patches for DPDK stable branches
 help / color / mirror / Atom feed
From: "Min Hu (Connor)" <humin29@huawei.com>
To: <stable@dpdk.org>
Cc: <ferruh.yigit@intel.com>, <xuemingl@nvidia.com>
Subject: [dpdk-stable] [PATCH 20.11 2/4] net/hns3: fix VF handling LSC event in secondary process
Date: Mon, 17 May 2021 21:50:43 +0800
Message-ID: <1621259445-5182-2-git-send-email-humin29@huawei.com> (raw)
In-Reply-To: <1621259445-5182-1-git-send-email-humin29@huawei.com>

From: Chengwen Feng <fengchengwen@huawei.com>

[ upstream commit dbbbad23e380773b37872df2195c4529fd93ca6f ]

VF will build two queues (csq: command send queue, crq: command receive
queue) with firmware, the crq may contain the following messages:
1) mailbox response message which was the ack of mailbox sync request.
2) PF's link status change message which may send by PF at anytime;

Currently, any threads in the primary and secondary processes could
send mailbox sync request, so it will need to process the crq messages
in there own thread context.

If the crq hold two messages: a) PF's link status change message, b)
mailbox response message when secondary process deals with the crq
messages, it will lead to report lsc event in secondary process
because it uses the policy of processing all pending messages at once.

We use the following scheme to solve it:
1) threads in secondary process could only process specifics messages
   (eg. mailbox response message) in crq, if the message processed, its
   opcode will rewrite with zero, then the intr thread in primary
   process will not process again.
2) threads other than intr thread in the primary process use the same
   processing logic as the threads in secondary process.
3) intr thread in the primary process could process all messages.

Fixes: 76a3836b98c4 ("net/hns3: fix setting default MAC address in bonding of VF")
Fixes: 463e748964f5 ("net/hns3: support mailbox")
Cc: stable@dpdk.org

Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
Signed-off-by: Min Hu (Connor) <humin29@huawei.com>
---
 drivers/net/hns3/hns3_mbx.c | 68 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 68 insertions(+)

diff --git a/drivers/net/hns3/hns3_mbx.c b/drivers/net/hns3/hns3_mbx.c
index 79ac16a..11e5235 100644
--- a/drivers/net/hns3/hns3_mbx.c
+++ b/drivers/net/hns3/hns3_mbx.c
@@ -407,6 +407,44 @@ hns3_handle_promisc_info(struct hns3_hw *hw, uint16_t promisc_en)
 	}
 }
 
+static void
+hns3_handle_mbx_msg_out_intr(struct hns3_hw *hw)
+{
+	struct hns3_cmq_ring *crq = &hw->cmq.crq;
+	struct hns3_mbx_pf_to_vf_cmd *req;
+	struct hns3_cmd_desc *desc;
+	uint32_t tail, next_to_use;
+	uint8_t opcode;
+	uint16_t flag;
+
+	tail = hns3_read_dev(hw, HNS3_CMDQ_RX_TAIL_REG);
+	next_to_use = crq->next_to_use;
+	while (next_to_use != tail) {
+		desc = &crq->desc[next_to_use];
+		req = (struct hns3_mbx_pf_to_vf_cmd *)desc->data;
+		opcode = req->msg[0] & 0xff;
+
+		flag = rte_le_to_cpu_16(crq->desc[next_to_use].flag);
+		if (!hns3_get_bit(flag, HNS3_CMDQ_RX_OUTVLD_B))
+			goto scan_next;
+
+		if (crq->desc[next_to_use].opcode == 0)
+			goto scan_next;
+
+		if (opcode == HNS3_MBX_PF_VF_RESP) {
+			hns3_handle_mbx_response(hw, req);
+			/*
+			 * Clear opcode to inform intr thread don't process
+			 * again.
+			 */
+			crq->desc[crq->next_to_use].opcode = 0;
+		}
+
+scan_next:
+		next_to_use = (next_to_use + 1) % hw->cmq.crq.desc_num;
+	}
+}
+
 void
 hns3_dev_handle_mbx_msg(struct hns3_hw *hw)
 {
@@ -418,6 +456,29 @@ hns3_dev_handle_mbx_msg(struct hns3_hw *hw)
 	uint16_t flag;
 	rte_spinlock_lock(&hw->cmq.crq.lock);
 
+	if (rte_eal_process_type() != RTE_PROC_PRIMARY ||
+	    !rte_thread_is_intr()) {
+		/*
+		 * Currently, any threads in the primary and secondary processes
+		 * could send mailbox sync request, so it will need to process
+		 * the crq message (which is the HNS3_MBX_PF_VF_RESP) in there
+		 * own thread context. It may also process other messages
+		 * because it uses the policy of processing all pending messages
+		 * at once.
+		 * But some messages such as HNS3_MBX_PUSH_LINK_STATUS could
+		 * only process within the intr thread in primary process,
+		 * otherwise it may lead to report lsc event in secondary
+		 * process.
+		 * So the threads other than intr thread in primary process
+		 * could only process HNS3_MBX_PF_VF_RESP message, if the
+		 * message processed, its opcode will rewrite with zero, then
+		 * the intr thread in primary process will not process again.
+		 */
+		hns3_handle_mbx_msg_out_intr(hw);
+		rte_spinlock_unlock(&hw->cmq.crq.lock);
+		return;
+	}
+
 	while (!hns3_cmd_crq_empty(hw)) {
 		if (rte_atomic16_read(&hw->reset.disable_cmd))
 			rte_spinlock_unlock(&hw->cmq.crq.lock);
@@ -439,6 +500,13 @@ hns3_dev_handle_mbx_msg(struct hns3_hw *hw)
 			continue;
 		}
 
+		if (desc->opcode == 0) {
+			/* Message already processed by other thread */
+			crq->desc[crq->next_to_use].flag = 0;
+			hns3_mbx_ring_ptr_move_crq(crq);
+			continue;
+		}
+
 		switch (opcode) {
 		case HNS3_MBX_PF_VF_RESP:
 			hns3_handle_mbx_response(hw, req);
-- 
2.7.4


  reply	other threads:[~2021-05-17 13:50 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-17 13:50 [dpdk-stable] [PATCH 20.11 1/4] net/hns3: fix possible mismatched response of mailbox Min Hu (Connor)
2021-05-17 13:50 ` Min Hu (Connor) [this message]
2021-05-18  4:15   ` [dpdk-stable] [PATCH 20.11 2/4] net/hns3: fix VF handling LSC event in secondary process Xueming(Steven) Li
2021-05-17 13:50 ` [dpdk-stable] [PATCH 20.11 3/4] net/hns3: fix timing in mailbox Min Hu (Connor)
2021-05-18  4:27   ` Xueming(Steven) Li
2021-05-17 13:50 ` [dpdk-stable] [PATCH 20.11 4/4] net/hns3: fix verification of NEON support Min Hu (Connor)
2021-05-18  4:27   ` Xueming(Steven) Li
2021-05-18  4:14 ` [dpdk-stable] [PATCH 20.11 1/4] net/hns3: fix possible mismatched response of mailbox Xueming(Steven) Li

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1621259445-5182-2-git-send-email-humin29@huawei.com \
    --to=humin29@huawei.com \
    --cc=ferruh.yigit@intel.com \
    --cc=stable@dpdk.org \
    --cc=xuemingl@nvidia.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

patches for DPDK stable branches

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://inbox.dpdk.org/stable/0 stable/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 stable stable/ https://inbox.dpdk.org/stable \
		stable@dpdk.org
	public-inbox-index stable

Example config snippet for mirrors.
Newsgroup available over NNTP:
	nntp://inbox.dpdk.org/inbox.dpdk.stable


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git