On 10/18/2017 08:34 AM, Christoph Hellwig wrote: > On Wed, Oct 18, 2017 at 08:15:46AM +0200, Hannes Reinecke wrote: >>> It's decidedly awful to have vendor/model-specific triggers in >>> scsi_error. >>> >>> What are the drawbacks of just always refiring on AC/0x44/ITF? >>> >> Hmm. 'Internal target failure' is not very descriptive, so it could mean >> anything. Hence the rather awkward approach. >> But I just checked with the qemu code, and that returns 0x44/0x00 only >> if some (internal) call returned with -ENOMEM. >> So I guess we're safe to always retry here. > > And even if not we should add a quirk so that we can just check a > bit instead of doing two string compares in the I/O completion path.. > Yeah, you are right. Will be adding a blacklist entry for this. Cheers, Hannes -- Dr. Hannes Reinecke Teamlead Storage & Networking hare@xxxxxxx +49 911 74053 688 SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg GF: F. Imendörffer, J. Smithard, J. Guild, D. Upmanyu, G. Norton HRB 21284 (AG Nürnberg)
