iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

RDMA/mlx5: Use odp instead of mr->umem in pagefault_mr

Browse Source 
These are the same thing since mr always comes from odp->private. It is
confusing to reference the same memory via two names.

Link: https://lore.kernel.org/r/20190819111710.18440-13-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This commit is contained in:
Jason Gunthorpe 2019-08-19 14:17:10 +03:00
parent a705f3e3a1
commit fba0e448a2
1 changed files with 5 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
drivers/infiniband/hw/mlx5/odp.c
View File

@ -601,7 +601,7 @@ next_mr:
start_idx = (io_virt - (mr->mmkey.iova & page_mask)) >> page_shift;
access_mask = ODP_READ_ALLOWED_BIT;
if (prefetch && !downgrade && !mr->umem->writable) {
if (prefetch && !downgrade && !odp->umem.writable) {
/* prefetch with write-access must
* be supported by the MR
*/
@ -609,7 +609,7 @@ next_mr:
goto out;
}
if (mr->umem->writable && !downgrade)
if (odp->umem.writable && !downgrade)
access_mask |= ODP_WRITE_ALLOWED_BIT;
current_seq = READ_ONCE(odp->notifiers_seq);
@ -619,8 +619,8 @@ next_mr:
*/
smp_rmb();
ret = ib_umem_odp_map_dma_pages(to_ib_umem_odp(mr->umem), io_virt, size,
access_mask, current_seq);
ret = ib_umem_odp_map_dma_pages(odp, io_virt, size, access_mask,
current_seq);
if (ret < 0)
goto out;
@ -628,8 +628,7 @@ next_mr:
np = ret;
mutex_lock(&odp->umem_mutex);
if (!ib_umem_mmu_notifier_retry(to_ib_umem_odp(mr->umem),
current_seq)) {
if (!ib_umem_mmu_notifier_retry(odp, current_seq)) {
/*
* No need to check whether the MTTs really belong to
* this MR, since ib_umem_odp_map_dma_pages already