Linux wireless drivers development
 help / color / mirror / Atom feed
From: Arik Nemtsov <arik@wizery.com>
To: <linux-wireless@vger.kernel.org>
Cc: Luciano Coelho <coelho@ti.com>, Arik Nemtsov <arik@wizery.com>
Subject: [PATCH 3/7] wl12xx: reset link Tx queues when freeing it
Date: Tue, 28 Feb 2012 00:41:30 +0200	[thread overview]
Message-ID: <1330382494-31148-3-git-send-email-arik@wizery.com> (raw)
In-Reply-To: <1330382494-31148-1-git-send-email-arik@wizery.com>

Before, the link was first freed (invalidating it in the map), and later
on vif removal, all valid wlvif-related links were reset. Since these
links were already invalid, we failed to reset them.
The bug was made worse by op_stop, which set the tx_queue_count to 0
arbitrarily. This resulted in a negative tx_queue_count in some scenarios.

Fix this by resetting the Tx-queues of a link when freeing it. Add a
WARN_ON and reset all link Tx-queues in op_stop, to avoid a negative
tx_queue_count.

Signed-off-by: Arik Nemtsov <arik@wizery.com>
---
 drivers/net/wireless/wl12xx/cmd.c  |    7 +++++++
 drivers/net/wireless/wl12xx/main.c |    1 -
 drivers/net/wireless/wl12xx/tx.c   |   13 ++++++++++---
 3 files changed, 17 insertions(+), 4 deletions(-)

diff --git a/drivers/net/wireless/wl12xx/cmd.c b/drivers/net/wireless/wl12xx/cmd.c
index ef994da..ae1f3d7 100644
--- a/drivers/net/wireless/wl12xx/cmd.c
+++ b/drivers/net/wireless/wl12xx/cmd.c
@@ -485,6 +485,13 @@ void wl12xx_free_link(struct wl1271 *wl, struct wl12xx_vif *wlvif, u8 *hlid)
 	__clear_bit(*hlid, wl->links_map);
 	__clear_bit(*hlid, wlvif->links_map);
 	spin_unlock_irqrestore(&wl->wl_lock, flags);
+
+	/*
+	 * At this point op_tx() will not add more packets to the queues. We
+	 * can purge them.
+	 */
+	wl1271_tx_reset_link_queues(wl, *hlid);
+
 	*hlid = WL12XX_INVALID_LINK_ID;
 }
 
diff --git a/drivers/net/wireless/wl12xx/main.c b/drivers/net/wireless/wl12xx/main.c
index adf9bbc..712e385 100644
--- a/drivers/net/wireless/wl12xx/main.c
+++ b/drivers/net/wireless/wl12xx/main.c
@@ -4228,7 +4228,6 @@ void wl1271_free_sta(struct wl1271 *wl, struct wl12xx_vif *wlvif, u8 hlid)
 	clear_bit(hlid, wlvif->ap.sta_hlid_map);
 	memset(wl->links[hlid].addr, 0, ETH_ALEN);
 	wl->links[hlid].ba_bitmap = 0;
-	wl1271_tx_reset_link_queues(wl, hlid);
 	__clear_bit(hlid, &wl->ap_ps_map);
 	__clear_bit(hlid, (unsigned long *)&wl->ap_fw_ps_map);
 	wl12xx_free_link(wl, wlvif, &hlid);
diff --git a/drivers/net/wireless/wl12xx/tx.c b/drivers/net/wireless/wl12xx/tx.c
index 6446e4d..311b5a2 100644
--- a/drivers/net/wireless/wl12xx/tx.c
+++ b/drivers/net/wireless/wl12xx/tx.c
@@ -527,6 +527,7 @@ static struct sk_buff *wl12xx_lnk_skb_dequeue(struct wl1271 *wl,
 	if (skb) {
 		int q = wl1271_tx_get_queue(skb_get_queue_mapping(skb));
 		spin_lock_irqsave(&wl->wl_lock, flags);
+		WARN_ON(wl->tx_queue_count[q] <= 0);
 		wl->tx_queue_count[q]--;
 		spin_unlock_irqrestore(&wl->wl_lock, flags);
 	}
@@ -602,6 +603,7 @@ static struct sk_buff *wl1271_skb_dequeue(struct wl1271 *wl)
 		skb = wl->dummy_packet;
 		q = wl1271_tx_get_queue(skb_get_queue_mapping(skb));
 		spin_lock_irqsave(&wl->wl_lock, flags);
+		WARN_ON(wl->tx_queue_count[q] <= 0);
 		wl->tx_queue_count[q]--;
 		spin_unlock_irqrestore(&wl->wl_lock, flags);
 	}
@@ -959,7 +961,6 @@ void wl12xx_tx_reset_wlvif(struct wl1271 *wl, struct wl12xx_vif *wlvif)
 		else
 			wlvif->sta.ba_rx_bitmap = 0;
 
-		wl1271_tx_reset_link_queues(wl, i);
 		wl->links[i].allocated_pkts = 0;
 		wl->links[i].prev_freed_pkts = 0;
 	}
@@ -973,8 +974,14 @@ void wl12xx_tx_reset(struct wl1271 *wl, bool reset_tx_queues)
 	struct sk_buff *skb;
 	struct ieee80211_tx_info *info;
 
-	for (i = 0; i < NUM_TX_QUEUES; i++)
-		wl->tx_queue_count[i] = 0;
+	/* only reset the queues if something bad happened */
+	if (WARN_ON(wl1271_tx_total_queue_count(wl) != 0)) {
+		for (i = 0; i < WL12XX_MAX_LINKS; i++)
+			wl1271_tx_reset_link_queues(wl, i);
+
+		for (i = 0; i < NUM_TX_QUEUES; i++)
+			wl->tx_queue_count[i] = 0;
+	}
 
 	wl->stopped_queues_map = 0;
 
-- 
1.7.5.4


  parent reply	other threads:[~2012-02-27 22:41 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-02-27 22:41 [PATCH 1/7] wl12xx: set the ELP entry delay to the FW dyn-ps timeout Arik Nemtsov
2012-02-27 22:41 ` [PATCH 2/7] wl12xx: change bits in the link_map under spin lock Arik Nemtsov
2012-02-27 22:41 ` Arik Nemtsov [this message]
2012-02-28  9:51   ` [PATCH 3/7] wl12xx: reset link Tx queues when freeing it Luciano Coelho
2012-02-28 10:09     ` Arik Nemtsov
2012-02-29  8:31     ` Kalle Valo
2012-02-29  8:39       ` Luciano Coelho
2012-02-27 22:41 ` [PATCH 4/7] wl12xx: avoid starving the system hlid Arik Nemtsov
2012-02-27 22:41 ` [PATCH 5/7] wl12xx: flush all Tx queues on tx_flush timeout Arik Nemtsov
2012-02-27 22:41 ` [PATCH 6/7] wl12xx: flush Tx during suspend and 802.11h chan switch Arik Nemtsov
2012-02-27 22:41 ` [PATCH 7/7] wl12xx: increase dynamic PS timeout to 200ms Arik Nemtsov
2012-02-28 12:05 ` [PATCH 1/7] wl12xx: set the ELP entry delay to the FW dyn-ps timeout Luciano Coelho

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=1330382494-31148-3-git-send-email-arik@wizery.com \
    --to=arik@wizery.com \
    --cc=coelho@ti.com \
    --cc=linux-wireless@vger.kernel.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