[SRU][N:nvidia][PATCH 3/8] net: mana: Fix race of mana_hwc_post_rx_wqe and new hwc response

[Jacob Martin]

From: Haiyang Zhang <haiyangz at microsoft.com>

BugLink: https://bugs.launchpad.net/bugs/2084598

The mana_hwc_rx_event_handler() / mana_hwc_handle_resp() calls
complete(&ctx->comp_event) before posting the wqe back. It's
possible that other callers, like mana_create_txq(), start the
next round of mana_hwc_send_request() before the posting of wqe.
And if the HW is fast enough to respond, it can hit no_wqe error
on the HW channel, then the response message is lost. The mana
driver may fail to create queues and open, because of waiting for
the HW response and timed out.
Sample dmesg:
[  528.610840] mana 39d4:00:02.0: HWC: Request timed out!
[  528.614452] mana 39d4:00:02.0: Failed to send mana message: -110, 0x0
[  528.618326] mana 39d4:00:02.0 enP14804s2: Failed to create WQ object: -110

To fix it, move posting of rx wqe before complete(&ctx->comp_event).

Cc: stable at vger.kernel.org
Fixes: ca9c54d2d6a5 ("net: mana: Add a driver for Microsoft Azure Network Adapter (MANA)")
Signed-off-by: Haiyang Zhang <haiyangz at microsoft.com>
Reviewed-by: Long Li <longli at microsoft.com>
Signed-off-by: David S. Miller <davem at davemloft.net>
(cherry-picked from commit 8af174ea863c72f25ce31cee3baad8a301c0cf0f netdev)
Signed-off-by: Vinicius Peixoto <vinicius.peixoto at canonical.com>
Acked-by: Aaron Jauregui <aaron.jauregui at canonical.com>
Acked-by: Thibault Ferrante <thibault.ferrante at canonical.com>
Signed-off-by: John Cabaj <john.cabaj at canonical.com>
(cherry picked from commit 40a4338c667f2c84e84b1fd5484e46ae3321e34f noble:linux-azure/master-next)
Signed-off-by: Jacob Martin <jacob.martin at canonical.com>
---
 .../net/ethernet/microsoft/mana/hw_channel.c  | 62 ++++++++++---------
 1 file changed, 34 insertions(+), 28 deletions(-)

diff --git a/drivers/net/ethernet/microsoft/mana/hw_channel.c b/drivers/net/ethernet/microsoft/mana/hw_channel.c
index 4dd43ac5a3cb..48b899834d0a 100644
--- a/drivers/net/ethernet/microsoft/mana/hw_channel.c
+++ b/drivers/net/ethernet/microsoft/mana/hw_channel.c
@@ -51,9 +51,33 @@ static int mana_hwc_verify_resp_msg(const struct hwc_caller_ctx *caller_ctx,
 	return 0;
 }
 
+static int mana_hwc_post_rx_wqe(const struct hwc_wq *hwc_rxq,
+				struct hwc_work_request *req)
+{
+	struct device *dev = hwc_rxq->hwc->dev;
+	struct gdma_sge *sge;
+	int err;
+
+	sge = &req->sge;
+	sge->address = (u64)req->buf_sge_addr;
+	sge->mem_key = hwc_rxq->msg_buf->gpa_mkey;
+	sge->size = req->buf_len;
+
+	memset(&req->wqe_req, 0, sizeof(struct gdma_wqe_request));
+	req->wqe_req.sgl = sge;
+	req->wqe_req.num_sge = 1;
+	req->wqe_req.client_data_unit = 0;
+
+	err = mana_gd_post_and_ring(hwc_rxq->gdma_wq, &req->wqe_req, NULL);
+	if (err)
+		dev_err(dev, "Failed to post WQE on HWC RQ: %d\n", err);
+	return err;
+}
+
 static void mana_hwc_handle_resp(struct hw_channel_context *hwc, u32 resp_len,
-				 const struct gdma_resp_hdr *resp_msg)
+				 struct hwc_work_request *rx_req)
 {
+	const struct gdma_resp_hdr *resp_msg = rx_req->buf_va;
 	struct hwc_caller_ctx *ctx;
 	int err;
 
@@ -61,6 +85,7 @@ static void mana_hwc_handle_resp(struct hw_channel_context *hwc, u32 resp_len,
 		      hwc->inflight_msg_res.map)) {
 		dev_err(hwc->dev, "hwc_rx: invalid msg_id = %u\n",
 			resp_msg->response.hwc_msg_id);
+		mana_hwc_post_rx_wqe(hwc->rxq, rx_req);
 		return;
 	}
 
@@ -74,30 +99,13 @@ static void mana_hwc_handle_resp(struct hw_channel_context *hwc, u32 resp_len,
 	memcpy(ctx->output_buf, resp_msg, resp_len);
 out:
 	ctx->error = err;
-	complete(&ctx->comp_event);
-}
-
-static int mana_hwc_post_rx_wqe(const struct hwc_wq *hwc_rxq,
-				struct hwc_work_request *req)
-{
-	struct device *dev = hwc_rxq->hwc->dev;
-	struct gdma_sge *sge;
-	int err;
-
-	sge = &req->sge;
-	sge->address = (u64)req->buf_sge_addr;
-	sge->mem_key = hwc_rxq->msg_buf->gpa_mkey;
-	sge->size = req->buf_len;
 
-	memset(&req->wqe_req, 0, sizeof(struct gdma_wqe_request));
-	req->wqe_req.sgl = sge;
-	req->wqe_req.num_sge = 1;
-	req->wqe_req.client_data_unit = 0;
+	/* Must post rx wqe before complete(), otherwise the next rx may
+	 * hit no_wqe error.
+	 */
+	mana_hwc_post_rx_wqe(hwc->rxq, rx_req);
 
-	err = mana_gd_post_and_ring(hwc_rxq->gdma_wq, &req->wqe_req, NULL);
-	if (err)
-		dev_err(dev, "Failed to post WQE on HWC RQ: %d\n", err);
-	return err;
+	complete(&ctx->comp_event);
 }
 
 static void mana_hwc_init_event_handler(void *ctx, struct gdma_queue *q_self,
@@ -234,14 +242,12 @@ static void mana_hwc_rx_event_handler(void *ctx, u32 gdma_rxq_id,
 		return;
 	}
 
-	mana_hwc_handle_resp(hwc, rx_oob->tx_oob_data_size, resp);
+	mana_hwc_handle_resp(hwc, rx_oob->tx_oob_data_size, rx_req);
 
-	/* Do no longer use 'resp', because the buffer is posted to the HW
-	 * in the below mana_hwc_post_rx_wqe().
+	/* Can no longer use 'resp', because the buffer is posted to the HW
+	 * in mana_hwc_handle_resp() above.
 	 */
 	resp = NULL;
-
-	mana_hwc_post_rx_wqe(hwc_rxq, rx_req);
 }
 
 static void mana_hwc_tx_event_handler(void *ctx, u32 gdma_txq_id,
-- 
2.43.0