Skip to contentSkip to conversation

New: When NETCONSOLE is enabled in kernel, computer crashes after 120seconds (approx)

Bugme-new Andrew Morton 3 participants 3 messages Jun 23, 2009 – Jul 16, 2009 1 message not in the archive

Subsystems (2)
Networking Network drivers
lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]/

Andrew Morton

author
#1 lore raw

(switched to email. Please respond via emailed reply-to-all, not via the bugzilla web interface).

Quoted 34 lines · On Wed, 17 Jun 2009 01:55:54 GMT [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=13553

Summary: When NETCONSOLE is enabled in kernel, computer crashes
after 120seconds (approx)
Product: Networking
Version: 2.5
Kernel Version: 2.6.29.4, 2.6.30
Platform: All
OS/Version: Linux
Tree: Mainline
Status: NEW
Severity: high
Priority: P1
Component: Other
AssignedTo: [email protected]
ReportedBy: [email protected]
Regression: No

00:00.0 Host bridge: Intel Corporation 440GX - 82443GX Host bridge 00:01.0 PCI bridge: Intel Corporation 440GX - 82443GX AGP bridge 00:07.0 ISA bridge: Intel Corporation 82371AB/EB/MB PIIX4 ISA (rev 02) 00:07.1 IDE interface: Intel Corporation 82371AB/EB/MB PIIX4 IDE (rev 01) 00:07.2 USB Controller: Intel Corporation 82371AB/EB/MB PIIX4 USB (rev 01) 00:07.3 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 02) 00:0b.0 SCSI storage controller: Adaptec AIC-7896U2/7897U2 00:0b.1 SCSI storage controller: Adaptec AIC-7896U2/7897U2 00:0d.0 Ethernet controller: Intel Corporation 82557/8/9/0/1 Ethernet Pro 100 (rev 08)
00:12.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
01:00.0 VGA compatible controller: ATI Technologies Inc Rage 128 RL/VR AGP

Quoted 9 lines

------- Comment #2 From David Hill 2009-06-17 02:55:56 (-) [reply] -------

With NETCONSOLE enabled, if I type:
ethtool -s eth1 speed 100 duplex full autoneg on

the computer freezes with kernel 2.6.29.4 and 2.6.30...

I can reproduce it anytime you want.

Interesting. I wonder what the significance is of the 120 seconds. I see no such timers in e100.c. Does the networking core have timers on such intervals?

#2 lore raw
Quoted 46 lines · Andrew Morton

(switched to email. Please respond via emailed reply-to-all, not via the bugzilla web interface).

Quoted 39 lines · On Wed, 17 Jun 2009 01:55:54 GMT [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=13553

Summary: When NETCONSOLE is enabled in kernel, computer crashes
after 120seconds (approx)
Product: Networking
Version: 2.5
Kernel Version: 2.6.29.4, 2.6.30
Platform: All
OS/Version: Linux
Tree: Mainline
Status: NEW
Severity: high
Priority: P1
Component: Other
AssignedTo: [email protected]
ReportedBy: [email protected]
Regression: No

00:00.0 Host bridge: Intel Corporation 440GX - 82443GX Host bridge 00:01.0 PCI bridge: Intel Corporation 440GX - 82443GX AGP bridge 00:07.0 ISA bridge: Intel Corporation 82371AB/EB/MB PIIX4 ISA (rev 02) 00:07.1 IDE interface: Intel Corporation 82371AB/EB/MB PIIX4 IDE (rev 01) 00:07.2 USB Controller: Intel Corporation 82371AB/EB/MB PIIX4 USB (rev 01) 00:07.3 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 02) 00:0b.0 SCSI storage controller: Adaptec AIC-7896U2/7897U2 00:0b.1 SCSI storage controller: Adaptec AIC-7896U2/7897U2 00:0d.0 Ethernet controller: Intel Corporation 82557/8/9/0/1 Ethernet Pro 100 (rev 08)
00:12.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
01:00.0 VGA compatible controller: ATI Technologies Inc Rage 128 RL/VR AGP

------- Comment #2 From David Hill 2009-06-17 02:55:56 (-) [reply] -------

With NETCONSOLE enabled, if I type:
ethtool -s eth1 speed 100 duplex full autoneg on

Quoted 9 lines
Quoted 4 lines

the computer freezes with kernel 2.6.29.4 and 2.6.30...

I can reproduce it anytime you want.

Interesting. I wonder what the significance is of the 120 seconds. I see no such timers in e100.c. Does the networking core have timers on such intervals?

My guess is the 120 seconds has less to do with the driver, and more to do with some other periodic event in the kernel that triggers a message getting written to the console, which in turn triggers whatever deadlock it is thats getting hit here. I imagine we could diagnose it pretty quick if a stack trace or vmcore could be captured on this. David, can you enable the NMI watchdog on this system to trigger a panic on the system after a deadlock? Then if you could enable a second serial console, or setup kdump to capture a vmcore on this system, we should be able to figure out whats going on. My guess is that in the e100 driver we're taking a lock in the ethtool set path, then calling printk, which winds up recursing into the driver, trying to take the same lock again. A stack trace will tell us for certain.

Regards
Neil

Quoted 5 lines

--
To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#3 lore raw

Re: [Bugme-new] [Bug 13553] New: When NETCONSOLE is enabled inkernel, computer crashes after 120seconds (approx)

Will try that in the next few days... sorry for the delay. I was on vacation for the last 2 weeks and thus, out of town :D

----- Original Message ----- From: "Neil Horman" <[email protected]>
To: "Andrew Morton" <[email protected]>
Cc: <[email protected]>; <[email protected]>; <[email protected]>; <[email protected]>
Sent: Tuesday, June 23, 2009 9:05 PM
Subject: Re: [Bugme-new] [Bug 13553] New: When NETCONSOLE is enabled inkernel, computer crashes after 120seconds (approx)

Quoted 86 lines · Andrew Morton
Quoted 48 lines

(switched to email. Please respond via emailed reply-to-all, not via the bugzilla web interface).

Quoted 41 lines · On Wed, 17 Jun 2009 01:55:54 GMT [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=13553

Summary: When NETCONSOLE is enabled in kernel, computer
crashes

                    after 120seconds (approx)
           Product: Networking
           Version: 2.5
    Kernel Version: 2.6.29.4, 2.6.30
          Platform: All
        OS/Version: Linux
              Tree: Mainline
            Status: NEW
          Severity: high
          Priority: P1
         Component: Other
        AssignedTo: [email protected]
        ReportedBy: [email protected]
        Regression: No

00:00.0 Host bridge: Intel Corporation 440GX - 82443GX Host bridge 00:01.0 PCI bridge: Intel Corporation 440GX - 82443GX AGP bridge 00:07.0 ISA bridge: Intel Corporation 82371AB/EB/MB PIIX4 ISA (rev 02) 00:07.1 IDE interface: Intel Corporation 82371AB/EB/MB PIIX4 IDE (rev 01) 00:07.2 USB Controller: Intel Corporation 82371AB/EB/MB PIIX4 USB (rev 01) 00:07.3 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 02) 00:0b.0 SCSI storage controller: Adaptec AIC-7896U2/7897U2 00:0b.1 SCSI storage controller: Adaptec AIC-7896U2/7897U2 00:0d.0 Ethernet controller: Intel Corporation 82557/8/9/0/1 Ethernet Pro 100 (rev 08)
00:12.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
01:00.0 VGA compatible controller: ATI Technologies Inc Rage 128 RL/VR AGP

------- Comment #2 From David Hill 2009-06-17 02:55:56 (-) [reply] -------

With NETCONSOLE enabled, if I type:
ethtool -s eth1 speed 100 duplex full autoneg on

Quoted 9 lines
Quoted 4 lines

the computer freezes with kernel 2.6.29.4 and 2.6.30...

I can reproduce it anytime you want.

Interesting. I wonder what the significance is of the 120 seconds. I see no such timers in e100.c. Does the networking core have timers on such intervals?

My guess is the 120 seconds has less to do with the driver, and more to do with some other periodic event in the kernel that triggers a message getting written to the console, which in turn triggers whatever deadlock it is thats getting hit here. I imagine we could diagnose it pretty quick if a stack trace or vmcore could be captured on this. David, can you enable the NMI watchdog on this system to trigger a panic on the system after a deadlock? Then if you could enable a second serial console, or setup kdump to capture a vmcore on this system, we should be able to figure out whats going on. My guess is that in the e100 driver we're taking a lock in the ethtool set path, then calling printk, which winds up recursing into the driver, trying to take the same lock again. A stack trace will tell us for certain.

Regards
Neil

Quoted 5 lines

--
To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Signature · 3 lines
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
Signature · 3 lines
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.