* Issue with pata_sil680
@ 2014-01-20 12:24 Guy Martin
2014-01-20 20:12 ` Helge Deller
0 siblings, 1 reply; 3+ messages in thread
From: Guy Martin @ 2014-01-20 12:24 UTC (permalink / raw
To: linux-parisc
Hi,
I tried to switch to the ATA drivers but I keep getting a panic whenever
I load the pata_sil680 module on my C8000.
Kconfig : http://bpaste.net/show/170067/
Kver : 3.12.8-gentoo
This is the backtrace :
insmoterminator ~ # insmod pata_sil680.ko
pata_sil680 0000:60:02.0: version 0.4.9
pata_sil680 0000:60:02.0: sil680: BA5_EN = 1 clock = 00
pata_sil680 0000:60:02.0: sil680: BA5_EN = 1 clock = 10
sil680: 133MHz clock.
Backtrace:
[<00000000401ad0c4>] handle_percpu_irq+0xbc/0xf8
[<00000000401a62c4>] generic_handle_irq+0x4c/0x68
[<00000000401062a4>] syscall_do_resched+0x28/0x34
[<000000004011e1f8>] execute_on_irq_stack+0x70/0xb8
[<000000004019486c>] enqueue_task_fair+0x5ac/0xbd0
[<0000000040188ec8>] check_preempt_curr+0x98/0xc8
[<0000000040188f34>] ttwu_do_wakeup+0x3c/0xd0
[<0000000040189184>] ttwu_do_activate.constprop.68+0xbc/0xd8
[<000000004018c2b8>] try_to_wake_up+0x238/0x450
[<000000004017ff8c>] __raw_notifier_call_chain+0x2c/0x40
[<000000004017ffd0>] raw_notifier_call_chain+0x30/0x48
[<00000000401b0f74>] timekeeping_update.constprop.9+0x5c/0xe0
[<0000000040188f34>] ttwu_do_wakeup+0x3c/0xd0
[<0000000040189184>] ttwu_do_activate.constprop.68+0xbc/0xd8
[<000000004018c2b8>] try_to_wake_up+0x238/0x450
[<000000004018c4fc>] default_wake_function+0x2c/0x40
Bad Address (null pointer deref?): Code=26 regs=0000000041f77350
(Addr=0000000000000000)
CPU: 1 PID: 0 Comm: swapper/1 Not tainted 3.12.8-gentoo #2
task: 000000007fc887e0 ti: 000000007fcc4000 task.ti: 000000007fcc4000
YZrvWESTHLNXBCVMcbcbcbcbOGFRQPDI
PSW: 00001000000001001111111100001110 Not tainted
r00-03 000000ff0804ff0e 0000000040987e60 00000000401ad0c4
0000000041f77250
r04-07 0000000040970660 0000000000000000 0000000000000000
0000000041f77320
r08-11 0000000000000047 00000000407c4880 0000000000000004
0000000000000001
r12-15 0000000040889a90 0000000000000047 0000000000000000
0000000000000001
r16-19 000000007fcc4250 000000000000001a 0000000040889a90
0000000001559000
r20-23 00000000407dec50 0000000041f7b020 fffc000000000000
feffffffffffffff
r24-27 0000000041f77040 0000000000000000 00000000407c4880
0000000040970660
r28-31 fefc000000000000 0000000041f77220 0000000041f77350
0100000000000000
sr00-03 0000000000217000 0000000000217000 0000000000000000
0000000000217000
sr04-07 0000000000000000 0000000000000000 0000000000000000
0000000000000000
IASQ: 0000000000000000 0000000000000000 IAOQ: 00000000401a7234
00000000401a7238
IIR: 0dc010dc ISR: 0000000000000000 IOR: 0000000000000000
CPU: 1 CR30: 000000007fcc4000 CR31: ffffe1691b8f2119
ORIG_R28: 0000000040189184
IAOQ[0]: handle_irq_event_percpu+0x9c/0x378
IAOQ[1]: handle_irq_event_percpu+0xa0/0x378
RP(r2): handle_percpu_irq+0xbc/0xf8
If I hack the driver to enable MMIO, I have the following backtrace :
scsi2 : pata_sil680
scsi3 : pata_sil680
ata1: PATA max UDMA/133 irq 71
ata2: PATA max UDMA/133 irq 71
scsi 3:0:0:0: CD-ROM HL-DT-ST DVDRAM GSA-4082B A208 PQ: 0
ANSI: 5
ata2.00: configured for UDMA/66
sr0: scsi3-mmc drive: 32x/32x writer dvd-ram cd/rw xa/form2 cdda tray
cdrom: Uniform CD-ROM driver Revision: 3.20
sr 3:0:0:0: Attached scsi CD-ROM sr0
sr 3:0:0:0: Attached scsi generic sg1 type 5
[<00000000401ad0c4>] handle_percpu_irq+0xbc/0xf8
[<00000000401a62c4>] generic_handle_irq+0x4c/0x68
[<00000000401062a4>] syscall_do_resched+0x28/0x34
With MMIO, it goes a little bit further as it detects the drive
correctly but eventually fails the same way.
As far as I understand the backtrace, it would seem that an IRQ handler
function pointer is set to NULL or some NULL pointer deref occurs in the
IRQ handler.
Also, this is not the original DVD drive from the C8000. I replaced it
with a DVD burner.
How idea what can be the issue ?
Thanks,
Guy
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: Issue with pata_sil680
2014-01-20 12:24 Issue with pata_sil680 Guy Martin
@ 2014-01-20 20:12 ` Helge Deller
2014-01-21 7:05 ` Guy Martin
0 siblings, 1 reply; 3+ messages in thread
From: Helge Deller @ 2014-01-20 20:12 UTC (permalink / raw
To: Guy Martin, linux-parisc
Hi Guy,
On 01/20/2014 01:24 PM, Guy Martin wrote:
> I tried to switch to the ATA drivers but I keep getting a panic
> whenever I load the pata_sil680 module on my C8000
>
> Kconfig : http://bpaste.net/show/170067/
> Kver : 3.12.8-gentoo
>
> This is the backtrace :
> insmoterminator ~ # insmod pata_sil680.ko
> pata_sil680 0000:60:02.0: version 0.4.9
> pata_sil680 0000:60:02.0: sil680: BA5_EN = 1 clock = 00
> pata_sil680 0000:60:02.0: sil680: BA5_EN = 1 clock = 10
> sil680: 133MHz clock.
> Backtrace:
> [<00000000401ad0c4>] handle_percpu_irq+0xbc/0xf8
> [<00000000401a62c4>] generic_handle_irq+0x4c/0x68
> [<00000000401062a4>] syscall_do_resched+0x28/0x34
> [<000000004011e1f8>] execute_on_irq_stack+0x70/0xb8
....
>
> If I hack the driver to enable MMIO, I have the following backtrace :
> scsi2 : pata_sil680
> scsi3 : pata_sil680
> ata1: PATA max UDMA/133 irq 71
> ata2: PATA max UDMA/133 irq 71
> scsi 3:0:0:0: CD-ROM HL-DT-ST DVDRAM GSA-4082B A208 PQ: 0 ANSI: 5
> ata2.00: configured for UDMA/66
> sr0: scsi3-mmc drive: 32x/32x writer dvd-ram cd/rw xa/form2 cdda tray
> cdrom: Uniform CD-ROM driver Revision: 3.20
> sr 3:0:0:0: Attached scsi CD-ROM sr0
> sr 3:0:0:0: Attached scsi generic sg1 type 5
> [<00000000401ad0c4>] handle_percpu_irq+0xbc/0xf8
> [<00000000401a62c4>] generic_handle_irq+0x4c/0x68
> [<00000000401062a4>] syscall_do_resched+0x28/0x34
>
>
> With MMIO, it goes a little bit further as it detects the drive correctly but eventually fails the same way.
>
> As far as I understand the backtrace, it would seem that an IRQ
> handler function pointer is set to NULL or some NULL pointer deref
> occurs in the IRQ handler. Also, this is not the original DVD drive
> from the C8000. I replaced it with a DVD burner.>
> How idea what can be the issue ?
I'm just guessing here since I would need to test myself.
But James Bottomley found once, that the secondary port of the cmd64x controller
was not completely wired physically up on parisc.
He posted a patch series here:
http://www.spinics.net/lists/linux-ide/msg40816.html
Maybe there is some similarity for pata_sil680 too, esp. since it seems from
the log above that you connected your DVD drive to the secondary port (scsi3).
What happens if you connect it to scsi2/the first pata port instead?
Helge
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: Issue with pata_sil680
2014-01-20 20:12 ` Helge Deller
@ 2014-01-21 7:05 ` Guy Martin
0 siblings, 0 replies; 3+ messages in thread
From: Guy Martin @ 2014-01-21 7:05 UTC (permalink / raw
To: Helge Deller; +Cc: linux-parisc
Hi Helge,
On 2014-01-20 21:12, Helge Deller wrote:
> Hi Guy,
>
> On 01/20/2014 01:24 PM, Guy Martin wrote:
>> I tried to switch to the ATA drivers but I keep getting a panic
>> whenever I load the pata_sil680 module on my C8000
>> With MMIO, it goes a little bit further as it detects the drive
>> correctly but eventually fails the same way.
>>
>> As far as I understand the backtrace, it would seem that an IRQ
>> handler function pointer is set to NULL or some NULL pointer deref
>> occurs in the IRQ handler. Also, this is not the original DVD drive
>> from the C8000. I replaced it with a DVD burner.>
>> How idea what can be the issue ?
>
> I'm just guessing here since I would need to test myself.
> But James Bottomley found once, that the secondary port of the cmd64x
> controller
> was not completely wired physically up on parisc.
> He posted a patch series here:
> http://www.spinics.net/lists/linux-ide/msg40816.html
> Maybe there is some similarity for pata_sil680 too, esp. since it seems
> from
> the log above that you connected your DVD drive to the secondary port
> (scsi3).
> What happens if you connect it to scsi2/the first pata port instead?
I have not checked how it is connected physically yet. It's probably the
jumper on the drive set for slave.
However, this drive works fine with the old IDE siimage driver.
Additionally, if I found out that if I hack the driver to only probe the
first drive,
then unload it, change it to probe both drives and reload it, it is
detected correctly.
Guy
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2014-01-21 7:05 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-01-20 12:24 Issue with pata_sil680 Guy Martin
2014-01-20 20:12 ` Helge Deller
2014-01-21 7:05 ` Guy Martin
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.