* Short notice: ipmi_si kills C8000
@ 2013-10-26 14:26 Rolf Eike Beer
2013-10-26 17:17 ` Jeroen Roovers
2013-10-31 21:40 ` Rolf Eike Beer
0 siblings, 2 replies; 13+ messages in thread
From: Rolf Eike Beer @ 2013-10-26 14:26 UTC (permalink / raw)
To: linux-parisc
[-- Attachment #1: Type: text/plain, Size: 238 bytes --]
I did "modprobe ipmi_si" on my C8000 (kernel 3.11.6), which immediately broke
the machine. I have no backtrace yet, but I'll try to get one next week. In
case someone has such a machine and the mood to debug this already… ;)
Eike
[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: Short notice: ipmi_si kills C8000
2013-10-26 14:26 Short notice: ipmi_si kills C8000 Rolf Eike Beer
@ 2013-10-26 17:17 ` Jeroen Roovers
2013-10-28 10:56 ` Thomas Bogendoerfer
2013-10-31 21:40 ` Rolf Eike Beer
1 sibling, 1 reply; 13+ messages in thread
From: Jeroen Roovers @ 2013-10-26 17:17 UTC (permalink / raw)
To: linux-parisc; +Cc: Rolf Eike Beer
On Sat, 26 Oct 2013 16:26:03 +0200
Rolf Eike Beer <eike@sf-mail.de> wrote:
> I did "modprobe ipmi_si" on my C8000 (kernel 3.11.6), which
> immediately broke the machine. I have no backtrace yet, but I'll try
> to get one next week. In case someone has such a machine and the mood
> to debug this already=E2=80=A6 ;)
[ 730.964000] ipmi message handler version 39.2
[ 731.092000] IPMI System Interface driver.
[ 731.148000] ipmi_si: Adding default-specified kcs state
machine[ 731.224000] ipmi_si: Trying default-specified kcs state
machine at i/o address 0xca2, slave address 0x0, irq 0 <Cpu0>
e800009800e00000 000000004042ab9c CC_ERR_CHECK_HPMC <Cpu0>
37000f7300e00000 8040004000000000 CC_ERR_CPU_CHECK_SUMMARY <Cpu0>
f600105e00e00000 fffffff0f0c00000 CC_MC_HPMC_MONARCH_SELECTED <Cpu0>
140003b200e00000 000000000000000b CC_ERR_HPMC_STATE_ENTRY <Cpu0>
5600100b00e00000 0000000000000194 CC_MC_OS_HPMC_LEN_ERR <Cpu0>
5600106400e00000 fffffff0f0435110 CC_MC_BR_TO_OS_HPMC_FAILED <Cpu0>
e800009800e00000 000000004042ab9c CC_ERR_CHECK_HPMC <Cpu0>
37000f7300e00000 8040004000000000 CC_ERR_CPU_CHECK_SUMMARY <Cpu0>
4000109f00e00000 0000000000000000 CC_MC_HPMC_INITIATED <Cpu0>
4000101900e00000 0000000000000000 CC_MC_MULTIPLE_HPMCS <Cpu0>
030010d500e00000 0000000000000000 CC_CPU_STOP
Confirmed.
jer
--
To unsubscribe from this list: send the line "unsubscribe linux-parisc"=
in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 13+ messages in thread* Re: Short notice: ipmi_si kills C8000
2013-10-26 17:17 ` Jeroen Roovers
@ 2013-10-28 10:56 ` Thomas Bogendoerfer
2013-10-29 9:37 ` Thomas Bogendoerfer
0 siblings, 1 reply; 13+ messages in thread
From: Thomas Bogendoerfer @ 2013-10-28 10:56 UTC (permalink / raw)
To: Jeroen Roovers; +Cc: linux-parisc, Rolf Eike Beer
On Sat, Oct 26, 2013 at 07:17:20PM +0200, Jeroen Roovers wrote:
> > I did "modprobe ipmi_si" on my C8000 (kernel 3.11.6), which
> > immediately broke the machine. I have no backtrace yet, but I'll try
> > to get one next week. In case someone has such a machine and the mood
> > to debug this already??? ;)
>
> [ 730.964000] ipmi message handler version 39.2
> [ 731.092000] IPMI System Interface driver.
> [ 731.148000] ipmi_si: Adding default-specified kcs state
> machine[ 731.224000] ipmi_si: Trying default-specified kcs state
> machine at i/o address 0xca2, slave address 0x0, irq 0 <Cpu0>
looks like stupid ISA auto probing is turned on, when ipmi is loaded
as a module :-( I'll have a look.
Thomas.
--
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea. [ RFC1925, 2.3 ]
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: Short notice: ipmi_si kills C8000
2013-10-28 10:56 ` Thomas Bogendoerfer
@ 2013-10-29 9:37 ` Thomas Bogendoerfer
2013-10-29 10:41 ` Rolf Eike Beer
0 siblings, 1 reply; 13+ messages in thread
From: Thomas Bogendoerfer @ 2013-10-29 9:37 UTC (permalink / raw)
To: Jeroen Roovers; +Cc: linux-parisc, Rolf Eike Beer
On Mon, Oct 28, 2013 at 11:56:12AM +0100, Thomas Bogendoerfer wrote:
> On Sat, Oct 26, 2013 at 07:17:20PM +0200, Jeroen Roovers wrote:
> > > I did "modprobe ipmi_si" on my C8000 (kernel 3.11.6), which
> > > immediately broke the machine. I have no backtrace yet, but I'll try
> > > to get one next week. In case someone has such a machine and the mood
> > > to debug this already??? ;)
> >
> > [ 730.964000] ipmi message handler version 39.2
> > [ 731.092000] IPMI System Interface driver.
> > [ 731.148000] ipmi_si: Adding default-specified kcs state
> > machine[ 731.224000] ipmi_si: Trying default-specified kcs state
> > machine at i/o address 0xca2, slave address 0x0, irq 0 <Cpu0>
>
> looks like stupid ISA auto probing is turned on, when ipmi is loaded
> as a module :-( I'll have a look.
my patch went into 3.12-rc1 so it's expected that 3.11.6 crashes without it.
Thomas.
--
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea. [ RFC1925, 2.3 ]
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: Short notice: ipmi_si kills C8000
2013-10-29 9:37 ` Thomas Bogendoerfer
@ 2013-10-29 10:41 ` Rolf Eike Beer
2013-10-29 16:41 ` Thomas Bogendoerfer
0 siblings, 1 reply; 13+ messages in thread
From: Rolf Eike Beer @ 2013-10-29 10:41 UTC (permalink / raw)
To: linux-parisc
Am 29.10.2013 10:37, schrieb Thomas Bogendoerfer:
> On Mon, Oct 28, 2013 at 11:56:12AM +0100, Thomas Bogendoerfer wrote:
>> On Sat, Oct 26, 2013 at 07:17:20PM +0200, Jeroen Roovers wrote:
>> > > I did "modprobe ipmi_si" on my C8000 (kernel 3.11.6), which
>> > > immediately broke the machine. I have no backtrace yet, but I'll try
>> > > to get one next week. In case someone has such a machine and the mood
>> > > to debug this already??? ;)
>> >
>> > [ 730.964000] ipmi message handler version 39.2
>> > [ 731.092000] IPMI System Interface driver.
>> > [ 731.148000] ipmi_si: Adding default-specified kcs state
>> > machine[ 731.224000] ipmi_si: Trying default-specified kcs state
>> > machine at i/o address 0xca2, slave address 0x0, irq 0 <Cpu0>
>>
>> looks like stupid ISA auto probing is turned on, when ipmi is loaded
>> as a module :-( I'll have a look.
>
> my patch went into 3.12-rc1 so it's expected that 3.11.6 crashes
> without it.
Oops, sorry, I misread the git log. I've now applied that patch to
3.11.6 and the module loads fine. Am I supposed to see anything in the
sysfs that I can query now?
Eike
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: Short notice: ipmi_si kills C8000
2013-10-29 10:41 ` Rolf Eike Beer
@ 2013-10-29 16:41 ` Thomas Bogendoerfer
2013-10-29 17:14 ` Jeroen Roovers
` (2 more replies)
0 siblings, 3 replies; 13+ messages in thread
From: Thomas Bogendoerfer @ 2013-10-29 16:41 UTC (permalink / raw)
To: Rolf Eike Beer; +Cc: linux-parisc
On Tue, Oct 29, 2013 at 11:41:08AM +0100, Rolf Eike Beer wrote:
> Oops, sorry, I misread the git log. I've now applied that patch to
no problem, maybe it's worth to send the fix to stable now it's
in 3.12
> 3.11.6 and the module loads fine. Am I supposed to see anything in
> the sysfs that I can query now?
I've played only with ipmitool, which gives something like
c8000:~# ipmitool sdr
Systemboard | 30 degrees C | ok
CPU 1 | 43 degrees C | ok
Ambient | 24.50 degrees C | ok
Main +12V | 12.25 Volts | ok
Main -12V | -12.45 Volts | ok
Main +5V | 5.00 Volts | ok
Main 3.3V | 3.35 Volts | ok
Battery +3V | 3.11 Volts | ok
Stdby +3.3V | 3.40 Volts | ok
Main +1.5V | 1.49 Volts | ok
VTT | 1.19 Volts | ok
Main +1.8V | 1.80 Volts | ok
FAN1 SYS | 6360 RPM | ok
FAN2 SYS | 6240 RPM | ok
FAN3 SYS | 6000 RPM | ok
FAN4 SYS | 6240 RPM | ok
FAN5 SYS | 5880 RPM | ok
FAN1 PSU | 2400 RPM | ok
FAN2 PSU | 2400 RPM | ok
SEL Level | 0 unspecified | ok
CPU 1 | 0x80 | ok
PSU | Not Readable | ns
PS1 | 0x02 | ok
FAN1 SYS | 0x01 | ok
FAN2 SYS | 0x01 | ok
FAN3 SYS | 0x01 | ok
FAN4 SYS | 0x01 | ok
FAN5 SYS | 0x01 | ok
FAN1 PSU | 0x01 | ok
FAN2 PSU | 0x01 | ok
Watchdog | 0x00 | ok
iRMC request | 0x00 | ok
System Mgmt SW | Not Readable | ns
local Mon. | 0x02 | ok
Pwr Btn override | 0x01 | ok
Software NMI | 0x00 | ok
System BIOS | Not Readable | ns
iRMC | Not Readable | ns
Thomas.
--
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea. [ RFC1925, 2.3 ]
^ permalink raw reply [flat|nested] 13+ messages in thread* Re: Short notice: ipmi_si kills C8000
2013-10-29 16:41 ` Thomas Bogendoerfer
@ 2013-10-29 17:14 ` Jeroen Roovers
2013-10-29 19:50 ` Helge Deller
2013-11-02 16:18 ` John David Anglin
2 siblings, 0 replies; 13+ messages in thread
From: Jeroen Roovers @ 2013-10-29 17:14 UTC (permalink / raw)
To: linux-parisc; +Cc: Thomas Bogendoerfer, Rolf Eike Beer
On Tue, 29 Oct 2013 17:41:51 +0100
Thomas Bogendoerfer <tsbogend@alpha.franken.de> wrote:
> I've played only with ipmitool, which gives something like
>
> c8000:~# ipmitool sdr
> Systemboard | 30 degrees C | ok
> CPU 1 | 43 degrees C | ok
Yes, that works nicely here, too (in 3.12 rc6).
jer
^ permalink raw reply [flat|nested] 13+ messages in thread* Re: Short notice: ipmi_si kills C8000
2013-10-29 16:41 ` Thomas Bogendoerfer
2013-10-29 17:14 ` Jeroen Roovers
@ 2013-10-29 19:50 ` Helge Deller
2013-10-31 12:06 ` Thomas Bogendoerfer
2013-11-02 16:18 ` John David Anglin
2 siblings, 1 reply; 13+ messages in thread
From: Helge Deller @ 2013-10-29 19:50 UTC (permalink / raw)
To: Thomas Bogendoerfer, Rolf Eike Beer; +Cc: linux-parisc
On 10/29/2013 05:41 PM, Thomas Bogendoerfer wrote:
> On Tue, Oct 29, 2013 at 11:41:08AM +0100, Rolf Eike Beer wrote:
>> Oops, sorry, I misread the git log. I've now applied that patch to
>
> no problem, maybe it's worth to send the fix to stable now it's
> in 3.12
Yes, I think it really makes sense to send this patch to stable kernel series.
Without this patch, any parisc box (ok - probably only the C8000) crashes
as soon as the ipmi driver is loaded.
Will you ask (e.g. if possible for kernel 3.0 and higher - not sure if it applies)?
Helge
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: Short notice: ipmi_si kills C8000
2013-10-29 19:50 ` Helge Deller
@ 2013-10-31 12:06 ` Thomas Bogendoerfer
2013-10-31 12:22 ` Helge Deller
0 siblings, 1 reply; 13+ messages in thread
From: Thomas Bogendoerfer @ 2013-10-31 12:06 UTC (permalink / raw)
To: Helge Deller; +Cc: Rolf Eike Beer, linux-parisc
On Tue, Oct 29, 2013 at 08:50:51PM +0100, Helge Deller wrote:
> Yes, I think it really makes sense to send this patch to stable kernel series.
> Without this patch, any parisc box (ok - probably only the C8000) crashes
> as soon as the ipmi driver is loaded.
>
> Will you ask (e.g. if possible for kernel 3.0 and higher - not sure if
> it applies)?
3.4 shows some fuzz and iirc there were some changes I needed to make for
inclusion into latest kernels. Afaik all other C8k went only into 3.10,
so that should be good enough. And the patch applies cleanly to 3.10.
I'll write mail to the stable team.
Thomas.
--
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea. [ RFC1925, 2.3 ]
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: Short notice: ipmi_si kills C8000
2013-10-31 12:06 ` Thomas Bogendoerfer
@ 2013-10-31 12:22 ` Helge Deller
0 siblings, 0 replies; 13+ messages in thread
From: Helge Deller @ 2013-10-31 12:22 UTC (permalink / raw)
To: Thomas Bogendoerfer; +Cc: Rolf Eike Beer, linux-parisc
On 10/31/2013 01:06 PM, Thomas Bogendoerfer wrote:
> On Tue, Oct 29, 2013 at 08:50:51PM +0100, Helge Deller wrote:
>> Yes, I think it really makes sense to send this patch to stable kernel series.
>> Without this patch, any parisc box (ok - probably only the C8000) crashes
>> as soon as the ipmi driver is loaded.
>>
>> Will you ask (e.g. if possible for kernel 3.0 and higher - not sure if
>> it applies)?
>
> 3.4 shows some fuzz and iirc there were some changes I needed to make for
> inclusion into latest kernels. Afaik all other C8k went only into 3.10,
> so that should be good enough. And the patch applies cleanly to 3.10.
> I'll write mail to the stable team.
If possible >= 3.8 would be good.
I'm not sure Debians main unstable installation kernel is 3.8...
Helge
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: Short notice: ipmi_si kills C8000
2013-10-29 16:41 ` Thomas Bogendoerfer
2013-10-29 17:14 ` Jeroen Roovers
2013-10-29 19:50 ` Helge Deller
@ 2013-11-02 16:18 ` John David Anglin
2 siblings, 0 replies; 13+ messages in thread
From: John David Anglin @ 2013-11-02 16:18 UTC (permalink / raw)
To: Thomas Bogendoerfer; +Cc: Rolf Eike Beer, linux-parisc
On 29-Oct-13, at 12:41 PM, Thomas Bogendoerfer wrote:
> I've played only with ipmitool, which gives something like
>
> c8000:~# ipmitool sdr
> Systemboard | 30 degrees C | ok
> CPU 1 | 43 degrees C | ok
> Ambient | 24.50 degrees C | ok
I played with IPMI a bit on rp3440. It works to a degree but so far I
haven't been able
to get it to show sensor data:
mx3210:/boot# ipmitool mc info
Device ID : 50
Device Revision : 1
Firmware Revision : 3.52
IPMI Version : 1.0
Manufacturer ID : 11
Manufacturer Name : Hewlett-Packard
Product ID : 33281 (0x8201)
Product Name : Unknown (0x8201)
Device Available : yes
Provides Device SDRs : yes
Additional Device Support :
Sensor Device
SDR Repository Device
SEL Device
FRU Inventory Device
IPMB Event Receiver
IPMB Event Generator
mx3210:/boot# ipmitool sdr info
SDR Version : 0x1
Record Count : 92
Free Space : 12336 bytes
Most recent Addition :
Most recent Erase :
SDR overflow : no
SDR Repository Update Support : unspecified
Delete SDR supported : no
Partial Add SDR supported : no
Reserve SDR repository supported : yes
SDR Repository Alloc info supported : yes
mx3210:/boot# ipmitool -v sdr list
Running Get PICMG Properties my_addr 0x20, transit 0, target 0
Error Response 0xc2 from Get PICMG Properities
No PICMG Extenstion discovered
Sensor ID : Chassis Intrus (0x1)
Entity ID : 23.1 (System Chassis)
Sensor Type (Discrete): Physical Security (0x05)
Sensor Reading : No Reading
Event Message Control : Per-threshold
OEM : 0
Get SDR 5000 command failed: Timeout
Get SDR 5000 command failed: Timeout
Get SDR 5000 command failed: Timeout
Get SDR 5000 command failed: Timeout
Get SDR 5000 command failed: Timeout
mx3210:/boot# ipmitool sdr type Temperature
Ambient Temp | D8h | ns | 23.1 | No Reading
Processor 0 Temp | D9h | ns | 3.1 | No Reading
Processor 1 Temp | DAh | ns | 3.2 | No Reading
mx3210:/boot# ipmitool fru print
FRU Device Description : Builtin FRU Device (ID 0)
Chassis Type : Rack Mount Chassis
Board Mfg Date : Mon May 24 14:50:00 2004
Board Mfg : JABIL
Board Product : rp3440 SYSTEM BOARD
Board Serial : 52JAPE4422169260
Board Part Number : A7136-60001
Board Extra : A1
Board Extra : 4419
Board Extra : D
Board Extra : 00000000000000000000000000000000
Product Manufacturer : hp
Product Extra : 09010000
FRU Device Description : I/O Backplane (ID 5)
Device not present (Timeout)
Get SDR 300d command failed: Timeout
Get SDR 300d command failed: Timeout
Get SDR 300d command failed: Timeout
Get SDR 300d command failed: Timeout
Get SDR 300d command failed: Timeout
mx3210:/boot# ipmitool -V
ipmitool version 1.8.13
Thoughts?
Dave
--
John David Anglin dave.anglin@bell.net
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: Short notice: ipmi_si kills C8000
2013-10-26 14:26 Short notice: ipmi_si kills C8000 Rolf Eike Beer
2013-10-26 17:17 ` Jeroen Roovers
@ 2013-10-31 21:40 ` Rolf Eike Beer
2013-11-05 12:23 ` Rolf Eike Beer
1 sibling, 1 reply; 13+ messages in thread
From: Rolf Eike Beer @ 2013-10-31 21:40 UTC (permalink / raw)
To: linux-parisc
[-- Attachment #1: Type: text/plain, Size: 813 bytes --]
Am Samstag, 26. Oktober 2013, 16:26:03 schrieb Rolf Eike Beer:
> I did "modprobe ipmi_si" on my C8000 (kernel 3.11.6), which immediately
> broke the machine. I have no backtrace yet, but I'll try to get one next
> week. In case someone has such a machine and the mood to debug this
> already… ;)
Ok, I now have IPMI monitoring on the C8000 working smoothly. Thanks for the
work on this.
While testing I installed by mistake the nagios IPMI check module and freeipmi
on the C3600 (running 3.11.2 at that time). Calling "check_ipmi_sensor -T
Temperature -H localhost" reproducible caused a HPMC. There was no IPMI module
loaded or even compiled on the machine so I think it has nothing to do with
the crash reported above. I'll try if I can get some additional information on
monday.
Eike
[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: Short notice: ipmi_si kills C8000
2013-10-31 21:40 ` Rolf Eike Beer
@ 2013-11-05 12:23 ` Rolf Eike Beer
0 siblings, 0 replies; 13+ messages in thread
From: Rolf Eike Beer @ 2013-11-05 12:23 UTC (permalink / raw)
To: linux-parisc
>> I did "modprobe ipmi_si" on my C8000 (kernel 3.11.6), which=20
>> immediately
>> broke the machine. I have no backtrace yet, but I'll try to get one=20
>> next
>> week. In case someone has such a machine and the mood to debug this
>> already=E2=80=A6 ;)
>=20
> Ok, I now have IPMI monitoring on the C8000 working smoothly. Thanks=20
> for the
> work on this.
>=20
> While testing I installed by mistake the nagios IPMI check module and=
=20
> freeipmi
> on the C3600 (running 3.11.2 at that time). Calling "check_ipmi_senso=
r=20
> -T
> Temperature -H localhost" reproducible caused a HPMC. There was no IP=
MI=20
> module
> loaded or even compiled on the machine so I think it has nothing to d=
o=20
> with
> the crash reported above. I'll try if I can get some additional=20
> information on
> monday.
I can reproduce this on the C8000, using both 3.11.6 and 3.12: if=20
ipmi_devintf is not loaded the machine will die when nagios does the=20
first IPMI check. Loading this module makes the machine survive. Sadly=20
it seems that I have screwed up my serial setup so I don't have a=20
backtrace of this.
Eike
--
To unsubscribe from this list: send the line "unsubscribe linux-parisc"=
in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 13+ messages in thread
end of thread, other threads:[~2013-11-05 12:23 UTC | newest]
Thread overview: 13+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-10-26 14:26 Short notice: ipmi_si kills C8000 Rolf Eike Beer
2013-10-26 17:17 ` Jeroen Roovers
2013-10-28 10:56 ` Thomas Bogendoerfer
2013-10-29 9:37 ` Thomas Bogendoerfer
2013-10-29 10:41 ` Rolf Eike Beer
2013-10-29 16:41 ` Thomas Bogendoerfer
2013-10-29 17:14 ` Jeroen Roovers
2013-10-29 19:50 ` Helge Deller
2013-10-31 12:06 ` Thomas Bogendoerfer
2013-10-31 12:22 ` Helge Deller
2013-11-02 16:18 ` John David Anglin
2013-10-31 21:40 ` Rolf Eike Beer
2013-11-05 12:23 ` Rolf Eike Beer
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox