2013-03-19 01:21:22

by Luis Chamberlain

Subject: [PATCH] ath9k: skip rearming rx_poll_timer when quiescing

From: "Luis R. Rodriguez" <[email protected]>

When mac80211 goes to suspend we disarm our timers
but users of mod_timer() must ensure to use
mod_timer_pending() where appropriate to avoid
re-arming them before coming up after suspend. This
is one case that was reported by Parag.

Issue reported:

WARNING: at net/mac80211/util.c:599 ieee80211_can_queue_work.isra.7+0x32/0x40 [mac80211]()
Hardware name: iMac12,1
queueing ieee80211 work while going to suspend
Pid: 0, comm: swapper/0 Tainted: PF O 3.8.2-206.fc18.x86_64 #1
Call Trace: Mar 16 09:39:17 Parags-iMac kernel: [ 3993.642992] <IRQ>
[<ffffffff8105e61f>] warn_slowpath_common+0x7f/0xc0
[<ffffffffa0581420>] ? ath_start_rx_poll+0x70/0x70 [ath9k]
<ffffffff8105e716>] warn_slowpath_fmt+0x46/0x50
[<ffffffffa045b542>] ieee80211_can_queue_work.isra.7+0x32/0x40

Other timers on 802.11 drivers, mac80211 and cfg80211
should be groomed to verify the same is avoided.

Reported-by: Parag Warudkar <[email protected]>
Tested-by: Parag Warudkar <[email protected]>
Cc: [email protected]
Signed-off-by: Luis R. Rodriguez <[email protected]>
---
drivers/net/wireless/ath/ath9k/link.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/wireless/ath/ath9k/link.c b/drivers/net/wireless/ath/ath9k/link.c
index 39c84ec..25aa894 100644
--- a/drivers/net/wireless/ath/ath9k/link.c
+++ b/drivers/net/wireless/ath/ath9k/link.c
@@ -162,8 +162,8 @@ void ath_start_rx_poll(struct ath_softc *sc, u8 nbeacon)
if (!test_bit(SC_OP_PRIM_STA_VIF, &sc->sc_flags))
return;

- mod_timer(&sc->rx_poll_timer, jiffies + msecs_to_jiffies
- (nbeacon * sc->cur_beacon_conf.beacon_interval));
+ mod_timer_pending(&sc->rx_poll_timer, jiffies + msecs_to_jiffies
+ (nbeacon * sc->cur_beacon_conf.beacon_interval));
}

void ath_rx_poll(unsigned long data)
--
1.7.10.4

2013-03-21 19:21:25

by Luis Chamberlain

Subject: Re: [PATCH] ath9k: skip rearming rx_poll_timer when quiescing

On Mon, Mar 18, 2013 at 6:14 PM, Luis R. Rodriguez
<[email protected]> wrote:
> From: "Luis R. Rodriguez" <[email protected]>
>
> When mac80211 goes to suspend we disarm our timers
> but users of mod_timer() must ensure to use
> mod_timer_pending() where appropriate to avoid
> re-arming them before coming up after suspend. This
> is one case that was reported by Parag.
>
> Issue reported:
>
> WARNING: at net/mac80211/util.c:599 ieee80211_can_queue_work.isra.7+0x32/0x40 [mac80211]()
> Hardware name: iMac12,1
> queueing ieee80211 work while going to suspend
> Pid: 0, comm: swapper/0 Tainted: PF O 3.8.2-206.fc18.x86_64 #1
> Call Trace: Mar 16 09:39:17 Parags-iMac kernel: [ 3993.642992] <IRQ>
> [<ffffffff8105e61f>] warn_slowpath_common+0x7f/0xc0
> [<ffffffffa0581420>] ? ath_start_rx_poll+0x70/0x70 [ath9k]
> <ffffffff8105e716>] warn_slowpath_fmt+0x46/0x50
> [<ffffffffa045b542>] ieee80211_can_queue_work.isra.7+0x32/0x40
>
> Other timers on 802.11 drivers, mac80211 and cfg80211
> should be groomed to verify the same is avoided.
>
> Reported-by: Parag Warudkar <[email protected]>
> Tested-by: Parag Warudkar <[email protected]>
> Cc: [email protected]
> Signed-off-by: Luis R. Rodriguez <[email protected]>

John, please hold off on this.

Luis