lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

IB/mlx5: Allow creation of QPs with zero-length work queues

Browse Source 
The current code attmepts to call ib_umem_get() even if the length is
zero, which causes a failure. Since the spec allows zero length work
queues, change the code so we don't call ib_umem_get() in those cases.

Signed-off-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
This commit is contained in:
Eli Cohen
2014-01-14 17:45:21 +02:00
committed by Roland Dreier
parent 05bdb2ab6b
commit 9e9c47d07d
1 changed files with 29 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

49
drivers/infiniband/hw/mlx5/qp.c
View File

@@ -523,12 +523,12 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd,
{ {
struct mlx5_ib_ucontext *context; struct mlx5_ib_ucontext *context;
struct mlx5_ib_create_qp ucmd; struct mlx5_ib_create_qp ucmd;
int page_shift; int page_shift = 0;
int uar_index; int uar_index;
int npages; int npages;
u32 offset; u32 offset = 0;
int uuarn; int uuarn;
int ncont; int ncont = 0;
int err; int err;
err = ib_copy_from_udata(&ucmd, udata, sizeof(ucmd)); err = ib_copy_from_udata(&ucmd, udata, sizeof(ucmd));
@@ -564,23 +564,29 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd,
if (err) if (err)
goto err_uuar; goto err_uuar;
qp->umem = ib_umem_get(pd->uobject->context, ucmd.buf_addr, if (ucmd.buf_addr && qp->buf_size) {
qp->buf_size, 0, 0); qp->umem = ib_umem_get(pd->uobject->context, ucmd.buf_addr,
if (IS_ERR(qp->umem)) { qp->buf_size, 0, 0);
mlx5_ib_dbg(dev, "umem_get failed\n"); if (IS_ERR(qp->umem)) {
err = PTR_ERR(qp->umem); mlx5_ib_dbg(dev, "umem_get failed\n");
goto err_uuar; err = PTR_ERR(qp->umem);
goto err_uuar;
}
} else {
qp->umem = NULL;
} }
mlx5_ib_cont_pages(qp->umem, ucmd.buf_addr, &npages, &page_shift, if (qp->umem) {
&ncont, NULL); mlx5_ib_cont_pages(qp->umem, ucmd.buf_addr, &npages, &page_shift,
err = mlx5_ib_get_buf_offset(ucmd.buf_addr, page_shift, &offset); &ncont, NULL);
if (err) { err = mlx5_ib_get_buf_offset(ucmd.buf_addr, page_shift, &offset);
mlx5_ib_warn(dev, "bad offset\n"); if (err) {
goto err_umem; mlx5_ib_warn(dev, "bad offset\n");
goto err_umem;
}
mlx5_ib_dbg(dev, "addr 0x%llx, size %d, npages %d, page_shift %d, ncont %d, offset %d\n",
ucmd.buf_addr, qp->buf_size, npages, page_shift, ncont, offset);
} }
mlx5_ib_dbg(dev, "addr 0x%llx, size %d, npages %d, page_shift %d, ncont %d, offset %d\n",
ucmd.buf_addr, qp->buf_size, npages, page_shift, ncont, offset);
*inlen = sizeof(**in) + sizeof(*(*in)->pas) * ncont; *inlen = sizeof(**in) + sizeof(*(*in)->pas) * ncont;
*in = mlx5_vzalloc(*inlen); *in = mlx5_vzalloc(*inlen);
@@ -588,7 +594,8 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd,
err = -ENOMEM; err = -ENOMEM;
goto err_umem; goto err_umem;
} }
mlx5_ib_populate_pas(dev, qp->umem, page_shift, (*in)->pas, 0); if (qp->umem)
mlx5_ib_populate_pas(dev, qp->umem, page_shift, (*in)->pas, 0);
(*in)->ctx.log_pg_sz_remote_qpn = (*in)->ctx.log_pg_sz_remote_qpn =
cpu_to_be32((page_shift - MLX5_ADAPTER_PAGE_SHIFT) << 24); cpu_to_be32((page_shift - MLX5_ADAPTER_PAGE_SHIFT) << 24);
(*in)->ctx.params2 = cpu_to_be32(offset << 6); (*in)->ctx.params2 = cpu_to_be32(offset << 6);
@@ -619,7 +626,8 @@ err_free:
mlx5_vfree(*in); mlx5_vfree(*in);
err_umem: err_umem:
ib_umem_release(qp->umem); if (qp->umem)
ib_umem_release(qp->umem);
err_uuar: err_uuar:
free_uuar(&context->uuari, uuarn); free_uuar(&context->uuari, uuarn);
@@ -632,7 +640,8 @@ static void destroy_qp_user(struct ib_pd *pd, struct mlx5_ib_qp *qp)
context = to_mucontext(pd->uobject->context); context = to_mucontext(pd->uobject->context);
mlx5_ib_db_unmap_user(context, &qp->db); mlx5_ib_db_unmap_user(context, &qp->db);
ib_umem_release(qp->umem); if (qp->umem)
ib_umem_release(qp->umem);
free_uuar(&context->uuari, qp->uuarn); free_uuar(&context->uuari, qp->uuarn);
} }