> -----Original Message----- > From: Christoph Hellwig [mailto:hch@xxxxxxxxxxxxx] > Sent: Thursday, October 23, 2014 3:48 AM > To: Elliott, Robert (Server Storage) > Cc: Bart Van Assche; Jens Axboe; Sagi Grimberg; Sebastian Parschauer; > Ming Lei; linux-scsi@xxxxxxxxxxxxxxx; linux-rdma; Scales, Webb; Don > Brace (PMC) > Subject: Re: [PATCH v2 10/12] IB/srp: Use block layer tags > > On Wed, Oct 22, 2014 at 10:03:24PM +0000, Elliott, Robert (Server > Storage) wrote: > > Have you tested this with scsi_mod.use_blk_mq=n? > > > > Trying similar changes in hpsa, we still receive some INQUIRY commands > > submitted through queuecommand with tag -1. They are for devices for > > which slave_alloc has not yet been run, implying this work needs to > > be done even earlier. Maybe the midlayer is missing a slave_alloc > > call somewhere? > > Did that version of hpsa really enable tagging in slave_alloc > or just in slave_configure? The latter would cause INQUIRY to be > sent untagged. Yes, it is slave_alloc, not slave_configure. However, it was looking at scmd->tag, which is always 0xff (at least in those early discovery commands). scmd->request->tag looks like it is the field that has the correct values. Also, I noticed that scmd->tag is just an 8 bit field, so it could never represent a large number of tags. Just to confirm: After calling scsi_init_shared_tag_map() in non-mq mode, will scmd->request->tag be based on controller-wide tag allocation (never using the same value at the same time for the request queues of multiple devices in that controller)? -- To unsubscribe from this list: 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
