diff options
| author | Sergey Gorenko <sergeygo@mellanox.com> | 2018-05-21 18:55:53 +0300 |
|---|---|---|
| committer | Doug Ledford <dledford@redhat.com> | 2018-05-31 20:07:17 +0300 |
| commit | 434dda422cb2d52af4c75857c3f2084fe8f3a2ba (patch) | |
| tree | c19c07db4098bcffea7064859c6187d1fc59294c /drivers/infiniband/ulp/iser/iser_verbs.c | |
| parent | 01298c0cd49f75bd2f133299d8f2655b533363eb (diff) | |
| download | linux-434dda422cb2d52af4c75857c3f2084fe8f3a2ba.tar.xz | |
IB/iser: Do not reduce max_sectors
The iSER driver reduces max_sectors. For example, if you load the
ib_iser module with max_sectors=1024, you will see that
/sys/class/block/<bdev>/queue/max_hw_sectors_kb is 508. It is an
incorrect value. The expected value is (max_sectors * sector_size) /
1024 = 512.
Reducing of max_sectors can cause performance degradation due to
unnecessary splitting of IO requests.
The number of pages per MR has been fixed here, so there is no longer
any need to reduce max_sectors.
Fixes: 9c674815d346 ("IB/iser: Fix max_sectors calculation")
Signed-off-by: Sergey Gorenko <sergeygo@mellanox.com>
Reviewed-by: Israel Rukshin <israelr@mellanox.com>
Reviewed-by: Max Gurtovoy <maxg@mellanox.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Sagi Grimberg <sagi@grimberg.me>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Diffstat (limited to 'drivers/infiniband/ulp/iser/iser_verbs.c')
| -rw-r--r-- | drivers/infiniband/ulp/iser/iser_verbs.c | 21 |
1 files changed, 18 insertions, 3 deletions
diff --git a/drivers/infiniband/ulp/iser/iser_verbs.c b/drivers/infiniband/ulp/iser/iser_verbs.c index 56b7240a3fc3..616d978cbf2b 100644 --- a/drivers/infiniband/ulp/iser/iser_verbs.c +++ b/drivers/infiniband/ulp/iser/iser_verbs.c @@ -703,19 +703,34 @@ iser_calc_scsi_params(struct iser_conn *iser_conn, unsigned int max_sectors) { struct iser_device *device = iser_conn->ib_conn.device; + struct ib_device_attr *attr = &device->ib_device->attrs; unsigned short sg_tablesize, sup_sg_tablesize; + unsigned short reserved_mr_pages; + + /* + * FRs without SG_GAPS or FMRs can only map up to a (device) page per + * entry, but if the first entry is misaligned we'll end up using two + * entries (head and tail) for a single page worth data, so one + * additional entry is required. + */ + if ((attr->device_cap_flags & IB_DEVICE_MEM_MGT_EXTENSIONS) && + (attr->device_cap_flags & IB_DEVICE_SG_GAPS_REG)) + reserved_mr_pages = 0; + else + reserved_mr_pages = 1; sg_tablesize = DIV_ROUND_UP(max_sectors * 512, SIZE_4K); - if (device->ib_device->attrs.device_cap_flags & - IB_DEVICE_MEM_MGT_EXTENSIONS) + if (attr->device_cap_flags & IB_DEVICE_MEM_MGT_EXTENSIONS) sup_sg_tablesize = min_t( uint, ISCSI_ISER_MAX_SG_TABLESIZE, - device->ib_device->attrs.max_fast_reg_page_list_len); + attr->max_fast_reg_page_list_len - reserved_mr_pages); else sup_sg_tablesize = ISCSI_ISER_MAX_SG_TABLESIZE; iser_conn->scsi_sg_tablesize = min(sg_tablesize, sup_sg_tablesize); + iser_conn->pages_per_mr = + iser_conn->scsi_sg_tablesize + reserved_mr_pages; } /** |