linux-wireless.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Kalle Valo <kvalo@codeaurora.org>
To: Wen Gong <wgong@codeaurora.org>
Cc: ath10k@lists.infradead.org, linux-wireless@vger.kernel.org
Subject: Re: [PATCH v6 3/3] ath10k: add workqueue for RX path of sdio
Date: Thu, 31 Oct 2019 11:08:18 +0200	[thread overview]
Message-ID: <8736f92sfx.fsf@kamboji.qca.qualcomm.com> (raw)
In-Reply-To: <1569402639-31720-4-git-send-email-wgong@codeaurora.org> (Wen Gong's message of "Wed, 25 Sep 2019 17:10:39 +0800")

Wen Gong <wgong@codeaurora.org> writes:

> For RX, it has two parts, one is to read data from sdio, another
> is to indicate the packets to upper stack. Recently it has only
> one thread to do all RX things, it results that it is sequential
> for RX and low throughout, change RX to parallel for the two parts
> will increase throughout.
>
> This patch move the indication to a workqueue, it results in
> significant performance improvement on RX path.
>
> Udp rx throughout is 200Mbps without this patch, and it arrives
> 400Mbps with this patch.
>
> Tested with QCA6174 SDIO with firmware
> WLAN.RMH.4.4.1-00017-QCARMSWPZ-1
>
> Signed-off-by: Wen Gong <wgong@codeaurora.org>

[...]

> +static void ath10k_rx_indication_async_work(struct work_struct *work)
> +{
> +	struct ath10k_sdio *ar_sdio = container_of(work, struct ath10k_sdio,
> +						   async_work_rx);
> +	struct ath10k *ar = ar_sdio->ar;
> +	struct ath10k_htc_ep *ep;
> +	struct ath10k_skb_cb *cb;
> +	struct sk_buff *skb;
> +
> +	while (true) {
> +		skb = skb_dequeue(&ar_sdio->rx_head);
> +		if (!skb)
> +			break;
> +		cb = ATH10K_SKB_CB(skb);
> +		ep = &ar->htc.endpoint[cb->eid];
> +		ep->ep_ops.ep_rx_complete(ar, skb);
> +	}
> +}

I just realised that this is RX path so we should use ATH10K_SKB_RXCB()
instead. I made the change below to this commit in pending branch:

https://git.kernel.org/pub/scm/linux/kernel/git/kvalo/ath.git/commit/?h=pending&id=28da1fe7a3ffa5c55c52328c21165d9efdf2e94c

diff --git a/drivers/net/wireless/ath/ath10k/core.h b/drivers/net/wireless/ath/ath10k/core.h
index af68eb5d0776..c5407f5080b2 100644
--- a/drivers/net/wireless/ath/ath10k/core.h
+++ b/drivers/net/wireless/ath/ath10k/core.h
@@ -124,6 +124,7 @@ struct ath10k_skb_cb {
 struct ath10k_skb_rxcb {
 	dma_addr_t paddr;
 	struct hlist_node hlist;
+	u8 eid;
 };
 
 static inline struct ath10k_skb_cb *ATH10K_SKB_CB(struct sk_buff *skb)
diff --git a/drivers/net/wireless/ath/ath10k/sdio.c b/drivers/net/wireless/ath/ath10k/sdio.c
index c34637881219..c7d09b07a382 100644
--- a/drivers/net/wireless/ath/ath10k/sdio.c
+++ b/drivers/net/wireless/ath/ath10k/sdio.c
@@ -419,7 +419,7 @@ static int ath10k_sdio_mbox_rx_process_packets(struct ath10k *ar,
 	struct ath10k_htc *htc = &ar->htc;
 	struct ath10k_sdio_rx_data *pkt;
 	struct ath10k_htc_ep *ep;
-	struct ath10k_skb_cb *cb;
+	struct ath10k_skb_rxcb *cb;
 	enum ath10k_htc_ep_id id;
 	int ret, i, *n_lookahead_local;
 	u32 *lookaheads_local;
@@ -466,7 +466,7 @@ static int ath10k_sdio_mbox_rx_process_packets(struct ath10k *ar,
 			goto out;
 
 		if (!pkt->trailer_only) {
-			cb = ATH10K_SKB_CB(pkt->skb);
+			cb = ATH10K_SKB_RXCB(pkt->skb);
 			cb->eid = id;
 
 			skb_queue_tail(&ar_sdio->rx_head, pkt->skb);
@@ -1333,14 +1333,14 @@ static void ath10k_rx_indication_async_work(struct work_struct *work)
 						   async_work_rx);
 	struct ath10k *ar = ar_sdio->ar;
 	struct ath10k_htc_ep *ep;
-	struct ath10k_skb_cb *cb;
+	struct ath10k_skb_rxcb *cb;
 	struct sk_buff *skb;
 
 	while (true) {
 		skb = skb_dequeue(&ar_sdio->rx_head);
 		if (!skb)
 			break;
-		cb = ATH10K_SKB_CB(skb);
+		cb = ATH10K_SKB_RXCB(skb);
 		ep = &ar->htc.endpoint[cb->eid];
 		ep->ep_ops.ep_rx_complete(ar, skb);
 	}

-- 
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

  parent reply	other threads:[~2019-10-31  9:08 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-09-25  9:10 [PATCH v6 0/3] ath10k: improve throughout of RX of sdio Wen Gong
2019-09-25  9:10 ` [PATCH v6 1/3] ath10k: enable RX bundle receive for sdio Wen Gong
2019-10-24  8:29   ` Kalle Valo
2019-11-25 11:49   ` Kalle Valo
2019-09-25  9:10 ` [PATCH v6 2/3] ath10k: change max RX bundle size from 8 to 32 " Wen Gong
2019-10-24  9:25   ` Kalle Valo
2019-10-24  9:40     ` Wen Gong
2019-10-24 10:14       ` Kalle Valo
2019-10-24  9:30   ` Kalle Valo
2019-10-24  9:47     ` Wen Gong
2019-10-24 10:05   ` Kalle Valo
2019-10-24 10:48     ` Wen Gong
2019-10-31  9:09       ` Kalle Valo
2019-11-01  7:41         ` Wen Gong
2019-09-25  9:10 ` [PATCH v6 3/3] ath10k: add workqueue for RX path of sdio Wen Gong
2019-10-24  8:14   ` Kalle Valo
2019-10-31  9:08   ` Kalle Valo [this message]
2019-11-01  7:42     ` Wen Gong
2019-11-11 10:47       ` Wen Gong
2019-11-11 12:23         ` Kalle Valo
2019-11-13  3:37           ` Wen Gong
2019-11-22 10:02             ` Kalle Valo

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=8736f92sfx.fsf@kamboji.qca.qualcomm.com \
    --to=kvalo@codeaurora.org \
    --cc=ath10k@lists.infradead.org \
    --cc=linux-wireless@vger.kernel.org \
    --cc=wgong@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).