> -----Original Message----- > From: linux-scsi-ow...@vger.kernel.org [mailto:linux-scsi- > ow...@vger.kernel.org] On Behalf Of Christoph Hellwig > Sent: Sunday, November 13, 2016 5:29 PM > To: Hannes Reinecke > Cc: Christoph Hellwig; Hannes Reinecke; Christoph Hellwig; Martin K. Petersen; > James Bottomley; Don Brace; linux-scsi@vger.kernel.org; Jens Axboe > Subject: Re: [PATCH] hpsa: scsi-mq support > > On Sun, Nov 13, 2016 at 10:44:47AM +0100, Hannes Reinecke wrote: > > One day to mark with bright red in the calendar. > > > > Christoph Hellwig is telling me _NOT_ to use scsi-mq. > > That's not what I'm doing. > > > This patch was done so see what would needed to be done to convert a > > legacy driver. > > As I was under the impression that scsi-mq is the way forward, seeing > > that it should be enabled per default. > > But I must have been mistaken. Apparently. > > What I am doing is to tell you you should not expose multiple queues unless the > hardware actually has multiple submissions queues. The blk-mq and scsi-mq > code works just fine with a single submission queue, and the hpsa driver in > particular works really well with scsi-mq and a single submission queue. E.g. the > RAID HBA on slide 19 of this presentations is an hpsa one:
I have similar results for MegaRaid where seen MR driver gives significant improvement for Single Submission Queue and multiple Completion Queue. Having said that scsi-mq is enabled but with single Queue is more than enough to maximize improvement of SCSI-MQ. Major advantage was seen while IO load is cross the boundary of Physical CPU socket. >From this discussion I understood that - Similar logical changes proposed for megaraid_sas and we are not really going to gain with fake multiple submission exposed to SML. Kashyap > > http://events.linuxfoundation.org/sites/events/files/slides/scsi.pdf > -- > To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of > a message to majord...@vger.kernel.org More majordomo info at > http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
