linux-wireless.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jouni Malinen <jouni@codeaurora.org>
To: Kalle Valo <kvalo@codeaurora.org>
Cc: ath11k@lists.infradead.org, linux-wireless@vger.kernel.org,
	P Praneesh <ppranees@codeaurora.org>,
	Karthikeyan Periyasamy <periyasa@codeaurora.org>,
	Jouni Malinen <jouni@codeaurora.org>
Subject: [PATCH 12/12] ath11k: avoid unnecessary lock contention in tx_completion path
Date: Wed, 16 Jun 2021 00:14:07 +0300	[thread overview]
Message-ID: <20210615211407.92233-13-jouni@codeaurora.org> (raw)
In-Reply-To: <20210615211407.92233-1-jouni@codeaurora.org>

From: P Praneesh <ppranees@codeaurora.org>

Lock is not needed for the readable idr operation, so avoid spin_lock_bh
for the idr_find() call. No need to disable the bottom half preempt if
it is already in bottom half context, so modify the spin_lock_bh to
spin_lock in the data tx completion handler.

Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2

Co-developed-by: Karthikeyan Periyasamy <periyasa@codeaurora.org>
Signed-off-by: Karthikeyan Periyasamy <periyasa@codeaurora.org>
Signed-off-by: P Praneesh <ppranees@codeaurora.org>
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
---
 drivers/net/wireless/ath/ath11k/dp_tx.c | 18 ++++++++----------
 1 file changed, 8 insertions(+), 10 deletions(-)

diff --git a/drivers/net/wireless/ath/ath11k/dp_tx.c b/drivers/net/wireless/ath/ath11k/dp_tx.c
index ab9ccf0eb274..659f9d98bc0c 100644
--- a/drivers/net/wireless/ath/ath11k/dp_tx.c
+++ b/drivers/net/wireless/ath/ath11k/dp_tx.c
@@ -288,19 +288,18 @@ static void ath11k_dp_tx_free_txbuf(struct ath11k_base *ab, u8 mac_id,
 	struct sk_buff *msdu;
 	struct ath11k_skb_cb *skb_cb;
 
-	spin_lock_bh(&tx_ring->tx_idr_lock);
 	msdu = idr_find(&tx_ring->txbuf_idr, msdu_id);
-	if (!msdu) {
+	if (unlikely(!msdu)) {
 		ath11k_warn(ab, "tx completion for unknown msdu_id %d\n",
 			    msdu_id);
-		spin_unlock_bh(&tx_ring->tx_idr_lock);
 		return;
 	}
 
 	skb_cb = ATH11K_SKB_CB(msdu);
 
+	spin_lock(&tx_ring->tx_idr_lock);
 	idr_remove(&tx_ring->txbuf_idr, msdu_id);
-	spin_unlock_bh(&tx_ring->tx_idr_lock);
+	spin_unlock(&tx_ring->tx_idr_lock);
 
 	dma_unmap_single(ab->dev, skb_cb->paddr, msdu->len, DMA_TO_DEVICE);
 	dev_kfree_skb_any(msdu);
@@ -320,12 +319,10 @@ ath11k_dp_tx_htt_tx_complete_buf(struct ath11k_base *ab,
 	struct ath11k_skb_cb *skb_cb;
 	struct ath11k *ar;
 
-	spin_lock_bh(&tx_ring->tx_idr_lock);
 	msdu = idr_find(&tx_ring->txbuf_idr, ts->msdu_id);
 	if (unlikely(!msdu)) {
 		ath11k_warn(ab, "htt tx completion for unknown msdu_id %d\n",
 			    ts->msdu_id);
-		spin_unlock_bh(&tx_ring->tx_idr_lock);
 		return;
 	}
 
@@ -334,8 +331,9 @@ ath11k_dp_tx_htt_tx_complete_buf(struct ath11k_base *ab,
 
 	ar = skb_cb->ar;
 
+	spin_lock(&tx_ring->tx_idr_lock);
 	idr_remove(&tx_ring->txbuf_idr, ts->msdu_id);
-	spin_unlock_bh(&tx_ring->tx_idr_lock);
+	spin_unlock(&tx_ring->tx_idr_lock);
 
 	if (atomic_dec_and_test(&ar->dp.num_tx_pending))
 		wake_up(&ar->dp.tx_empty_waitq);
@@ -584,16 +582,16 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id)
 			continue;
 		}
 
-		spin_lock_bh(&tx_ring->tx_idr_lock);
 		msdu = idr_find(&tx_ring->txbuf_idr, msdu_id);
 		if (unlikely(!msdu)) {
 			ath11k_warn(ab, "tx completion for unknown msdu_id %d\n",
 				    msdu_id);
-			spin_unlock_bh(&tx_ring->tx_idr_lock);
 			continue;
 		}
+
+		spin_lock(&tx_ring->tx_idr_lock);
 		idr_remove(&tx_ring->txbuf_idr, msdu_id);
-		spin_unlock_bh(&tx_ring->tx_idr_lock);
+		spin_unlock(&tx_ring->tx_idr_lock);
 
 		ar = ab->pdevs[mac_id].ar;
 
-- 
2.25.1


  parent reply	other threads:[~2021-06-15 21:15 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-15 21:13 [PATCH 00/12] ath11k: optimizations in data path Jouni Malinen
2021-06-15 21:13 ` [PATCH 01/12] ath11k: disable unused CE8 interrupts for ipq8074 Jouni Malinen
2021-06-15 21:13 ` [PATCH 02/12] ath11k: allocate dst ring descriptors from cacheable memory Jouni Malinen
2021-06-28 15:19   ` Jeff Johnson
2021-06-15 21:13 ` [PATCH 03/12] ath11k: modify dp_rx desc access wrapper calls inline Jouni Malinen
2021-06-15 21:13 ` [PATCH 04/12] ath11k: avoid additional access to ath11k_hal_srng_dst_num_free Jouni Malinen
2021-06-15 21:14 ` [PATCH 05/12] ath11k: avoid active pdev check for each msdu Jouni Malinen
2021-06-15 21:14 ` [PATCH 06/12] ath11k: remove usage quota while processing rx packets Jouni Malinen
2021-06-15 21:14 ` [PATCH 07/12] ath11k: add branch predictors in process_rx Jouni Malinen
2021-06-15 21:14 ` [PATCH 08/12] ath11k: allocate HAL_WBM2SW_RELEASE ring from cacheable memory Jouni Malinen
2021-06-15 21:14 ` [PATCH 09/12] ath11k: remove mod operator in dst ring processing Jouni Malinen
2021-06-15 21:14 ` [PATCH 10/12] ath11k: avoid while loop in ring selection of tx completion interrupt Jouni Malinen
2021-06-15 21:14 ` [PATCH 11/12] ath11k: add branch predictors in dp_tx path Jouni Malinen
2021-06-15 21:14 ` Jouni Malinen [this message]
2021-06-29 17:35   ` [PATCH 12/12] ath11k: avoid unnecessary lock contention in tx_completion path Felix Fietkau
2021-06-30  5:17     ` P Praneesh

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=20210615211407.92233-13-jouni@codeaurora.org \
    --to=jouni@codeaurora.org \
    --cc=ath11k@lists.infradead.org \
    --cc=kvalo@codeaurora.org \
    --cc=linux-wireless@vger.kernel.org \
    --cc=periyasa@codeaurora.org \
    --cc=ppranees@codeaurora.org \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).