> -----Original Message----- > From: megaraidlinux....@broadcom.com > [mailto:megaraidlinux....@broadcom.com] On Behalf Of Ming Lei > Sent: Tuesday, June 18, 2019 6:05 AM > To: Christoph Hellwig <h...@lst.de> > Cc: Martin K . Petersen <martin.peter...@oracle.com>; Sagi Grimberg > <s...@grimberg.me>; Max Gurtovoy <m...@mellanox.com>; Bart Van > Assche <bvanass...@acm.org>; linux-rdma <linux-r...@vger.kernel.org>; > Linux SCSI List <linux-s...@vger.kernel.org>; > megaraidlinux....@broadcom.com; mpt-fusionlinux....@broadcom.com; > linux-hyp...@vger.kernel.org; Linux Kernel Mailing List <linux- > ker...@vger.kernel.org> > Subject: Re: [PATCH 1/8] scsi: add a host / host template field for the > virt > boundary > > On Mon, Jun 17, 2019 at 8:21 PM Christoph Hellwig <h...@lst.de> wrote: > > > > This allows drivers setting it up easily instead of branching out to > > block layer calls in slave_alloc, and ensures the upgraded > > max_segment_size setting gets picked up by the DMA layer. > > > > Signed-off-by: Christoph Hellwig <h...@lst.de> > > --- > > drivers/scsi/hosts.c | 3 +++ > > drivers/scsi/scsi_lib.c | 3 ++- > > include/scsi/scsi_host.h | 3 +++ > > 3 files changed, 8 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c index > > ff0d8c6a8d0c..55522b7162d3 100644 > > --- a/drivers/scsi/hosts.c > > +++ b/drivers/scsi/hosts.c > > @@ -462,6 +462,9 @@ struct Scsi_Host *scsi_host_alloc(struct > scsi_host_template *sht, int privsize) > > else > > shost->dma_boundary = 0xffffffff; > > > > + if (sht->virt_boundary_mask) > > + shost->virt_boundary_mask = sht->virt_boundary_mask; > > + > > device_initialize(&shost->shost_gendev); > > dev_set_name(&shost->shost_gendev, "host%d", shost->host_no); > > shost->shost_gendev.bus = &scsi_bus_type; diff --git > > a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index > > 65d0a10c76ad..d333bb6b1c59 100644 > > --- a/drivers/scsi/scsi_lib.c > > +++ b/drivers/scsi/scsi_lib.c > > @@ -1775,7 +1775,8 @@ void __scsi_init_queue(struct Scsi_Host *shost, > struct request_queue *q) > > dma_set_seg_boundary(dev, shost->dma_boundary); > > > > blk_queue_max_segment_size(q, shost->max_segment_size); > > - dma_set_max_seg_size(dev, shost->max_segment_size); > > + blk_queue_virt_boundary(q, shost->virt_boundary_mask); > > + dma_set_max_seg_size(dev, queue_max_segment_size(q)); > > The patch looks fine, also suggest to make sure that max_segment_size is > block-size aligned, and un-aligned max segment size has caused trouble on > mmc.
I validated changes on latest and few older series controllers. Post changes, I noticed max_segment_size is updated. find /sys/ -name max_segment_size |grep sdb |xargs grep -r . /sys/devices/pci0000:3a/0000:3a:00.0/0000:3b:00.0/0000:3c:04.0/0000:40:00.0/0000:41:00.0/0000:42:00.0/host0/target0:2:12/0:2:12:0/block/sdb/queue/max_segment_size:4294967295 I verify that single SGE having 1MB transfer length is working fine. Acked-by: Kashyap Desai < kashyap.de...@broadcom.com> > > Thanks, > Ming Lei >
