> -----Original Message----- > From: Lukas Kolbe [mailto:lkolbe@xxxxxxxxxxxxxxxxxxxxxxxx] > Sent: Thursday, November 19, 2009 4:19 PM > To: Desai, Kashyap > Cc: linux-scsi@xxxxxxxxxxxxxxx; sfrey@xxxxxxxxxxxxxxxxxxxxxxxx > Subject: RE: Bug 14579 (was: RE: Bug 14579 - Devices disappear... and Bug > 14577 - Data corruption with Adaptec) > > Am Donnerstag, den 19.11.2009, 16:00 +0530 schrieb Desai, Kashyap: > > > > > -----Original Message----- > > > From: Lukas Kolbe [mailto:lkolbe@xxxxxxxxxxxxxxxxxxxxxxxx] > > > Sent: Thursday, November 19, 2009 3:48 PM > > > To: Desai, Kashyap > > > Cc: linux-scsi@xxxxxxxxxxxxxxx; sfrey@xxxxxxxxxxxxxxxxxxxxxxxx > > > Subject: Bug 14579 (was: RE: Bug 14579 - Devices disappear... and Bug > > > 14577 - Data corruption with Adaptec) > > > > > > Am Donnerstag, den 19.11.2009, 10:43 +0530 schrieb Desai, Kashyap: > > > > OK. So I get it now. You are working to solve major issue which is > > > > related to your RAID 60 + Adaptec controller. If you want to upgrade > > > > LSI driver, we can always go for latest driver. I can provide you > > > > latest driver source tar ball. Back porting to 2.6.30 is fine. > > > > > > That would be great, thank you. It seems the Adaptec-problem will > still > > > take a while to debug, so if we can get further with the LSI issue it > > > would be much appreciated. It might take a few days of testing after > we > > > build 2.6.30 with the new mtpfusion. > > > > > I think, without solving one issue (Adaptec related), I am not in > > position to understand what is an issue w.r.t LSI ? Pls clarify. If > > you can provide me Logs for LSI driver and provide me pointer to look > > inside, I can do something on this. > > It's all in Bug #14579, see > http://bugzilla.kernel.org/show_bug.cgi?id=14579 > > Basically, after a while writing to the tape drive that is attached to > the LSI controller, the bus either resets itself without finding the > tape drives again, or we get an oops and the system crashes. This is a > seperate problem from the one we have with the Adaptec RAID on 2.6.32, > that is the reason why we opened two bugs. > Looking at oops message attached to bug14579, I am still unable to map anything which might be related to LSI driver, in sort I am not getting any pointers related to fusion driver from OOPS message . Better way to provide Oops will be console redirect. > If you can provide us with the current mptfusion, we can build it for > 2.6.30 and test wether the bus reset and/or oops do still happen. If > they don't, you made a few people very happy :) > I have attached latest fusion driver version 3.04.13 for upstream kernel. You can run *compile* script inside fusion folder to compile it. You have to copy *.ko to /lib/modules/`uname -r`/ and create mkinitrd once again to reflect driver update in mkinitrd. Please try > -- > Kind regards, > Lukas >
Attachment:
fusion_3.04.13.tar.gz
Description: fusion_3.04.13.tar.gz
