boot in -git22

ata_exec_internal crash at boot in -git22

[Andi Kleen]

One of the systems tested in autoboot crashes at boot with with -git22.
This is a AMD 2 socket Opteron NUMA system.

The tester was a little flakey and happened to hit the x86-merge-broke-
compilation window, so the last good data point I have is 2.6.23-rc9.

-Andi

megasas: 00.00.03.05 Mon Oct 02 11:21:32 PDT 2006
ACPI: PCI Interrupt 0000:03:05.0[A] -> GSI 17 (level, low) -> IRQ 17
ata1: SATA max UDMA/100 cmd 0xFFFFC20000006C80 ctl 0xFFFFC20000006C8A bmdma 0xFFFFC20000006C00 irq 17
ata2: SATA max UDMA/100 cmd 0xFFFFC20000006CC0 ctl 0xFFFFC20000006CCA bmdma 0xFFFFC20000006C08 irq 17
ata3: SATA max UDMA/100 cmd 0xFFFFC20000006E80 ctl 0xFFFFC20000006E8A bmdma 0xFFFFC20000006E00 irq 17
ata4: SATA max UDMA/100 cmd 0xFFFFC20000006EC0 ctl 0xFFFFC20000006ECA bmdma 0xFFFFC20000006E08 irq 17
scsi0 : sata_sil
ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
ata1.00: ATA-7, max UDMA/133, 234441648 sectors: LBA48 NCQ (depth 0/32)
ata1.00: ata1: dev 0 multi count 16
Unable to handle kernel paging request at 0000000000002277 RIP:
 [<ffffffff80259b03>] pfn_to_page+0x1f/0x32
PGD 0
Oops: 0000 [1] SMP
CPU 2
Modules linked in:
Pid: 915, comm: scsi_eh_0 Not tainted 2.6.19-git22 #24
RIP: 0010:[<ffffffff80259b03>]  [<ffffffff80259b03>] pfn_to_page+0x1f/0x32
RSP: 0000:ffff810004fbbbb8  EFLAGS: 00010216
RAX: 0000000000000040 RBX: 0000000000000000 RCX: 0000000000000020
RDX: 000000000000000f RSI: ffff810004fbbc40 RDI: 00000007f0000000
RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000001 R11: ffffffff80352a9b R12: 0000000000000003
R13: 0000000000000000 R14: ffff810004fbbc40 R15: ffff8100f7c2c708
FS:  0000000000000000(0000) GS:ffff8101000eabc0(0000) knlGS:0000000000000000
CS:  0010 DS: 0018 ES: 0018 CR0: 000000008005003b
CR2: 0000000000002277 CR3: 0000000000201000 CR4: 00000000000006e0
Process scsi_eh_0 (pid: 915, threadinfo ffff810004fba000, task ffff8100f7a54870)
Stack:  ffffffff8043f1d1 0000000000000000 0000000000000000 0000000000000000
 0000000000000000 0000000000000000 ffff8100f7c2c508 ffff8100f7c2c708
 ffff8100f7c2c708 0000000000000000 ffff8100f7c2c508 0000000000000000
Call Trace:
 [<ffffffff8043f1d1>] ata_exec_internal+0x5a/0x96
 [<ffffffff8043fab3>] ata_set_mode+0x415/0x564
 [<ffffffff80445b2f>] ata_do_eh+0x11af/0x1493
 [<ffffffff804464a2>] ata_scsi_error+0x26f/0x509
 [<ffffffff80403a67>] scsi_error_handler+0x73/0x67b
 [<ffffffff80242f04>] kthread+0xd1/0x101
 [<ffffffff8020a318>] child_rip+0xa/0x12


Code: 48 2b ba 68 22 00 00 48 6b c7 38 48 03 82 58 22 00 00 c3 48
RIP  [<ffffffff80259b03>] pfn_to_page+0x1f/0x32
 RSP <ffff810004fbbbb8>
CR2: 0000000000002277

Re: ata_exec_internal crash at boot in -git22

[Jens Axboe]

On Mon, Oct 22 2007, Andi Kleen wrote:
> 
> One of the systems tested in autoboot crashes at boot with with -git22.
> This is a AMD 2 socket Opteron NUMA system.
> 
> The tester was a little flakey and happened to hit the x86-merge-broke-
> compilation window, so the last good data point I have is 2.6.23-rc9.

Andi, can you test with this patch applied?

http://brick.kernel.dk/sg-git.patch

-- 
Jens Axboe

Re: ata_exec_internal crash at boot in -git22

[Andi Kleen]

On Monday 22 October 2007 20:26:45 Jens Axboe wrote:
> On Mon, Oct 22 2007, Andi Kleen wrote:
> > 
> > One of the systems tested in autoboot crashes at boot with with -git22.
> > This is a AMD 2 socket Opteron NUMA system.
> > 
> > The tester was a little flakey and happened to hit the x86-merge-broke-
> > compilation window, so the last good data point I have is 2.6.23-rc9.
> 
> Andi, can you test with this patch applied?
> 
> http://brick.kernel.dk/sg-git.patch
Sorry was a mistake (cue: there is no 2.6.23-git22 yet). It turned out
to be an old broken kernel. Current -git seems to boot.
Sorry for the noise.

-Andi

Re: ata_exec_internal crash at boot in -git22

[Jens Axboe]

On Mon, Oct 22 2007, Andi Kleen wrote:
> On Monday 22 October 2007 20:26:45 Jens Axboe wrote:
> > On Mon, Oct 22 2007, Andi Kleen wrote:
> > > 
> > > One of the systems tested in autoboot crashes at boot with with -git22.
> > > This is a AMD 2 socket Opteron NUMA system.
> > > 
> > > The tester was a little flakey and happened to hit the x86-merge-broke-
> > > compilation window, so the last good data point I have is 2.6.23-rc9.
> > 
> > Andi, can you test with this patch applied?
> > 
> > http://brick.kernel.dk/sg-git.patch
> Sorry was a mistake (cue: there is no 2.6.23-git22 yet). It turned out
> to be an old broken kernel. Current -git seems to boot.
> Sorry for the noise.

OK, all for the better :-)

-- 
Jens Axboe