On Fri, 2016-11-04 at 09:33 -0700, [email protected] wrote: > From: Michael Hernandez <[email protected]> > > Tell the SCSI layer how many hardware queues we have based on the > number of max queue pairs created. The number of max queue pairs > created will depend on number of MSI X vector count or number of CPU's > in a system. > > This feature can be turned on via CONFIG_SCSI_MQ_DEFAULT or passing > scsi_mod.use_blk_mq=Y as a parameter to the kernel > Queue pair creation depend on module parameter "ql2xmqsupport", which > need to be enabled to create queue pair. >
I don't understand this change at all. Setting ->nr_hw_queues causes the block layer to allocate a number of queues for that Scsi_Host object, but it does not appear as if this code uses that functionality. There is nothing in the patch that examines the tag to see which queue the request came in on, in order to map it to a hardware queue. Instead, this patch seems to be reworking the mechanism involved in NPIV vport creation, which creates an entirely separate Scsi_Host object. The driver was already creating separate request queues to the card for vports, so what does this have to do with Block-MQ? -Ewan -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
