From mboxrd@z Thu Jan 1 00:00:00 1970 From: hch@lst.de (Christoph Hellwig) Date: Tue, 16 Jul 2019 14:05:47 +0200 Subject: [PATCH 2/3] nvme: Retrieve the required IO queue entry size from the controller In-Reply-To: References: <20190716004649.17799-1-benh@kernel.crashing.org> <20190716004649.17799-2-benh@kernel.crashing.org> <20190716060430.GB29414@lst.de> <20190716093301.GA32562@lst.de> Message-ID: <20190716120547.GA2388@lst.de> On Tue, Jul 16, 2019@08:58:28PM +1000, Benjamin Herrenschmidt wrote: > The main risk is if existing controllers return crap in SQES and we try > to then use that crap. The rest should essentially be NOPs. > > Maybe I should add some kind of printk to warn in case we use/detect a > non-standard size. That would help diagnosing issues. Given that the spec currently requires bits 0 to 3 of SQES to be 6 we might as well not check SQES and just hardcode it to 6 or 7 depending on the quirk. That actually was my initial idea, I just suggested using the SQES naming and indexing.