* kernel boot error.
@ 2003-02-05 0:41 Jason Ormes
2003-02-05 0:43 ` Andrew Clausen
0 siblings, 1 reply; 11+ messages in thread
From: Jason Ormes @ 2003-02-05 0:41 UTC (permalink / raw)
To: linux-mips
hello,
can someone help me with this error? Is this because the network failed?
scsi0 : QLogic ISP1020 SCSI on PCI bus 00 device 00 irq 4 I/O base 0x8200000
scsi1 : QLogic ISP1020 SCSI on PCI bus 00 device 08 irq 5 I/O base 0x8400000
eth0: Auto-Negotiation unsuccessful, trying force link mode
Slice A got dbe at 0xffffffff801882fc
Hub information:
ERR_INT_PEND = 0x200100
Hub has valid error information:
Overrun is set. Error stack may contain additional information.
Hub error address is 02400208
Incoming message command 0x9e
Supplemental field of incoming message is 0x7f8
T5 Rn (for RRB only) is 0x0
Error type is Uncached Partial Read PRERR
Cpu 0
$0 : 0000000000000000 0000000014001ce0 a8000000013b1040 a8000000013b1000
$4 : a8000000013b1080 0000000000000000 0000000000000040 ffffffff80284688
$8 : 0000000014001ce1 0000000000000000 a8000000012d3c70 0000000000000004
$12 : 0000000000000000 a8000000013b1080 a8000000012d3c80 ffffffff8026e468
$16 : a8000000013b1040 0000000000000002 a8000000013b0400 a8000000003840c0
$20 : a800000000384000 0000000000000040 0000000000000001 a8000000013a7e00
$24 : 0000000000000040 0000000000000020
$28 : a8000000012d0000 a8000000012d3af0 0000000000000000 ffffffff8010ecd8
Hi : 0000000000000000
Lo : 0000000000000040
epc : ffffffff801882fc Not tainted
badvaddr: 0000000000000000
Status : 14001ce2 [ KX SX UX KERNEL EXL ]
Cause : 0000901c
Index: 0 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 1 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 2 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 3 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 4 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 5 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 6 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 7 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 8 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 9 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 10 pgmask=4kb va=c0000fff80000000 asid=00
[pa=00000000000 c=0 d=0 v=0 g=0] [pa=00000000000 c=0 d=0 v=0 g=0]
Index: 11 pgmask=4kb va=c0000fff80000000 asid=00
Jason Ormes
^ permalink raw reply [flat|nested] 11+ messages in thread* Re: kernel boot error.
2003-02-05 0:41 kernel boot error Jason Ormes
@ 2003-02-05 0:43 ` Andrew Clausen
2003-02-05 1:12 ` Jason Ormes
` (2 more replies)
0 siblings, 3 replies; 11+ messages in thread
From: Andrew Clausen @ 2003-02-05 0:43 UTC (permalink / raw)
To: Jason Ormes; +Cc: linux-mips
On Tue, Feb 04, 2003 at 06:41:10PM -0600, Jason Ormes wrote:
> hello,
>
> can someone help me with this error? Is this because the network failed?
I'm getting exactly the same problem. What machine are you using?
I'm using an ip27 (origin 200), and an acenic network card.
It seems that there all kinds of PCI hacks in the ip27 support,
and I'm currently trying to figure out how to get this card working...
Cheers,
Andrew
^ permalink raw reply [flat|nested] 11+ messages in thread* Re: kernel boot error.
2003-02-05 0:43 ` Andrew Clausen
@ 2003-02-05 1:12 ` Jason Ormes
2003-02-05 1:14 ` Andrew Clausen
2003-02-05 1:37 ` Tibor Polgar
2003-02-05 16:48 ` Ralf Baechle
2 siblings, 1 reply; 11+ messages in thread
From: Jason Ormes @ 2003-02-05 1:12 UTC (permalink / raw)
To: Andrew Clausen; +Cc: linux-mips
Its an ip27 origin 200 with what hinv -v reports as a
adapter IOC3 Rev 1, (pci id 2)
controller multi function SuperIO
controller Ethernet Rev 1
I did a little searching online
anhttp://www.scd.ucar.edu/nets/docs/procs/SGI-100mbps/SGI-auto.htmld found a
lot of references to origins having problems with the autonegotiation timing
out to fast, but the only fix that I've found has to do with editing part of
the kernel. here's a link to one that I found.
http://www.scd.ucar.edu/nets/docs/procs/SGI-100mbps/SGI-auto.html
could this be part of the problem?
Thanks for the quick response, atleast I know I'm not alone.
Jason
On Tuesday 04 February 2003 06:43 pm, Andrew Clausen wrote:
> On Tue, Feb 04, 2003 at 06:41:10PM -0600, Jason Ormes wrote:
> > hello,
> >
> > can someone help me with this error? Is this because the network failed?
>
> I'm getting exactly the same problem. What machine are you using?
> I'm using an ip27 (origin 200), and an acenic network card.
>
> It seems that there all kinds of PCI hacks in the ip27 support,
> and I'm currently trying to figure out how to get this card working...
>
> Cheers,
> Andrew
^ permalink raw reply [flat|nested] 11+ messages in thread* Re: kernel boot error.
2003-02-05 1:12 ` Jason Ormes
@ 2003-02-05 1:14 ` Andrew Clausen
0 siblings, 0 replies; 11+ messages in thread
From: Andrew Clausen @ 2003-02-05 1:14 UTC (permalink / raw)
To: Jason Ormes; +Cc: linux-mips
On Tue, Feb 04, 2003 at 07:12:28PM -0600, Jason Ormes wrote:
> Its an ip27 origin 200 with what hinv -v reports as a
>
> adapter IOC3 Rev 1, (pci id 2)
> controller multi function SuperIO
> controller Ethernet Rev 1
>
> I did a little searching online
> anhttp://www.scd.ucar.edu/nets/docs/procs/SGI-100mbps/SGI-auto.htmld found a
> lot of references to origins having problems with the autonegotiation timing
> out to fast, but the only fix that I've found has to do with editing part of
> the kernel. here's a link to one that I found.
> http://www.scd.ucar.edu/nets/docs/procs/SGI-100mbps/SGI-auto.html
>
> could this be part of the problem?
I doubt it. This problem is a PCI bus issue. I'm still
investigating... but it could be a multitude of things.
Have a look in mips/sgi-ip27/ip27-pci.c at all the functions
with "fixup" in their name. Things like byte-swapping, configuring
IO addresses, etc. I suspect we need something like this for
other cards.
Cheers,
Andrew
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: kernel boot error.
2003-02-05 0:43 ` Andrew Clausen
2003-02-05 1:12 ` Jason Ormes
@ 2003-02-05 1:37 ` Tibor Polgar
2003-02-05 3:06 ` Andrew Clausen
2003-02-05 16:48 ` Ralf Baechle
2 siblings, 1 reply; 11+ messages in thread
From: Tibor Polgar @ 2003-02-05 1:37 UTC (permalink / raw)
To: Andrew Clausen; +Cc: Jason Ormes, linux-mips
> I'm getting exactly the same problem. What machine are you using?
> I'm using an ip27 (origin 200), and an acenic network card.
>
> It seems that there all kinds of PCI hacks in the ip27 support,
> and I'm currently trying to figure out how to get this card working...
My buddy and I used to work at Alteon with my buddy doing most of the original
NIC firmware coding. I don't remember who did the SGI driver side coding.
The linux driver was done by Jes Sorensen using our OpenDriver kit. Let me
know if we can help in any way. Is the Origin an SGI machine? If so, i do
recall we had to do some special casing to get the card to work correctly.
This was 4 years ago ....
Tibor
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: kernel boot error.
2003-02-05 1:37 ` Tibor Polgar
@ 2003-02-05 3:06 ` Andrew Clausen
2003-02-05 14:03 ` Ralf Baechle
0 siblings, 1 reply; 11+ messages in thread
From: Andrew Clausen @ 2003-02-05 3:06 UTC (permalink / raw)
To: Tibor Polgar; +Cc: Jason Ormes, linux-mips
On Tue, Feb 04, 2003 at 05:37:00PM -0800, Tibor Polgar wrote:
> > I'm getting exactly the same problem. What machine are you using?
> > I'm using an ip27 (origin 200), and an acenic network card.
> >
> > It seems that there all kinds of PCI hacks in the ip27 support,
> > and I'm currently trying to figure out how to get this card working...
>
> My buddy and I used to work at Alteon with my buddy doing most of the original
> NIC firmware coding. I don't remember who did the SGI driver side coding.
> The linux driver was done by Jes Sorensen using our OpenDriver kit. Let me
> know if we can help in any way. Is the Origin an SGI machine?
Yep. It's sitting in an SGI machine room.
> If so, i do
> recall we had to do some special casing to get the card to work correctly.
Yeah, that would be right. Have you had a look at pci_fixup_ioc3()?
(That's the network card that seems to come with the Origin 200). I
bet it's something similar.
Just, the base address the card (PCI bus?) is spitting out is very odd:
eth0: SGI AceNIC Gigabit Ethernet at 0xfe7fc000, irq 8
The card is in slot 6, so I'd expect the base address to be 0x8900000.
Anyway, it dies on this:
writel(HW_RESET | (HW_RESET << 24), ®s->HostCtrl);
with:
®s->HostCtrl=9200000008900040
or
®s->HostCtrl=92000000fe7fc040
(depending on the base address... I hard coded in a more
sane one, but it still crashes)
Cheers,
Andrew
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: kernel boot error.
2003-02-05 3:06 ` Andrew Clausen
@ 2003-02-05 14:03 ` Ralf Baechle
2003-02-06 0:34 ` Andrew Clausen
0 siblings, 1 reply; 11+ messages in thread
From: Ralf Baechle @ 2003-02-05 14:03 UTC (permalink / raw)
To: Andrew Clausen; +Cc: Tibor Polgar, Jason Ormes, linux-mips
On Wed, Feb 05, 2003 at 02:06:25PM +1100, Andrew Clausen wrote:
> > If so, i do
> > recall we had to do some special casing to get the card to work correctly.
>
> Yeah, that would be right. Have you had a look at pci_fixup_ioc3()?
> (That's the network card that seems to come with the Origin 200). I
> bet it's something similar.
Pci_fixup_ioc3() is only necessary for the IOC3 nic. It's a PCI board
that's about as broken are it only can be. The board runs on in PCI busses
clocked at 33MHz. It only partially decodes the PCI config address space.
Attempted access to one of the nimplemented registers of the IOC3 will
result in access to another register. That's too buggy for any OS to cope
with without that special kludge pci_fixup_ioc3.
> Just, the base address the card (PCI bus?) is spitting out is very odd:
>
> eth0: SGI AceNIC Gigabit Ethernet at 0xfe7fc000, irq 8
>
> The card is in slot 6, so I'd expect the base address to be 0x8900000.
> Anyway, it dies on this:
Query the address using the usual Linux PCI bus stuff from <linux/pci.h>.
Anything else is doomed, especially guessing ...
Ralf
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: kernel boot error.
2003-02-05 14:03 ` Ralf Baechle
@ 2003-02-06 0:34 ` Andrew Clausen
0 siblings, 0 replies; 11+ messages in thread
From: Andrew Clausen @ 2003-02-06 0:34 UTC (permalink / raw)
To: Ralf Baechle; +Cc: Tibor Polgar, Jason Ormes, linux-mips
On Wed, Feb 05, 2003 at 03:03:40PM +0100, Ralf Baechle wrote:
> > Just, the base address the card (PCI bus?) is spitting out is very odd:
> >
> > eth0: SGI AceNIC Gigabit Ethernet at 0xfe7fc000, irq 8
> >
> > The card is in slot 6, so I'd expect the base address to be 0x8900000.
> > Anyway, it dies on this:
>
> Query the address using the usual Linux PCI bus stuff from <linux/pci.h>.
> Anything else is doomed, especially guessing ...
That stuff is returning 0xfe7fc000. That is sane for an Intel
configuration, but totally insane for Origins, and has no chance
of working.
Cheers,
Andrew
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: kernel boot error.
2003-02-05 0:43 ` Andrew Clausen
2003-02-05 1:12 ` Jason Ormes
2003-02-05 1:37 ` Tibor Polgar
@ 2003-02-05 16:48 ` Ralf Baechle
2 siblings, 0 replies; 11+ messages in thread
From: Ralf Baechle @ 2003-02-05 16:48 UTC (permalink / raw)
To: Andrew Clausen; +Cc: Jason Ormes, linux-mips
On Wed, Feb 05, 2003 at 11:43:45AM +1100, Andrew Clausen wrote:
> On Tue, Feb 04, 2003 at 06:41:10PM -0600, Jason Ormes wrote:
> > hello,
> >
> > can someone help me with this error? Is this because the network failed?
>
> I'm getting exactly the same problem. What machine are you using?
> I'm using an ip27 (origin 200), and an acenic network card.
>
> It seems that there all kinds of PCI hacks in the ip27 support,
> and I'm currently trying to figure out how to get this card working...
His particular machine is a uniprocessor machine, a very rare configuration.
In all the years I'm working with Origins this is just the second I
encounter. Note that disabling one of the processor doesn't suffice;
this problem really only seems to hit machines with one physical processor.
Ralf
^ permalink raw reply [flat|nested] 11+ messages in thread
[parent not found: <ralf@linux-mips.org>]
* Re: kernel boot error.
[not found] <ralf@linux-mips.org>
@ 2003-02-05 17:07 ` Tor Arntsen
2003-02-06 11:25 ` Ralf Baechle
0 siblings, 1 reply; 11+ messages in thread
From: Tor Arntsen @ 2003-02-05 17:07 UTC (permalink / raw)
To: linux-mips
On Feb 5, 17:49, Ralf Baechle wrote:
>> On Tue, Feb 04, 2003 at 06:41:10PM -0600, Jason Ormes wrote:
>> I'm using an ip27 (origin 200), and an acenic network card.
[...]
>His particular machine is a uniprocessor machine, a very rare configuration.
>In all the years I'm working with Origins this is just the second I
>encounter. Note that disabling one of the processor doesn't suffice;
>this problem really only seems to hit machines with one physical processor.
1-CPU Origin-200's are not that uncommon.. I have one right here. Although
we mostly used to buy 2-CPU Origins we also bought 1-CPU Origins now and
then, depending on what we planned to use them for. It's all about price
in the end.
-Tor
^ permalink raw reply [flat|nested] 11+ messages in thread* Re: kernel boot error.
2003-02-05 17:07 ` Tor Arntsen
@ 2003-02-06 11:25 ` Ralf Baechle
0 siblings, 0 replies; 11+ messages in thread
From: Ralf Baechle @ 2003-02-06 11:25 UTC (permalink / raw)
To: Tor Arntsen; +Cc: linux-mips
On Wed, Feb 05, 2003 at 06:07:49PM +0100, Tor Arntsen wrote:
> >His particular machine is a uniprocessor machine, a very rare configuration.
> >In all the years I'm working with Origins this is just the second I
> >encounter. Note that disabling one of the processor doesn't suffice;
> >this problem really only seems to hit machines with one physical processor.
>
> 1-CPU Origin-200's are not that uncommon.. I have one right here. Although
> we mostly used to buy 2-CPU Origins we also bought 1-CPU Origins now and
> then, depending on what we planned to use them for. It's all about price
> in the end.
Maybe my machine room at SGI was just equipped to well :-)
Ralf
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2003-02-06 11:26 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-02-05 0:41 kernel boot error Jason Ormes
2003-02-05 0:43 ` Andrew Clausen
2003-02-05 1:12 ` Jason Ormes
2003-02-05 1:14 ` Andrew Clausen
2003-02-05 1:37 ` Tibor Polgar
2003-02-05 3:06 ` Andrew Clausen
2003-02-05 14:03 ` Ralf Baechle
2003-02-06 0:34 ` Andrew Clausen
2003-02-05 16:48 ` Ralf Baechle
[not found] <ralf@linux-mips.org>
2003-02-05 17:07 ` Tor Arntsen
2003-02-06 11:25 ` Ralf Baechle
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox