[linux-scsi re-added to cc list]
On Mon, 2005-09-12 at 12:58 +0200, Bernd Schubert wrote:
> Hello James,
>
> > > scsi: host 4 channel 0 id 0 lun 0x306727c0706727c0 has a LUN larger than
> > > currently supported.
> >
> > This looks symptomatic of a report luns failure
>
> Hmm, so far we always believed this is a bug in the firmware of our scsi-ide
> raid array.
It is ... the array firmware isn't sending the correct response to
report luns, I think. Try the attached patch; it will blacklist the
array and prevent us from sending REPORT LUNS commands.
> >
> > > EIP is at ahd_send_async+0xde/0x2a0 [aic79xx]
> >
> > This I'm not sure about. There are some fixes that may correct this in
> > the current kernel tree head (i.e. beyond 2.6.13), if you could give
> > that a go.
>
> Well, the oopses do not always occur, only sometimes. On Friday the server
> locked up a second time and we switched to 2.6.13. With three reboots no oops
> did happen, though this doesn't tell anything as we also could reboot 2.6.11
> several times without this oops.
> Well, now we a probably experiencing a network bug in 2.6.13 and the chance it
> pretty high that we have to switch back to 2.6.11, no chance for a kernel
> beyond 2.6.13.
OK ... it looks like an internal refcounting error in the aic79xx; that
piece of the driver was overhauled for post 2.6.13; however, that
doesn't guarantee that this particular issue was fixed ...
James
diff --git a/drivers/scsi/scsi_devinfo.c b/drivers/scsi/scsi_devinfo.c
--- a/drivers/scsi/scsi_devinfo.c
+++ b/drivers/scsi/scsi_devinfo.c
@@ -110,6 +110,7 @@ static struct {
{"RELISYS", "Scorpio", NULL, BLIST_NOLUN}, /* responds to all lun */
{"SANKYO", "CP525", "6.64", BLIST_NOLUN}, /* causes failed REQ SENSE, extra reset */
{"TEXEL", "CD-ROM", "1.06", BLIST_NOLUN},
+ {"transtec", "T5008", "0001", BLIST_NOREPORTLUN },
{"YAMAHA", "CDR100", "1.00", BLIST_NOLUN}, /* locks up */
{"YAMAHA", "CDR102", "1.00", BLIST_NOLUN}, /* locks up */
{"YAMAHA", "CRW8424S", "1.0", BLIST_NOLUN}, /* locks up */
-
: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
