| Message ID | 20160407221359.GA16686@infradead.org (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
On 04/07/16 15:14, Christoph Hellwig wrote: > The patch below fixes the issue for me. But for SRP I'm still seeing > lots of these failures even with the equivalent patch applied: > > [ 426.027905] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053203] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053265] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053270] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053273] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053276] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053279] scsi host8: ib_srp: Failed to map data (-12) > > It seems the actual reproducer is xfs_repair as run by xfstests. Hello Christoph, Thanks for the report. I will try to reproduce this behavior and will see whether I can find the root cause. Bart. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 4/8/2016 1:13 AM, Christoph Hellwig wrote: > The patch below fixes the issue for me. But for SRP I'm still seeing > lots of these failures even with the equivalent patch applied: > > [ 426.027905] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053203] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053265] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053270] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053273] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053276] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053279] scsi host8: ib_srp: Failed to map data (-12) > > It seems the actual reproducer is xfs_repair as run by xfstests. > > diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c > index 60b169a..3ac4755 100644 > --- a/drivers/infiniband/ulp/srp/ib_srp.c > +++ b/drivers/infiniband/ulp/srp/ib_srp.c > @@ -3205,6 +3205,7 @@ static ssize_t srp_create_target(struct device *dev, > target_host->max_id = 1; > target_host->max_lun = -1LL; > target_host->max_cmd_len = sizeof ((struct srp_cmd *) (void *) 0L)->cdb; > + target_host->max_segment_size = PAGE_SIZE; > > target = host_to_target(target_host); Hi Christoph, did you add similar assignment in iser code or only used the scsi patches ? -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Sun, Apr 10, 2016 at 04:30:11PM +0300, Max Gurtovoy wrote: > Hi Christoph, > did you add similar assignment in iser code or only used the scsi patches ? Yes, to fix it I had to do the assignment in iser. Not sure why I didn't manage to send it out. Let me prepare a proper patch series. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 04/07/2016 03:13 PM, Christoph Hellwig wrote: > The patch below fixes the issue for me. But for SRP I'm still seeing > lots of these failures even with the equivalent patch applied: > > [ 426.027905] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053203] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053265] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053270] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053273] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053276] scsi host8: ib_srp: Failed to map data (-12) > [ 426.053279] scsi host8: ib_srp: Failed to map data (-12) > > It seems the actual reproducer is xfs_repair as run by xfstests. Hello Christoph, So far I have not yet been able to reproduce the ib_srp mapping failure with kernel v4.6-rc3. How was ib_srp configured in your test (cat /etc/modprobe.d/ib_srp.conf)? Thanks, Bart. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 60b169a..3ac4755 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -3205,6 +3205,7 @@ static ssize_t srp_create_target(struct device *dev, target_host->max_id = 1; target_host->max_lun = -1LL; target_host->max_cmd_len = sizeof ((struct srp_cmd *) (void *) 0L)->cdb; + target_host->max_segment_size = PAGE_SIZE; target = host_to_target(target_host); diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 8c6e318..ae331e0 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -2120,7 +2120,8 @@ static void __scsi_init_queue(struct Scsi_Host *shost, struct request_queue *q) blk_queue_segment_boundary(q, shost->dma_boundary); dma_set_seg_boundary(dev, shost->dma_boundary); - blk_queue_max_segment_size(q, dma_get_max_seg_size(dev)); + blk_queue_max_segment_size(q, + min(shost->max_segment_size, dma_get_max_seg_size(dev))); if (!shost->use_clustering) q->limits.cluster = 0; diff --git a/include/scsi/scsi_host.h b/include/scsi/scsi_host.h index fcfa3d7..f11d3fe 100644 --- a/include/scsi/scsi_host.h +++ b/include/scsi/scsi_host.h @@ -621,6 +621,7 @@ struct Scsi_Host { short unsigned int sg_tablesize; short unsigned int sg_prot_tablesize; unsigned int max_sectors; + unsigned int max_segment_size; unsigned long dma_boundary; /* * In scsi-mq mode, the number of hardware queues supported by the LLD.