Re: [PATCH] ath5k: Invoke irqsafe version of ieee80211_tx_status() to avoid deadlock

[Thomas Pedersen <thomas@cozybit.com>]

On Fri, Aug 26, 2011 at 7:22 AM, John W. Linville
<linville@tuxdriver.com> wrote:
> On Wed, Aug 24, 2011 at 06:34:24PM -0700, Thomas Pedersen wrote:
>> From: Javier Cardona <javier@cozybit.com>
>>
>> This driver reports transmission status to the upper layer
>> (ath5k_tx_frame_completed()) while holding the lock on the transmission
>> queue (txq->lock).  Under failure conditions, the mesh stack will
>> attempt to send PERR messages to the previous sender of the failed
>> frame.  When that happens the driver will attempt to re-acquire the
>> txq->lock lock causing a deadlock.  There are two possible fixes for
>> this, (1) we could defer the transmission of the PERR frame until the
>> lock is released or (2) release the lock before invoking
>> ieee80211_tx_status().  The ath9k driver implements the second approach
>> (see ath_tx_complete() in ath9k/xmit.c) as well as the rt2x00 and b43
>> drivers.  The iwl driver, on the other hand, avoids this problem by
>> invoking  ieee80211_tx_status_irqsafe() which effectively defers
>> processing of transmission feedback status.  This last approach is the
>> least intrusive is implemented here.
>>
>> Reported by Pedro Larbig (ASPj)
>> ---
>>  drivers/net/wireless/ath/ath5k/base.c |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>
> Missing Signed-off-by...
>
Yikes. Also, it looks like ieee80211_tx_status() should not be called
from irq context. Will resubmit a v2 with signoff and comment shortly.

Thomas