On 5/30/19 6:25 AM, Max Gurtovoy wrote:
This new WR will be used to perform PI (protection information) handover
using the new API. Using the new API, the user will post a single WR that
will internally perform all the needed actions to complete PI operation.
This new WR will use a memory region that was allocated as
IB_MR_TYPE_INTEGRITY and was mapped using ib_map_mr_sg_pi to perform the
registration. In the old API, in order to perform a signature handover
operation, each ULP should perform the following:
1. Map and register the data buffers.
2. Map and register the protection buffers.
3. Post a special reg WR to configure the signature handover operation
layout.
4. Invalidate the signature memory key.
5. Invalidate protection buffers memory key.
6. Invalidate data buffers memory key.
In the new API, the mapping of both data and protection buffers is
performed using a single call to ib_map_mr_sg_pi function. Also the
registration of the buffers and the configuration of the signature
operation layout is done by a single new work request called
IB_WR_REG_MR_INTEGRITY.
This patch implements this operation for mlx5 devices that are capable to
offload data integrity generation/validation while performing the actual
buffer transfer.
This patch will not remove the old signature API that is used by the iSER
initiator and target drivers. This will be done in the future.
In the internal implementation, for each IB_WR_REG_MR_INTEGRITY work
request, we are using a single UMR operation to register both data and
protection buffers using KLM's.
Afterwards, another UMR operation will describe the strided block format.
These will be followed by 2 SET_PSV operations to set the memory/wire
domains initial signature parameters passed by the user.
In the end of the whole transaction, only the signature memory key
(the one that exposed for the RDMA operation) will be invalidated.
Signed-off-by: Max Gurtovoy <maxg@xxxxxxxxxxxx>
Signed-off-by: Israel Rukshin <israelr@xxxxxxxxxxxx>
Reviewed-by: Sagi Grimberg <sagi@xxxxxxxxxxx>
---
drivers/infiniband/hw/mlx5/qp.c | 218 ++++++++++++++++++++++++++++++++++++----
include/linux/mlx5/qp.h | 3 +-
include/rdma/ib_verbs.h | 1 +
3 files changed, 201 insertions(+), 21 deletions(-)
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 65d82e40871c..7c9fd335d43d 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -4172,7 +4172,7 @@ static __be64 sig_mkey_mask(void)
static void set_reg_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr,
struct mlx5_ib_mr *mr, u8 flags)
{
- int size = mr->ndescs * mr->desc_size;
+ int size = (mr->ndescs + mr->meta_ndescs) * mr->desc_size;
memset(umr, 0, sizeof(*umr));
@@ -4303,7 +4303,7 @@ static void set_reg_mkey_seg(struct mlx5_mkey_seg *seg,
struct mlx5_ib_mr *mr,
u32 key, int access)
{
- int ndescs = ALIGN(mr->ndescs, 8) >> 1;
+ int ndescs = ALIGN(mr->ndescs + mr->meta_ndescs, 8) >> 1;
memset(seg, 0, sizeof(*seg));
@@ -4354,7 +4354,7 @@ static void set_reg_data_seg(struct mlx5_wqe_data_seg *dseg,
struct mlx5_ib_mr *mr,
struct mlx5_ib_pd *pd)
{
- int bcount = mr->desc_size * mr->ndescs;
+ int bcount = mr->desc_size * (mr->ndescs + mr->meta_ndescs);
dseg->addr = cpu_to_be64(mr->desc_map);
dseg->byte_count = cpu_to_be32(ALIGN(bcount, 64));
@@ -4547,23 +4547,52 @@ static int mlx5_set_bsf(struct ib_mr *sig_mr,
return 0;
}
-static int set_sig_data_segment(const struct ib_sig_handover_wr *wr,
- struct mlx5_ib_qp *qp, void **seg,
- int *size, void **cur_edge)
+static int set_sig_data_segment(const struct ib_send_wr *send_wr,
+ struct ib_mr *sig_mr,
+ struct ib_sig_attrs *sig_attrs,
+ struct mlx5_ib_qp *qp, void **seg, int *size,
+ void **cur_edge)
{
- struct ib_sig_attrs *sig_attrs = wr->sig_attrs;
- struct ib_mr *sig_mr = wr->sig_mr;
struct mlx5_bsf *bsf;
- u32 data_len = wr->wr.sg_list->length;
- u32 data_key = wr->wr.sg_list->lkey;
- u64 data_va = wr->wr.sg_list->addr;
+ u32 data_len;
+ u32 data_key;
+ u64 data_va;
+ u32 prot_len = 0;
+ u32 prot_key = 0;
+ u64 prot_va = 0;
+ bool prot = false;
int ret;
int wqe_size;
- if (!wr->prot ||
- (data_key == wr->prot->lkey &&
- data_va == wr->prot->addr &&
- data_len == wr->prot->length)) {
+ if (send_wr->opcode == IB_WR_REG_SIG_MR) {
+ const struct ib_sig_handover_wr *wr = sig_handover_wr(send_wr);
+
+ data_len = wr->wr.sg_list->length;
+ data_key = wr->wr.sg_list->lkey;
+ data_va = wr->wr.sg_list->addr;
+ if (wr->prot) {
+ prot_len = wr->prot->length;
+ prot_key = wr->prot->lkey;
+ prot_va = wr->prot->addr;
+ prot = true;
+ }
+ } else {
+ struct mlx5_ib_mr *mr = to_mmr(sig_mr);
+ struct mlx5_ib_mr *pi_mr = mr->pi_mr;
+
+ data_len = pi_mr->data_length;
+ data_key = pi_mr->ibmr.lkey;
+ data_va = pi_mr->ibmr.iova;
+ if (pi_mr->meta_ndescs) {
+ prot_len = pi_mr->meta_length;
+ prot_key = pi_mr->ibmr.lkey;
+ prot_va = pi_mr->ibmr.iova + data_len;
+ prot = true;
+ }
+ }
+
+ if (!prot || (data_key == prot_key && data_va == prot_va &&
+ data_len == prot_len)) {
Worth explaining in a comment that this is either insert/strip or
interleaved case..
Other than that,
Reviewed-by: Sagi Grimberg <sagi@xxxxxxxxxxx>
(again)
