diff mbox series

[v20,07/20] nvme-tcp: RX DDGST offload

Message ID 20231122134833.20825-8-aaptel@nvidia.com (mailing list archive)
State Deferred
Headers show
Series nvme-tcp receive offloads | expand

Checks

Context Check Description
netdev/series_format fail Series longer than 15 patches (and no cover letter)
netdev/codegen success Generated files up to date
netdev/tree_selection success Guessed tree name to be net-next, async
netdev/fixes_present success Fixes tag not required for -next series
netdev/header_inline success No static functions without inline keyword in header files
netdev/build_32bit success Errors and warnings before: 1127 this patch: 1127
netdev/cc_maintainers success CCed 5 of 5 maintainers
netdev/build_clang success Errors and warnings before: 1154 this patch: 1154
netdev/verify_signedoff success Signed-off-by tag matches author and committer
netdev/deprecated_api success None detected
netdev/check_selftest success No net selftest shell script
netdev/verify_fixes success No Fixes tag
netdev/build_allmodconfig_warn success Errors and warnings before: 1154 this patch: 1154
netdev/checkpatch warning WARNING: line length of 85 exceeds 80 columns
netdev/build_clang_rust success No Rust files in patch. Skipping build
netdev/kdoc success Errors and warnings before: 0 this patch: 0
netdev/source_inline fail Was 0 now: 4

Commit Message

Aurelien Aptel Nov. 22, 2023, 1:48 p.m. UTC 
From: Yoray Zack <yorayz@nvidia.com>

Enable rx side of DDGST offload when supported.

At the end of the capsule, check if all the skb bits are on, and if not
recalculate the DDGST in SW and check it.

Signed-off-by: Yoray Zack <yorayz@nvidia.com>
Signed-off-by: Boris Pismenny <borisp@nvidia.com>
Signed-off-by: Ben Ben-Ishay <benishay@nvidia.com>
Signed-off-by: Or Gerlitz <ogerlitz@nvidia.com>
Signed-off-by: Shai Malin <smalin@nvidia.com>
Signed-off-by: Aurelien Aptel <aaptel@nvidia.com>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
---
 drivers/nvme/host/tcp.c | 84 ++++++++++++++++++++++++++++++++++++++---
 1 file changed, 79 insertions(+), 5 deletions(-)

Comments

Sagi Grimberg Nov. 28, 2023, 10:42 a.m. UTC | #1 
On 11/22/23 15:48, Aurelien Aptel wrote:
> From: Yoray Zack <yorayz@nvidia.com>
> 
> Enable rx side of DDGST offload when supported.
> 
> At the end of the capsule, check if all the skb bits are on, and if not
> recalculate the DDGST in SW and check it.
> 
> Signed-off-by: Yoray Zack <yorayz@nvidia.com>
> Signed-off-by: Boris Pismenny <borisp@nvidia.com>
> Signed-off-by: Ben Ben-Ishay <benishay@nvidia.com>
> Signed-off-by: Or Gerlitz <ogerlitz@nvidia.com>
> Signed-off-by: Shai Malin <smalin@nvidia.com>
> Signed-off-by: Aurelien Aptel <aaptel@nvidia.com>
> Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
> ---
>   drivers/nvme/host/tcp.c | 84 ++++++++++++++++++++++++++++++++++++++---
>   1 file changed, 79 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
> index 680d909eb3fb..5537f04a62fd 100644
> --- a/drivers/nvme/host/tcp.c
> +++ b/drivers/nvme/host/tcp.c
> @@ -141,6 +141,7 @@ enum nvme_tcp_queue_flags {
>   	NVME_TCP_Q_LIVE		= 1,
>   	NVME_TCP_Q_POLLING	= 2,
>   	NVME_TCP_Q_OFF_DDP	= 3,
> +	NVME_TCP_Q_OFF_DDGST_RX = 4,
>   };
>   
>   enum nvme_tcp_recv_state {
> @@ -178,6 +179,7 @@ struct nvme_tcp_queue {
>   	 *   is pending (ULP_DDP_RESYNC_PENDING).
>   	 */
>   	atomic64_t		resync_tcp_seq;
> +	bool			ddp_ddgst_valid;
>   #endif
>   
>   	/* send state */
> @@ -360,6 +362,33 @@ nvme_tcp_get_ddp_netdev_with_limits(struct nvme_tcp_ctrl *ctrl)
>   	return netdev;
>   }
>   
> +static inline bool nvme_tcp_ddp_ddgst_ok(struct nvme_tcp_queue *queue)
> +{
> +	return queue->ddp_ddgst_valid;
> +}
> +
> +static inline void nvme_tcp_ddp_ddgst_update(struct nvme_tcp_queue *queue,
> +					     struct sk_buff *skb)
> +{
> +	if (queue->ddp_ddgst_valid)
> +		queue->ddp_ddgst_valid = skb_is_ulp_crc(skb);
> +}
> +
> +static void nvme_tcp_ddp_ddgst_recalc(struct ahash_request *hash,
> +				      struct request *rq,
> +				      __le32 *ddgst)
> +{
> +	struct nvme_tcp_request *req;
> +
> +	if (!rq)
> +		return;

How is this even possible? And what happens down the road if this is
indeed a null rq?

> +
> +	req = blk_mq_rq_to_pdu(rq);
> +	ahash_request_set_crypt(hash, req->ddp.sg_table.sgl, (u8 *)ddgst,
> +				req->data_len);
> +	crypto_ahash_digest(hash);
> +}
> +
>   static bool nvme_tcp_resync_request(struct sock *sk, u32 seq, u32 flags);
>   static void nvme_tcp_ddp_teardown_done(void *ddp_ctx);
>   static const struct ulp_ddp_ulp_ops nvme_tcp_ddp_ulp_ops = {
> @@ -430,6 +459,8 @@ static void nvme_tcp_setup_ddp(struct nvme_tcp_queue *queue,
>   static int nvme_tcp_offload_socket(struct nvme_tcp_queue *queue)
>   {
>   	struct ulp_ddp_config config = {.type = ULP_DDP_NVME};
> +	bool offload_ddgst_rx = ulp_ddp_is_cap_active(queue->ctrl->ddp_netdev,
> +						      ULP_DDP_CAP_NVME_TCP_DDGST_RX);

Not sure a local variable is needed here.
Aurelien Aptel Nov. 29, 2023, 1:56 p.m. UTC | #2 
Sagi Grimberg <sagi@grimberg.me> writes:
>> +static void nvme_tcp_ddp_ddgst_recalc(struct ahash_request *hash,
>> +                                   struct request *rq,
>> +                                   __le32 *ddgst)
>> +{
>> +     struct nvme_tcp_request *req;
>> +
>> +     if (!rq)
>> +             return;
>
> How is this even possible? And what happens down the road if this is
> indeed a null rq?

You are correct, this isn't possible. We can see the req is fetched
earlier in nvme_tcp_recv_pdu() so it must exist.
We will remove it.

>> +
>> +     req = blk_mq_rq_to_pdu(rq);
>> +     ahash_request_set_crypt(hash, req->ddp.sg_table.sgl, (u8 *)ddgst,
>> +                             req->data_len);
>> +     crypto_ahash_digest(hash);
>> +}
>> +
>>   static bool nvme_tcp_resync_request(struct sock *sk, u32 seq, u32 flags);
>>   static void nvme_tcp_ddp_teardown_done(void *ddp_ctx);
>>   static const struct ulp_ddp_ulp_ops nvme_tcp_ddp_ulp_ops = {
>> @@ -430,6 +459,8 @@ static void nvme_tcp_setup_ddp(struct nvme_tcp_queue *queue,
>>   static int nvme_tcp_offload_socket(struct nvme_tcp_queue *queue)
>>   {
>>       struct ulp_ddp_config config = {.type = ULP_DDP_NVME};
>> +     bool offload_ddgst_rx = ulp_ddp_is_cap_active(queue->ctrl->ddp_netdev,
>> +                                                   ULP_DDP_CAP_NVME_TCP_DDGST_RX);
>
> Not sure a local variable is needed here.

Ok we will remove it.

Thanks
diff mbox series

Patch

diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index 680d909eb3fb..5537f04a62fd 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -141,6 +141,7 @@  enum nvme_tcp_queue_flags {
 	NVME_TCP_Q_LIVE		= 1,
 	NVME_TCP_Q_POLLING	= 2,
 	NVME_TCP_Q_OFF_DDP	= 3,
+	NVME_TCP_Q_OFF_DDGST_RX = 4,
 };
 
 enum nvme_tcp_recv_state {
@@ -178,6 +179,7 @@  struct nvme_tcp_queue {
 	 *   is pending (ULP_DDP_RESYNC_PENDING).
 	 */
 	atomic64_t		resync_tcp_seq;
+	bool			ddp_ddgst_valid;
 #endif
 
 	/* send state */
@@ -360,6 +362,33 @@  nvme_tcp_get_ddp_netdev_with_limits(struct nvme_tcp_ctrl *ctrl)
 	return netdev;
 }
 
+static inline bool nvme_tcp_ddp_ddgst_ok(struct nvme_tcp_queue *queue)
+{
+	return queue->ddp_ddgst_valid;
+}
+
+static inline void nvme_tcp_ddp_ddgst_update(struct nvme_tcp_queue *queue,
+					     struct sk_buff *skb)
+{
+	if (queue->ddp_ddgst_valid)
+		queue->ddp_ddgst_valid = skb_is_ulp_crc(skb);
+}
+
+static void nvme_tcp_ddp_ddgst_recalc(struct ahash_request *hash,
+				      struct request *rq,
+				      __le32 *ddgst)
+{
+	struct nvme_tcp_request *req;
+
+	if (!rq)
+		return;
+
+	req = blk_mq_rq_to_pdu(rq);
+	ahash_request_set_crypt(hash, req->ddp.sg_table.sgl, (u8 *)ddgst,
+				req->data_len);
+	crypto_ahash_digest(hash);
+}
+
 static bool nvme_tcp_resync_request(struct sock *sk, u32 seq, u32 flags);
 static void nvme_tcp_ddp_teardown_done(void *ddp_ctx);
 static const struct ulp_ddp_ulp_ops nvme_tcp_ddp_ulp_ops = {
@@ -430,6 +459,8 @@  static void nvme_tcp_setup_ddp(struct nvme_tcp_queue *queue,
 static int nvme_tcp_offload_socket(struct nvme_tcp_queue *queue)
 {
 	struct ulp_ddp_config config = {.type = ULP_DDP_NVME};
+	bool offload_ddgst_rx = ulp_ddp_is_cap_active(queue->ctrl->ddp_netdev,
+						      ULP_DDP_CAP_NVME_TCP_DDGST_RX);
 	int ret;
 
 	config.nvmeotcp.pfv = NVME_TCP_PFV_1_0;
@@ -449,6 +480,8 @@  static int nvme_tcp_offload_socket(struct nvme_tcp_queue *queue)
 		return ret;
 
 	set_bit(NVME_TCP_Q_OFF_DDP, &queue->flags);
+	if (queue->data_digest && offload_ddgst_rx)
+		set_bit(NVME_TCP_Q_OFF_DDGST_RX, &queue->flags);
 
 	return 0;
 }
@@ -456,6 +489,7 @@  static int nvme_tcp_offload_socket(struct nvme_tcp_queue *queue)
 static void nvme_tcp_unoffload_socket(struct nvme_tcp_queue *queue)
 {
 	clear_bit(NVME_TCP_Q_OFF_DDP, &queue->flags);
+	clear_bit(NVME_TCP_Q_OFF_DDGST_RX, &queue->flags);
 	ulp_ddp_sk_del(queue->ctrl->ddp_netdev, queue->sock->sk);
 }
 
@@ -555,6 +589,20 @@  static void nvme_tcp_resync_response(struct nvme_tcp_queue *queue,
 				     struct sk_buff *skb, unsigned int offset)
 {}
 
+static inline bool nvme_tcp_ddp_ddgst_ok(struct nvme_tcp_queue *queue)
+{
+	return false;
+}
+
+static inline void nvme_tcp_ddp_ddgst_update(struct nvme_tcp_queue *queue,
+					     struct sk_buff *skb)
+{}
+
+static void nvme_tcp_ddp_ddgst_recalc(struct ahash_request *hash,
+				      struct request *rq,
+				      __le32 *ddgst)
+{}
+
 #endif
 
 static void nvme_tcp_init_iter(struct nvme_tcp_request *req,
@@ -815,6 +863,9 @@  static void nvme_tcp_init_recv_ctx(struct nvme_tcp_queue *queue)
 	queue->pdu_offset = 0;
 	queue->data_remaining = -1;
 	queue->ddgst_remaining = 0;
+#ifdef CONFIG_ULP_DDP
+	queue->ddp_ddgst_valid = true;
+#endif
 }
 
 static void nvme_tcp_error_recovery(struct nvme_ctrl *ctrl)
@@ -1082,6 +1133,10 @@  static int nvme_tcp_recv_data(struct nvme_tcp_queue *queue, struct sk_buff *skb,
 		nvme_cid_to_rq(nvme_tcp_tagset(queue), pdu->command_id);
 	struct nvme_tcp_request *req = blk_mq_rq_to_pdu(rq);
 
+	if (queue->data_digest &&
+	    test_bit(NVME_TCP_Q_OFF_DDGST_RX, &queue->flags))
+		nvme_tcp_ddp_ddgst_update(queue, skb);
+
 	while (true) {
 		int recv_len, ret;
 
@@ -1110,7 +1165,8 @@  static int nvme_tcp_recv_data(struct nvme_tcp_queue *queue, struct sk_buff *skb,
 		recv_len = min_t(size_t, recv_len,
 				iov_iter_count(&req->iter));
 
-		if (queue->data_digest)
+		if (queue->data_digest &&
+		    !test_bit(NVME_TCP_Q_OFF_DDGST_RX, &queue->flags))
 			ret = skb_copy_and_hash_datagram_iter(skb, *offset,
 				&req->iter, recv_len, queue->rcv_hash);
 		else
@@ -1152,8 +1208,11 @@  static int nvme_tcp_recv_ddgst(struct nvme_tcp_queue *queue,
 	char *ddgst = (char *)&queue->recv_ddgst;
 	size_t recv_len = min_t(size_t, *len, queue->ddgst_remaining);
 	off_t off = NVME_TCP_DIGEST_LENGTH - queue->ddgst_remaining;
+	struct request *rq;
 	int ret;
 
+	if (test_bit(NVME_TCP_Q_OFF_DDGST_RX, &queue->flags))
+		nvme_tcp_ddp_ddgst_update(queue, skb);
 	ret = skb_copy_bits(skb, *offset, &ddgst[off], recv_len);
 	if (unlikely(ret))
 		return ret;
@@ -1164,9 +1223,25 @@  static int nvme_tcp_recv_ddgst(struct nvme_tcp_queue *queue,
 	if (queue->ddgst_remaining)
 		return 0;
 
+	rq = nvme_cid_to_rq(nvme_tcp_tagset(queue),
+			    pdu->command_id);
+
+	if (test_bit(NVME_TCP_Q_OFF_DDGST_RX, &queue->flags)) {
+		/*
+		 * If HW successfully offloaded the digest
+		 * verification, we can skip it
+		 */
+		if (nvme_tcp_ddp_ddgst_ok(queue))
+			goto out;
+		/*
+		 * Otherwise we have to recalculate and verify the
+		 * digest with the software-fallback
+		 */
+		nvme_tcp_ddp_ddgst_recalc(queue->rcv_hash, rq,
+					  &queue->exp_ddgst);
+	}
+
 	if (queue->recv_ddgst != queue->exp_ddgst) {
-		struct request *rq = nvme_cid_to_rq(nvme_tcp_tagset(queue),
-					pdu->command_id);
 		struct nvme_tcp_request *req = blk_mq_rq_to_pdu(rq);
 
 		req->status = cpu_to_le16(NVME_SC_DATA_XFER_ERROR);
@@ -1177,9 +1252,8 @@  static int nvme_tcp_recv_ddgst(struct nvme_tcp_queue *queue,
 			le32_to_cpu(queue->exp_ddgst));
 	}
 
+out:
 	if (pdu->hdr.flags & NVME_TCP_F_DATA_SUCCESS) {
-		struct request *rq = nvme_cid_to_rq(nvme_tcp_tagset(queue),
-					pdu->command_id);
 		struct nvme_tcp_request *req = blk_mq_rq_to_pdu(rq);
 
 		nvme_tcp_end_request(rq, le16_to_cpu(req->status));