mptbase Code={Abort}

A "me too" here as well.

Fedora 8 x86_64 2.6.24.4-64.fc8, 8GB RAM, LSI SAS1068 B1 with latest BIOS and FW, attached to 15 x WD7500AYYS SATA via a Vitesse port expander.

While doing initial sync up on an md RAID 5 of the first eight drives:

Apr 13 13:53:16 flash2 kernel: mptbase: ioc0: LogInfo(0x31123000): Originator={PL}, Code={Abort}, SubCode(0x3000)
Apr 13 13:53:18 flash2 kernel: mptbase: ioc0: LogInfo(0x31120403): Originator={PL}, Code={Abort}, SubCode(0x0403)
Apr 13 13:53:21 flash2 kernel: mptbase: ioc0: LogInfo(0x31123000): Originator={PL}, Code={Abort}, SubCode(0x3000)
Apr 13 13:53:21 flash2 kernel: mptbase: ioc0: LogInfo(0x31120403): Originator={PL}, Code={Abort}, SubCode(0x0403)
Apr 13 13:53:23 flash2 kernel: mptbase: ioc0: LogInfo(0x31123000): Originator={PL}, Code={Abort}, SubCode(0x3000)
Apr 13 13:53:24 flash2 kernel: mptbase: ioc0: LogInfo(0x31120403): Originator={PL}, Code={Abort}, SubCode(0x0403)
Apr 13 13:53:34 flash2 kernel: mptbase: ioc0: LogInfo(0x31123000): Originator={PL}, Code={Abort}, SubCode(0x3000)
Apr 13 13:53:34 flash2 kernel: mptbase: ioc0: LogInfo(0x31120403): Originator={PL}, Code={Abort}, SubCode(0x0403)
Apr 13 13:53:36 flash2 kernel: mptbase: ioc0: LogInfo(0x31123000): Originator={PL}, Code={Abort}, SubCode(0x3000)
Apr 13 13:53:38 flash2 kernel: mptbase: ioc0: LogInfo(0x31120403): Originator={PL}, Code={Abort}, SubCode(0x0403)
Apr 13 13:53:42 flash2 kernel: mptbase: ioc0: LogInfo(0x31123000): Originator={PL}, Code={Abort}, SubCode(0x3000)
Apr 13 13:53:42 flash2 kernel: mptbase: ioc0: LogInfo(0x31120403): Originator={PL}, Code={Abort}, SubCode(0x0403)
Apr 13 13:53:44 flash2 kernel: mptbase: ioc0: LogInfo(0x31123000): Originator={PL}, Code={Abort}, SubCode(0x3000)
Apr 13 13:53:46 flash2 kernel: mptbase: ioc0: LogInfo(0x31120403): Originator={PL}, Code={Abort}, SubCode(0x0403)
Apr 13 13:53:50 flash2 kernel: mptbase: ioc0: LogInfo(0x31123000): Originator={PL}, Code={Abort}, SubCode(0x3000)

Will try compiling latest mpt fusion driver today.

Have sent a similar mail, including lsiutil dump offlist to eric at LSI.

Regards,

Richard

I have already done it. Nothing has changed.

Could you let me know when he replies, please? I would like know whether it is something serious or can be ignored.

I have noticed that turning off NCQ significantly reduces amount of the errors.

I have set for each disk (sda,sdb... ($dev)):
echo 1 > /sys/block/$dev/device/queue_depth

and the same value for "SATA Maximum Queue Depth" using lsiutils.

Better but the error message still did not disappear.

Regards,
Sebastian

"Computer Science is no more about computers
 than astronomy is about telescopes"
                        Edsger Wybe Dijkstra

Sebastian Wasilewski
Division of Physical Biochemistry
MRC National Institute for Medical Research
The Ridgeway, London NW7 1AA
+442088162089

[email protected]
[email protected]