public-mirrors/linux
1
0
Fork 0
mirror of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git synced 2025-08-05 16:54:27 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

RDMA/mlx5: Remove ncont from mlx5_ib_cont_pages()

Browse source 
This is the same as ib_umem_num_dma_blocks(umem, 1UL << page_shift), have
the callers compute it directly.

Link: https://lore.kernel.org/r/20201026131936.1335664-7-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
This commit is contained in:
Jason Gunthorpe 2020-10-26 15:19:35 +02:00
parent 95741ee3f0
commit 7db0eea916
7 changed files with 46 additions and 52 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

30
drivers/infiniband/hw/mlx5/cq.c
View file

@ -746,8 +746,9 @@ static int create_cq_user(struct mlx5_ib_dev *dev, struct ib_udata *udata,
if (err) if (err)
goto err_umem; goto err_umem;
mlx5_ib_cont_pages(cq->buf.umem, ucmd.buf_addr, 0, &npages, &page_shift, mlx5_ib_cont_pages(cq->buf.umem, ucmd.buf_addr, 0, &npages,
&ncont); &page_shift);
ncont = ib_umem_num_dma_blocks(cq->buf.umem, 1UL << page_shift);
mlx5_ib_dbg(dev, "addr 0x%llx, size %u, npages %d, page_shift %d, ncont %d\n", mlx5_ib_dbg(dev, "addr 0x%llx, size %u, npages %d, page_shift %d, ncont %d\n",
ucmd.buf_addr, entries * ucmd.cqe_size, npages, page_shift, ncont); ucmd.buf_addr, entries * ucmd.cqe_size, npages, page_shift, ncont);
@ -1128,7 +1129,7 @@ int mlx5_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period)
} }
static int resize_user(struct mlx5_ib_dev *dev, struct mlx5_ib_cq *cq, static int resize_user(struct mlx5_ib_dev *dev, struct mlx5_ib_cq *cq,
int entries, struct ib_udata *udata, int *npas, int entries, struct ib_udata *udata,
int *page_shift, int *cqe_size) int *page_shift, int *cqe_size)
{ {
struct mlx5_ib_resize_cq ucmd; struct mlx5_ib_resize_cq ucmd;
@ -1155,7 +1156,7 @@ static int resize_user(struct mlx5_ib_dev *dev, struct mlx5_ib_cq *cq,
return err; return err;
} }
mlx5_ib_cont_pages(umem, ucmd.buf_addr, 0, &npages, page_shift, npas); mlx5_ib_cont_pages(umem, ucmd.buf_addr, 0, &npages, page_shift);
cq->resize_umem = umem; cq->resize_umem = umem;
*cqe_size = ucmd.cqe_size; *cqe_size = ucmd.cqe_size;
@ -1276,22 +1277,23 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
mutex_lock(&cq->resize_mutex); mutex_lock(&cq->resize_mutex);
if (udata) { if (udata) {
err = resize_user(dev, cq, entries, udata, &npas, &page_shift, err = resize_user(dev, cq, entries, udata, &page_shift,
&cqe_size); &cqe_size);
if (err)
goto ex;
npas = ib_umem_num_dma_blocks(cq->resize_umem, 1UL << page_shift);
} else { } else {
struct mlx5_frag_buf *frag_buf;
cqe_size = 64; cqe_size = 64;
err = resize_kernel(dev, cq, entries, cqe_size); err = resize_kernel(dev, cq, entries, cqe_size);
if (!err) { if (err)
struct mlx5_frag_buf *frag_buf = &cq->resize_buf->frag_buf; goto ex;
frag_buf = &cq->resize_buf->frag_buf;
npas = frag_buf->npages; npas = frag_buf->npages;
page_shift = frag_buf->page_shift; page_shift = frag_buf->page_shift;
}
} }
if (err)
goto ex;
inlen = MLX5_ST_SZ_BYTES(modify_cq_in) + inlen = MLX5_ST_SZ_BYTES(modify_cq_in) +
MLX5_FLD_SZ_BYTES(modify_cq_in, pas[0]) * npas; MLX5_FLD_SZ_BYTES(modify_cq_in, pas[0]) * npas;

12
drivers/infiniband/hw/mlx5/devx.c
View file

@ -95,7 +95,6 @@ struct devx_umem {
struct ib_umem *umem; struct ib_umem *umem;
u32 page_offset; u32 page_offset;
int page_shift; int page_shift;
int ncont;
u32 dinlen; u32 dinlen;
u32 dinbox[MLX5_ST_SZ_DW(general_obj_in_cmd_hdr)]; u32 dinbox[MLX5_ST_SZ_DW(general_obj_in_cmd_hdr)];
}; };
@ -2083,7 +2082,7 @@ static int devx_umem_get(struct mlx5_ib_dev *dev, struct ib_ucontext *ucontext,
mlx5_ib_cont_pages(obj->umem, obj->umem->address, mlx5_ib_cont_pages(obj->umem, obj->umem->address,
MLX5_MKEY_PAGE_SHIFT_MASK, &npages, MLX5_MKEY_PAGE_SHIFT_MASK, &npages,
&obj->page_shift, &obj->ncont); &obj->page_shift);
if (!npages) { if (!npages) {
ib_umem_release(obj->umem); ib_umem_release(obj->umem);
@ -2100,8 +2099,10 @@ static int devx_umem_reg_cmd_alloc(struct uverbs_attr_bundle *attrs,
struct devx_umem *obj, struct devx_umem *obj,
struct devx_umem_reg_cmd *cmd) struct devx_umem_reg_cmd *cmd)
{ {
cmd->inlen = MLX5_ST_SZ_BYTES(create_umem_in) + cmd->inlen =
(MLX5_ST_SZ_BYTES(mtt) * obj->ncont); MLX5_ST_SZ_BYTES(create_umem_in) +
(MLX5_ST_SZ_BYTES(mtt) *
ib_umem_num_dma_blocks(obj->umem, 1UL << obj->page_shift));
cmd->in = uverbs_zalloc(attrs, cmd->inlen); cmd->in = uverbs_zalloc(attrs, cmd->inlen);
return PTR_ERR_OR_ZERO(cmd->in); return PTR_ERR_OR_ZERO(cmd->in);
} }
@ -2117,7 +2118,8 @@ static void devx_umem_reg_cmd_build(struct mlx5_ib_dev *dev,
mtt = (__be64 *)MLX5_ADDR_OF(umem, umem, mtt); mtt = (__be64 *)MLX5_ADDR_OF(umem, umem, mtt);
MLX5_SET(create_umem_in, cmd->in, opcode, MLX5_CMD_OP_CREATE_UMEM); MLX5_SET(create_umem_in, cmd->in, opcode, MLX5_CMD_OP_CREATE_UMEM);
MLX5_SET64(umem, umem, num_of_mtt, obj->ncont); MLX5_SET64(umem, umem, num_of_mtt,
ib_umem_num_dma_blocks(obj->umem, 1UL << obj->page_shift));
MLX5_SET(umem, umem, log_page_size, obj->page_shift - MLX5_SET(umem, umem, log_page_size, obj->page_shift -
MLX5_ADAPTER_PAGE_SHIFT); MLX5_ADAPTER_PAGE_SHIFT);
MLX5_SET(umem, umem, page_offset, obj->page_offset); MLX5_SET(umem, umem, page_offset, obj->page_offset);

15
drivers/infiniband/hw/mlx5/mem.c
View file

@ -41,12 +41,9 @@
* @max_page_shift: high limit for page_shift - 0 means no limit * @max_page_shift: high limit for page_shift - 0 means no limit
* @count: number of PAGE_SIZE pages covered by umem * @count: number of PAGE_SIZE pages covered by umem
* @shift: page shift for the compound pages found in the region * @shift: page shift for the compound pages found in the region
* @ncont: number of compund pages
*/ */
void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr,
unsigned long max_page_shift, unsigned long max_page_shift, int *count, int *shift)
int *count, int *shift,
int *ncont)
{ {
unsigned long tmp; unsigned long tmp;
unsigned long m; unsigned long m;
@ -60,8 +57,7 @@ void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr,
struct ib_umem_odp *odp = to_ib_umem_odp(umem); struct ib_umem_odp *odp = to_ib_umem_odp(umem);
*shift = odp->page_shift; *shift = odp->page_shift;
*ncont = ib_umem_odp_num_pages(odp); *count = ib_umem_num_pages(umem);
*count = *ncont << (*shift - PAGE_SHIFT);
return; return;
} }
@ -90,13 +86,10 @@ void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr,
i += len; i += len;
} }
if (i) { if (i)
m = min_t(unsigned long, ilog2(roundup_pow_of_two(i)), m); m = min_t(unsigned long, ilog2(roundup_pow_of_two(i)), m);
*ncont = DIV_ROUND_UP(i, (1 << m)); else
} else {
m = 0; m = 0;
*ncont = 0;
}
*shift = PAGE_SHIFT + m; *shift = PAGE_SHIFT + m;
*count = i; *count = i;
} }

3
drivers/infiniband/hw/mlx5/mlx5_ib.h
View file

@ -1231,8 +1231,7 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
struct ib_port_attr *props); struct ib_port_attr *props);
void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr,
unsigned long max_page_shift, unsigned long max_page_shift,
int *count, int *shift, int *count, int *shift);
int *ncont);
void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem, void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
int page_shift, size_t offset, size_t num_pages, int page_shift, size_t offset, size_t num_pages,
__be64 *pas, int access_flags); __be64 *pas, int access_flags);

6
drivers/infiniband/hw/mlx5/mr.c
View file

@ -959,10 +959,9 @@ static struct mlx5_ib_mr *alloc_mr_from_cache(struct ib_pd *pd,
struct mlx5_ib_mr *mr; struct mlx5_ib_mr *mr;
int npages; int npages;
int page_shift; int page_shift;
int ncont;
mlx5_ib_cont_pages(umem, iova, MLX5_MKEY_PAGE_SHIFT_MASK, &npages, mlx5_ib_cont_pages(umem, iova, MLX5_MKEY_PAGE_SHIFT_MASK, &npages,
&page_shift, &ncont); &page_shift);
ent = mr_cache_ent_from_order(dev, order_base_2(ib_umem_num_dma_blocks( ent = mr_cache_ent_from_order(dev, order_base_2(ib_umem_num_dma_blocks(
umem, 1UL << page_shift))); umem, 1UL << page_shift)));
if (!ent) if (!ent)
@ -1153,7 +1152,6 @@ static struct mlx5_ib_mr *reg_create(struct ib_mr *ibmr, struct ib_pd *pd,
int page_shift; int page_shift;
__be64 *pas; __be64 *pas;
int npages; int npages;
int ncont;
void *mkc; void *mkc;
int inlen; int inlen;
u32 *in; u32 *in;
@ -1165,7 +1163,7 @@ static struct mlx5_ib_mr *reg_create(struct ib_mr *ibmr, struct ib_pd *pd,
return ERR_PTR(-ENOMEM); return ERR_PTR(-ENOMEM);
mlx5_ib_cont_pages(umem, iova, MLX5_MKEY_PAGE_SHIFT_MASK, &npages, mlx5_ib_cont_pages(umem, iova, MLX5_MKEY_PAGE_SHIFT_MASK, &npages,
&page_shift, &ncont); &page_shift);
mr->page_shift = page_shift; mr->page_shift = page_shift;
mr->ibmr.pd = pd; mr->ibmr.pd = pd;

26
drivers/infiniband/hw/mlx5/qp.c
View file

@ -781,7 +781,7 @@ int bfregn_to_uar_index(struct mlx5_ib_dev *dev,
static int mlx5_ib_umem_get(struct mlx5_ib_dev *dev, struct ib_udata *udata, static int mlx5_ib_umem_get(struct mlx5_ib_dev *dev, struct ib_udata *udata,
unsigned long addr, size_t size, unsigned long addr, size_t size,
struct ib_umem **umem, int *npages, int *page_shift, struct ib_umem **umem, int *npages, int *page_shift,
int *ncont, u32 *offset) u32 *offset)
{ {
int err; int err;
@ -791,7 +791,7 @@ static int mlx5_ib_umem_get(struct mlx5_ib_dev *dev, struct ib_udata *udata,
return PTR_ERR(*umem); return PTR_ERR(*umem);
} }
mlx5_ib_cont_pages(*umem, addr, 0, npages, page_shift, ncont); mlx5_ib_cont_pages(*umem, addr, 0, npages, page_shift);
err = mlx5_ib_get_buf_offset(addr, *page_shift, offset); err = mlx5_ib_get_buf_offset(addr, *page_shift, offset);
if (err) { if (err) {
@ -799,8 +799,8 @@ static int mlx5_ib_umem_get(struct mlx5_ib_dev *dev, struct ib_udata *udata,
goto err_umem; goto err_umem;
} }
mlx5_ib_dbg(dev, "addr 0x%lx, size %zu, npages %d, page_shift %d, ncont %d, offset %d\n", mlx5_ib_dbg(dev, "addr 0x%lx, size %zu, npages %d, page_shift %d, offset %d\n",
addr, size, *npages, *page_shift, *ncont, *offset); addr, size, *npages, *page_shift, *offset);
return 0; return 0;
@ -836,7 +836,6 @@ static int create_user_rq(struct mlx5_ib_dev *dev, struct ib_pd *pd,
int page_shift = 0; int page_shift = 0;
int npages; int npages;
u32 offset = 0; u32 offset = 0;
int ncont = 0;
int err; int err;
if (!ucmd->buf_addr) if (!ucmd->buf_addr)
@ -849,8 +848,7 @@ static int create_user_rq(struct mlx5_ib_dev *dev, struct ib_pd *pd,
return err; return err;
} }
mlx5_ib_cont_pages(rwq->umem, ucmd->buf_addr, 0, &npages, &page_shift, mlx5_ib_cont_pages(rwq->umem, ucmd->buf_addr, 0, &npages, &page_shift);
&ncont);
err = mlx5_ib_get_buf_offset(ucmd->buf_addr, page_shift, err = mlx5_ib_get_buf_offset(ucmd->buf_addr, page_shift,
&rwq->rq_page_offset); &rwq->rq_page_offset);
if (err) { if (err) {
@ -858,14 +856,14 @@ static int create_user_rq(struct mlx5_ib_dev *dev, struct ib_pd *pd,
goto err_umem; goto err_umem;
} }
rwq->rq_num_pas = ncont; rwq->rq_num_pas = ib_umem_num_dma_blocks(rwq->umem, 1UL << page_shift);
rwq->page_shift = page_shift; rwq->page_shift = page_shift;
rwq->log_page_size = page_shift - MLX5_ADAPTER_PAGE_SHIFT; rwq->log_page_size = page_shift - MLX5_ADAPTER_PAGE_SHIFT;
rwq->wq_sig = !!(ucmd->flags & MLX5_WQ_FLAG_SIGNATURE); rwq->wq_sig = !!(ucmd->flags & MLX5_WQ_FLAG_SIGNATURE);
mlx5_ib_dbg(dev, "addr 0x%llx, size %zd, npages %d, page_shift %d, ncont %d, offset %d\n", mlx5_ib_dbg(dev, "addr 0x%llx, size %zd, npages %d, page_shift %d, ncont %d, offset %d\n",
(unsigned long long)ucmd->buf_addr, rwq->buf_size, (unsigned long long)ucmd->buf_addr, rwq->buf_size,
npages, page_shift, ncont, offset); npages, page_shift, rwq->rq_num_pas, offset);
err = mlx5_ib_db_map_user(ucontext, udata, ucmd->db_addr, &rwq->db); err = mlx5_ib_db_map_user(ucontext, udata, ucmd->db_addr, &rwq->db);
if (err) { if (err) {
@ -952,9 +950,10 @@ static int _create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd,
ubuffer->buf_addr = ucmd->buf_addr; ubuffer->buf_addr = ucmd->buf_addr;
err = mlx5_ib_umem_get(dev, udata, ubuffer->buf_addr, err = mlx5_ib_umem_get(dev, udata, ubuffer->buf_addr,
ubuffer->buf_size, &ubuffer->umem, ubuffer->buf_size, &ubuffer->umem,
&npages, &page_shift, &ncont, &offset); &npages, &page_shift, &offset);
if (err) if (err)
goto err_bfreg; goto err_bfreg;
ncont = ib_umem_num_dma_blocks(ubuffer->umem, 1UL << page_shift);
} else { } else {
ubuffer->umem = NULL; ubuffer->umem = NULL;
} }
@ -1211,16 +1210,17 @@ static int create_raw_packet_qp_sq(struct mlx5_ib_dev *dev,
int err; int err;
int page_shift = 0; int page_shift = 0;
int npages; int npages;
int ncont = 0;
u32 offset = 0; u32 offset = 0;
err = mlx5_ib_umem_get(dev, udata, ubuffer->buf_addr, ubuffer->buf_size, err = mlx5_ib_umem_get(dev, udata, ubuffer->buf_addr, ubuffer->buf_size,
&sq->ubuffer.umem, &npages, &page_shift, &ncont, &sq->ubuffer.umem, &npages, &page_shift,
&offset); &offset);
if (err) if (err)
return err; return err;
inlen = MLX5_ST_SZ_BYTES(create_sq_in) + sizeof(u64) * ncont; inlen = MLX5_ST_SZ_BYTES(create_sq_in) +
sizeof(u64) * ib_umem_num_dma_blocks(sq->ubuffer.umem,
1UL << page_shift);
in = kvzalloc(inlen, GFP_KERNEL); in = kvzalloc(inlen, GFP_KERNEL);
if (!in) { if (!in) {
err = -ENOMEM; err = -ENOMEM;

6
drivers/infiniband/hw/mlx5/srq.c
View file

@ -53,7 +53,6 @@ static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq,
int err; int err;
int npages; int npages;
int page_shift; int page_shift;
int ncont;
u32 offset; u32 offset;
u32 uidx = MLX5_IB_DEFAULT_UIDX; u32 uidx = MLX5_IB_DEFAULT_UIDX;
@ -88,7 +87,7 @@ static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq,
} }
mlx5_ib_cont_pages(srq->umem, ucmd.buf_addr, 0, &npages, mlx5_ib_cont_pages(srq->umem, ucmd.buf_addr, 0, &npages,
&page_shift, &ncont); &page_shift);
err = mlx5_ib_get_buf_offset(ucmd.buf_addr, page_shift, err = mlx5_ib_get_buf_offset(ucmd.buf_addr, page_shift,
&offset); &offset);
if (err) { if (err) {
@ -96,7 +95,8 @@ static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq,
goto err_umem; goto err_umem;
} }
in->pas = kvcalloc(ncont, sizeof(*in->pas), GFP_KERNEL); in->pas = kvcalloc(ib_umem_num_dma_blocks(srq->umem, 1UL << page_shift),
sizeof(*in->pas), GFP_KERNEL);
if (!in->pas) { if (!in->pas) {
err = -ENOMEM; err = -ENOMEM;
goto err_umem; goto err_umem;