Bug 13045 - WATCHDOG i6300ESB doesn't reboot the machine
Summary: WATCHDOG i6300ESB doesn't reboot the machine
Status: ASSIGNED
Alias: None
Product: Other
Classification: Unclassified
Component: Modules (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: Wim Van Sebroeck
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2009-04-08 13:35 UTC by Ralf Hildebrandt
Modified: 2012-10-02 18:19 UTC (History)
0 users

See Also:
Kernel Version: 3.6.0
Subsystem:
Regression: No
Bisected commit-id:


Attachments
lspci -v output (7.79 KB, text/plain)
2010-03-08 14:13 UTC, Ralf Hildebrandt
Details
dmesg output (29.75 KB, text/plain)
2010-03-08 14:14 UTC, Ralf Hildebrandt
Details

Description Ralf Hildebrandt 2009-04-08 13:35:54 UTC
recently installed the watchdog package along with the i6300ESB
module. I killed the watchdog process and waited for the reboot. To no
avail. Is my test flawed? I even stopped the watchdog process,
unloaded and loaded the i6300esb module: No reboot.

This is an HP-Server.

dmesg output:

[  140.017922] i6300ESB timer: initialized (0xf86cc000). heartbeat=30 sec (nowayout=0)

lspci -vv output:

00:1c.0 PCI bridge: Intel Corporation 6300ESB 64-bit PCI-X Bridge (rev 02) (prog-if 00 [Normal decode])
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr+ Stepping- SERR+ FastB2B- DisINTx-
        Status: Cap+ 66MHz+ UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 64, Cache Line Size: 64 bytes
        Bus: primary=00, secondary=02, subordinate=02, sec-latency=64
        I/O behind bridge: 00004000-00004fff
        Memory behind bridge: fdf00000-fdffffff
        Prefetchable memory behind bridge: 0000000088000000-00000000880fffff
        Secondary status: 66MHz+ FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- <SERR- <PERR-
        BridgeCtl: Parity+ SERR+ NoISA- VGA- MAbort- >Reset- FastB2B-
                PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn-
        Capabilities: [50] PCI-X bridge device
                Secondary Status: 64bit+ 133MHz- SCD- USC- SCO- SRD- Freq=66MHz
                Status: Dev=00:1c.0 64bit+ 133MHz- SCD- USC- SCO- SRD-
                Upstream: Capacity=65535 CommitmentLimit=0
                Downstream: Capacity=65535 CommitmentLimit=0

# uname -a
Linux mail-ausfall.charite.de 2.6.29.1 #1 SMP Mon Apr 6 15:25:51 CEST 2009 i686 GNU/Linux

# cat /etc/issue
Debian GNU/Linux squeeze/sid \n \l

# lsmod |grep esb
i6300esb                4852  0

# ll /dev/watchdog
crw-rw---- 1 root root 10, 130  6. Apr 15:31 /dev/watchdog
Comment 1 Wim Van Sebroeck 2009-04-09 05:42:26 UTC
Will Investigate.

Wim.
Comment 2 Ralf Hildebrandt 2009-04-09 09:54:01 UTC
If you need any more info, just let me know.
Comment 3 Ralf Hildebrandt 2009-04-12 09:46:18 UTC
Still there in 2.6.30-rc1-git5
Comment 4 Wim Van Sebroeck 2009-04-12 17:22:43 UTC
Hmmm, I can see that the timer is counting down and that it indeed does not reboot after the timeout/heartbeat period.

Will start debugging...

Kind regards,
Wim.
Comment 5 Ralf Hildebrandt 2009-04-12 18:40:47 UTC
OK. I have a testing machine as a spare, so I can test whatever you produce :)
Comment 6 Ralf Hildebrandt 2009-04-26 09:33:45 UTC
Still there in 2.6.30-rc3-git1:

# uname -a
Linux mail.charite.de 2.6.30-rc3-git1 #1 SMP Sun Apr 26 10:49:49 CEST 2009 i686 GNU/Linux
Comment 7 Ralf Hildebrandt 2009-04-29 13:07:58 UTC
Is there any debug mode for the module so I can poke around?

[   11.064957] i6300ESB timer: Intel 6300ESB WatchDog Timer Driver v0.04

# uname -a
Linux mail.charite.de 2.6.30-rc3-git6 #1 SMP Wed Apr 29 14:56:23 CEST 2009 i686 GNU/Linux
Comment 8 Wim Van Sebroeck 2009-04-29 18:39:29 UTC
No debug code is foreseen. I need to finish another bug first and then I will continue on this one.

I'll keep you posted.

Kind regards,
Wim.
Comment 9 Ralf Hildebrandt 2009-06-09 15:00:39 UTC
I noticed some changes to the esb6300 module, so I gave a new kernel a spin.

Linux proxy-cvk-1 2.6.30-rc8 #1 SMP Wed Jun 3 11:10:19 CEST 2009 i686 GNU/Linux
and i6300esb still doesn't reboot the server
Comment 10 Ralf Hildebrandt 2009-10-08 09:45:23 UTC
# uname -a
Linux mail-ausfall.charite.de 2.6.31.2 #1 SMP Wed Oct 7 16:18:31 CEST 2009 i686 GNU/Linux

# dmesg |fgrep -i esb
[   13.234208] i6300ESB timer: Intel 6300ESB WatchDog Timer Driver v0.04
[   13.234397] i6300ESB timer: initialized (0xf835c000). heartbeat=30 sec (nowayout=0)

Still not working.
Comment 11 Ralf Hildebrandt 2009-12-15 10:03:18 UTC
# uname -a
Linux proxy-cvk-1 2.6.32.1 #1 SMP Tue Dec 15 10:49:28 CET 2009 i686 GNU/Linux

# dmesg |fgrep -i esb
[   16.107182] i6300ESB timer: Intel 6300ESB WatchDog Timer Driver v0.04
[   16.107351] i6300ESB timer: initialized (0xf8566000). heartbeat=30 sec (nowayout=0)

Still not working.
Comment 12 Ralf Hildebrandt 2010-02-04 14:45:51 UTC
Still no go in 2.6.33-rc6
Comment 13 Wim Van Sebroeck 2010-03-08 10:43:19 UTC
Please sent full lspci and dmesg as attachment.
Can you also try with iTCO_wdt instead of i6300esb?
Comment 14 Ralf Hildebrandt 2010-03-08 14:13:33 UTC
Created attachment 25407 [details]
lspci -v output
Comment 15 Ralf Hildebrandt 2010-03-08 14:14:00 UTC
Created attachment 25408 [details]
dmesg output
Comment 16 Ralf Hildebrandt 2010-03-08 14:44:38 UTC
Trying to modprobe iTCO_wdt results in:

[  893.708670] i6300ESB timer: Watchdog Module Unloaded.
[  895.364696] iTCO_vendor_support: vendor-support=0
[  895.366733] iTCO_wdt: Intel TCO WatchDog Timer Driver v1.05
[  895.366877] iTCO_wdt: failed to reset NO_REBOOT flag, reboot disabled by hardware
[  895.366961] iTCO_wdt: No card detected
Comment 17 Ralf Hildebrandt 2010-04-07 13:24:02 UTC
Tried the new version 0.05 today:

mail-ausfall:~# uname -a
Linux mail-ausfall.charite.de 2.6.34-rc3 #1 SMP Wed Apr 7 14:20:31 CEST 2010 i686 GNU/Linux
mail-ausfall:~# dmesg |fgrep -i esb
[    9.002647] i6300ESB timer: Intel 6300ESB WatchDog Timer Driver v0.05
[    9.002809] i6300ESB timer: initialized (0xf855a000). heartbeat=30 sec (nowayout=0)

Doesn't work
Comment 18 Ralf Hildebrandt 2010-06-25 13:12:10 UTC
mail-ausfall:~# uname -a
Linux mail-ausfall.charite.de 2.6.34 #1 SMP Thu Jun 24 14:24:22 CEST 2010 i686 GNU/Linux
mail-ausfall:~# dmesg |fgrep -i esb
[    7.312181] i6300ESB timer: Intel 6300ESB WatchDog Timer Driver v0.05
[    7.312344] i6300ESB timer: initialized (0xf8534000). heartbeat=30 sec (nowayout=0)
Comment 19 Ralf Hildebrandt 2012-02-22 22:10:19 UTC
Still not working (same machine, new name):

# uname -a
Linux mail2.charite.de 3.2.6 #1 SMP Tue Feb 14 13:57:34 CET 2012 i686 GNU/Linux

mail2:~# dmesg |fgrep -i esb
[    3.774844] i6300ESB timer: Intel 6300ESB WatchDog Timer Driver v0.05
[    3.774916] i6300ESB timer: initialized (0xf8416000). heartbeat=30 sec (nowayout=0)
Comment 20 Ralf Hildebrandt 2012-10-02 18:19:11 UTC
# uname -a
Linux mail2.charite.de 3.6.0 #1 SMP Mon Oct 1 17:16:02 CEST 2012 i686 GNU/Linux

mail2:~# dmesg |fgrep -i esb
[    3.719745] i6300esb: Intel 6300ESB WatchDog Timer Driver v0.05
[    3.719848] i6300esb: initialized (0xf8404000). heartbeat=30 sec (nowayout=0)
[62773.643306] Modules linked in: nfnetlink_log nfnetlink ipv6 tg3 microcode rng_core psmouse hpwdt container thermal button processor i6300esb

Note You need to log in before you can comment on or make changes to this bug.