Olaf Hering wrote:
On Mon, Jun 05, Brian King wrote:
Brian King wrote:
...
Setting sdev->allow_restart in struct scsi_device will cause sense
key/code/qual of 02/04/02 (not ready, initialization command required)
to wake up the scsi error handler and will force scsi core to issue
a start unit command to the disk. I added this a while back to handle
ipr raid arrays which need a start unit command each time the adapter
gets reset.
...
Please try the attached patch. It adds a sysfs device attribute "allow_restart".
...
Is it safe to have this attribute be 1 per default, and let the user
disable it for problematic drives? The IO errors bother me alot...
Matthew Dharm pointed out that it may confuse USB devices:
http://marc.theaimsgroup.com/?l=linux-scsi&m=107702811830956
Side note: May plan for sbp2 is to switch the flag on only for models
(or, as it is currently done in -mm, for vendor IDs) which are known to
need the flag. The flag won't hurt with most SBP-2 HDDs; most of them
don't need it though. But I am worried about how the unknown few quirky
ones may react. On the other hand, it is likely that MS's SBP-2 stack is
behaving like Linux with allow_restart=1. I somehow doubt they have a
whitelist or blacklist for it.
What about enabling it for all direct access and RBC devices, except USB
direct access devices? (Or equivalently, enable it for all devices in
scsi/sd but switch it off again in usb-storage?) Or would that be too
daring?
--
Stefan Richter
-=====-=-==- -=== -=-==
http://arcgraph.de/sr/
-
: 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
