[PATCH 3.9 0/2] two iwlwifi fixes

[Stanislaw Gruszka]

Intel recomends to include those two patches in Fedora.

Emmanuel Grumbach (2):
  iwlwifi: pcie: fix race in queue unmapping
  iwlwifi: pcie: wake the queue if stopped when being unmapped

drivers/net/wireless/iwlwifi/pcie/tx.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

[PATCH 3.9 1/2] iwlwifi: pcie: fix race in queue unmapping

[Stanislaw Gruszka]

From: Emmanuel Grumbach

When a queue is disabled, it frees all its entries. Later,
the op_mode might still get notifications from the firmware
that triggers to free entries in the tx queue. The transport
should be prepared for these races and know to ignore
reclaim calls on queues that have been disabled and whose
entries have been freed.

Cc: [e-mail]
Signed-off-by: Emmanuel Grumbach 
Signed-off-by: Johannes Berg 
---
 drivers/net/wireless/iwlwifi/pcie/tx.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/drivers/net/wireless/iwlwifi/pcie/tx.c b/drivers/net/wireless/iwlwifi/pcie/tx.c
index cb5c679..faaf77c 100644
--- a/drivers/net/wireless/iwlwifi/pcie/tx.c
+++ b/drivers/net/wireless/iwlwifi/pcie/tx.c
@@ -578,9 +578,12 @@ static void iwl_pcie_txq_unmap(struct iwl_trans *trans, int txq_id)

spin_lock_bh(&txq->lock);
    while (q->write_ptr != q->read_ptr) {
+		IWL_DEBUG_TX_REPLY(trans, "Q %d Free %d\n",
+				   txq_id, q->read_ptr);
    	iwl_pcie_txq_free_tfd(trans, txq);
    	q->read_ptr = iwl_queue_inc_wrap(q->read_ptr, q->n_bd);
    }
+	txq->active = false;
    spin_unlock_bh(&txq->lock);
 }

@@ -929,6 +932,12 @@ void iwl_trans_pcie_reclaim(struct iwl_trans *trans, int txq_id, int ssn,

spin_lock_bh(&txq->lock);

+	if (!txq->active) {
+		IWL_DEBUG_TX_QUEUES(trans, "Q %d inactive - ignoring idx %d\n",
+				    txq_id, ssn);
+		goto out;
+	}
+
    if (txq->q.read_ptr == tfd_num)
    	goto out;

@@ -1105,6 +1114,7 @@ void iwl_trans_pcie_txq_enable(struct iwl_trans *trans, int txq_id, int fifo,
    	       (fifo << SCD_QUEUE_STTS_REG_POS_TXF) |
    	       (1 << SCD_QUEUE_STTS_REG_POS_WSL) |
    	       SCD_QUEUE_STTS_REG_MSK);
+	trans_pcie->txq[txq_id].active = true;
    IWL_DEBUG_TX_QUEUES(trans, "Activate queue %d on FIFO %d WrPtr: %d\n",
    		    txq_id, fifo, ssn & 0xff);
 }

[PATCH 3.9 2/2] iwlwifi: pcie: wake the queue if stopped when being unmapped

[Stanislaw Gruszka]

From: Emmanuel Grumbach

When the queue is unmapped while it was so loaded that
mac80211's was stopped, we need to wake the queue after
having freed all the packets in the queue.
Not doing so can result in weird stuff like:

* run lots of traffic (mac80211's queue gets stopped)
* RFKILL
* de-assert RFKILL
* no traffic

Cc: [e-mail]
Signed-off-by: Emmanuel Grumbach 
Signed-off-by: Johannes Berg 
---
 drivers/net/wireless/iwlwifi/pcie/tx.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/wireless/iwlwifi/pcie/tx.c b/drivers/net/wireless/iwlwifi/pcie/tx.c
index faaf77c..4e7b8d4 100644
--- a/drivers/net/wireless/iwlwifi/pcie/tx.c
+++ b/drivers/net/wireless/iwlwifi/pcie/tx.c
@@ -585,6 +585,9 @@ static void iwl_pcie_txq_unmap(struct iwl_trans *trans, int txq_id)
    }
    txq->active = false;
    spin_unlock_bh(&txq->lock);
+
+	/* just in case - this queue may have been stopped */
+	iwl_wake_queue(trans, txq);
 }

/*

[Josh Boyer]

On Fri, Jun 21, 2013 at 03:21:50PM +0200, Stanislaw Gruszka wrote:
> Intel recomends to include those two patches in Fedora.
>
> Emmanuel Grumbach (2):
>   iwlwifi: pcie: fix race in queue unmapping
>   iwlwifi: pcie: wake the queue if stopped when being unmapped
>
> drivers/net/wireless/iwlwifi/pcie/tx.c | 13 +++++++++++++
>  1 file changed, 13 insertions(+)

Ok.  I can grab these.

They don't seem to be in the net_39.mbox that DaveM sent to stable a
couple days ago.  Are they still in someone else's tree at the moment?

josh

[Stanislaw Gruszka]

On Fri, Jun 21, 2013 at 09:29:03AM -0400, Josh Boyer wrote:
> On Fri, Jun 21, 2013 at 03:21:50PM +0200, Stanislaw Gruszka wrote:
> > Intel recomends to include those two patches in Fedora.
> >
> > Emmanuel Grumbach (2):
> >   iwlwifi: pcie: fix race in queue unmapping
> >   iwlwifi: pcie: wake the queue if stopped when being unmapped
> >
> > drivers/net/wireless/iwlwifi/pcie/tx.c | 13 +++++++++++++
> >  1 file changed, 13 insertions(+)
>
> Ok.  I can grab these.
>
> They don't seem to be in the net_39.mbox that DaveM sent to stable a
> couple days ago.  Are they still in someone else's tree at the moment?

They were refused by upstream maintainers as fixes proper to 3.10-rc5 or
later -rc. Instead patches were pushed to -next and queued to 3.11. Anyway
fixes are small, look sane and are intended to -stable.

Stanislaw

[Josh Boyer]

On Fri, Jun 21, 2013 at 03:51:16PM +0200, Stanislaw Gruszka wrote:
> On Fri, Jun 21, 2013 at 09:29:03AM -0400, Josh Boyer wrote:
> > On Fri, Jun 21, 2013 at 03:21:50PM +0200, Stanislaw Gruszka wrote:
> > > Intel recomends to include those two patches in Fedora.
> > >
> > > Emmanuel Grumbach (2):
> > >   iwlwifi: pcie: fix race in queue unmapping
> > >   iwlwifi: pcie: wake the queue if stopped when being unmapped
> > >
> > > drivers/net/wireless/iwlwifi/pcie/tx.c | 13 +++++++++++++
> > >  1 file changed, 13 insertions(+)
> >
> > Ok.  I can grab these.
> >
> > They don't seem to be in the net_39.mbox that DaveM sent to stable a
> > couple days ago.  Are they still in someone else's tree at the moment?
>
> They were refused by upstream maintainers as fixes proper to 3.10-rc5 or
> later -rc. Instead patches were pushed to -next and queued to 3.11. Anyway
> fixes are small, look sane and are intended to -stable.

OK.  I've applied them to all branches on 3.9.  Thanks!

josh