On Thu, Jun 08, 2017 at 09:49:35AM +0200, Johannes Thumshirn wrote: > On 06/08/2017 09:44 AM, Christoph Hellwig wrote: > >> + if (sg_zeroout_area(req->sg, req->sg_cnt, NVME_IDENTIFY_DATA_SIZE, off) > > > > Shouldn;t the third argument be NVME_IDENTIFY_DATA_SIZE - off in theory? > > It's probably fine as is as the S/G helpers deal with overflows > > gracefully, but still.. > > Ahm yes. *doh* > > Can you fix it up when applying?
Sure.
