Skip to contentSkip to conversation

Kernel NMI error

DISCUSSION Richard J Moore 2 participants 2 messages Sep 16, 2003 – Sep 17, 2003

lore
Share & tools
lore
#0 lore raw

NMI can occur for a number fo reasons:

I/O check from a card
Parity error from memory
DMA timeout
Watchdog timer expiry
INT 2 instruction

To investigate this further you need to look back though your Syslog for any messages that might indicate intmittant h/w related errors or watchdog reports.

One way to cause an intermittent NMI is by having power management enabled with devices that don't support it. They see the world caving in when power goes away and sometimes generate an NMI. If you're in a power saving process then you have probably switched to System Management Mode, in which case the NMI will be held pending the processor returning from SMM.

Do you have a crash dump associtated with this problem?

--
Richard J Moore
IBM Linux Technology Centre

Quoted 44 lines · msrinath

Hello Everybody,

Can anyone help me on this?

Recently one of our servers running RedHat linux 7.2 with 2.4.7-10 SMP kernel generated the following log and the system rebooted. This system has 2 CPUs.

Sep 16 01:34:24 cbesc ftpd[30753]: FTP LOGIN FROM 16.128.157.7
[16.128.157.7], scuser
Sep 16 01:36:48 cbesc ftpd[30753]: FTP session closed
Sep 16 01:54:30 cbesc kernel: Uhhuh. NMI received for unknown reason 35.
Sep 16 01:54:30 cbesc kernel: Dazed and confused, but trying to continue
Sep 16 01:54:30 cbesc kernel: Do you have a strange power saving mode
enabled?
Sep 16 01:54:30 cbesc kernel: eth0: card reports no resources.
Sep 16 01:58:09 cbesc syslogd 1.4.1: restart.
Sep 16 01:58:09 cbesc syslog: syslogd startup succeeded

This is the first time we have faced this problem. The ethernet card used is intel eepro 100. The details are shown below.

Sep 16 07:33:53 cbesc kernel: eepro100.c:v1.09j-t 9/29/99 Donald Becker
http://cesdis.gsfc.nasa.gov/linux/drivers/eepro100.html
Sep 16 07:33:53 cbesc kernel: eepro100.c: $Revision: 1.36 $ 2000/11/17
Modified by Andrey V. Savochkin <[email protected]> and others
Sep 16 07:33:53 cbesc kernel: eth0: Intel Corporation 82557 [Ethernet Pro
100], 00:A0:C9:A0:B7:71, IRQ 16.
Sep 16 07:33:53 cbesc kernel:   Receiver lock-up bug exists -- enabling
work-around.
Sep 16 07:33:53 cbesc kernel:   Board assembly 668081-004, Physical
connectors present: RJ45
Sep 16 07:33:53 cbesc kernel:   Primary interface chip i82555 PHY #1.
Sep 16 07:33:54 cbesc kernel:   General self-test: passed.
Sep 16 07:33:54 cbesc kernel:   Serial sub-system self-test: passed.
Sep 16 07:33:54 cbesc kernel:   Internal registers self-test: passed.
Sep 16 07:33:54 cbesc kernel:   ROM checksum self-test: passed

(0x3c15c8f1). Sep 16 07:33:54 cbesc kernel: Receiver lock-up workaround
activated.

Please let me know why this happened and whether it indicates any hardware problem in the system.

Quoted 6 lines

Please send a CC to my email address, since I have not subscribed to the list.

Thanks & Regards,

- Srinath.

-------------------------------------------------------

Signature · 2 lines
Richard J Moore
IBM Linux Technology Centre
#1 lore raw

Thank you very much for the reply.

I went through the syslog and did not find any kernel related error messages.

I don't know much about power saving mode. We have not done anything specific to enable/disable power saving mode or System Management mode.

I am not aware of crash dumps. Please let me know how to get it.

Thanks & Regards
- Srinath

-----Original Message-----
From: Richard J Moore [mailto:[email protected]]
Sent: 16 September 2003 21:54
To: msrinath
Cc: Linux Kernel Mailing List
Subject: Fwd: Re: Kernel NMI error

NMI can occur for a number fo reasons:

I/O check from a card
Parity error from memory
DMA timeout
Watchdog timer expiry
INT 2 instruction

To investigate this further you need to look back though your Syslog for any messages that might indicate intmittant h/w related errors or watchdog reports.

One way to cause an intermittent NMI is by having power management enabled with devices that don't support it. They see the world caving in when power goes away and sometimes generate an NMI. If you're in a power saving process then you have probably switched to System Management Mode, in which case the NMI will be held pending the processor returning from SMM.

Do you have a crash dump associtated with this problem?

--
Richard J Moore
IBM Linux Technology Centre

Quoted 6 lines · msrinath

Hello Everybody,

Can anyone help me on this?

Recently one of our servers running RedHat linux 7.2 with 2.4.7-10 SMP kernel generated the following log and the system rebooted. This system

has

Quoted 38 lines

2 CPUs.

Sep 16 01:34:24 cbesc ftpd[30753]: FTP LOGIN FROM 16.128.157.7
[16.128.157.7], scuser
Sep 16 01:36:48 cbesc ftpd[30753]: FTP session closed
Sep 16 01:54:30 cbesc kernel: Uhhuh. NMI received for unknown reason 35.
Sep 16 01:54:30 cbesc kernel: Dazed and confused, but trying to continue
Sep 16 01:54:30 cbesc kernel: Do you have a strange power saving mode
enabled?
Sep 16 01:54:30 cbesc kernel: eth0: card reports no resources.
Sep 16 01:58:09 cbesc syslogd 1.4.1: restart.
Sep 16 01:58:09 cbesc syslog: syslogd startup succeeded

This is the first time we have faced this problem. The ethernet card used is intel eepro 100. The details are shown below.

Sep 16 07:33:53 cbesc kernel: eepro100.c:v1.09j-t 9/29/99 Donald Becker
http://cesdis.gsfc.nasa.gov/linux/drivers/eepro100.html
Sep 16 07:33:53 cbesc kernel: eepro100.c: $Revision: 1.36 $ 2000/11/17
Modified by Andrey V. Savochkin <[email protected]> and others
Sep 16 07:33:53 cbesc kernel: eth0: Intel Corporation 82557 [Ethernet Pro
100], 00:A0:C9:A0:B7:71, IRQ 16.
Sep 16 07:33:53 cbesc kernel:   Receiver lock-up bug exists -- enabling
work-around.
Sep 16 07:33:53 cbesc kernel:   Board assembly 668081-004, Physical
connectors present: RJ45
Sep 16 07:33:53 cbesc kernel:   Primary interface chip i82555 PHY #1.
Sep 16 07:33:54 cbesc kernel:   General self-test: passed.
Sep 16 07:33:54 cbesc kernel:   Serial sub-system self-test: passed.
Sep 16 07:33:54 cbesc kernel:   Internal registers self-test: passed.
Sep 16 07:33:54 cbesc kernel:   ROM checksum self-test: passed

(0x3c15c8f1). Sep 16 07:33:54 cbesc kernel: Receiver lock-up workaround
activated.

Please let me know why this happened and whether it indicates any hardware problem in the system.

Quoted 6 lines

Please send a CC to my email address, since I have not subscribed to the list.

Thanks & Regards,

- Srinath.

-------------------------------------------------------

--
Richard J Moore
IBM Linux Technology Centre

--
This message has been scanned for viruses and dangerous content by Kaspersky on bpl Server, and is believed to be clean.
bpl www.kaspersky.com
.

Signature · 5 lines
This message has been scanned for viruses and
dangerous content by Kaspersky on bpl Server, and is
believed to be clean.
bpl www.kaspersky.com
.