iwlwifi: fix dma mappings and skbs leak

Since commit commit 470058e0ad82fcfaaffd57307d8bf8c094e8e9d7
"iwlwifi: avoid Tx queue memory allocation in interface down" we do
not unmap dma and free skbs when down device and there is pending
transfer. What in consequence may cause that system hung (waiting
for free skb's) when performing shutdown at iptables module unload.

DMA leak manifest itself following warning:

WARNING: at lib/dma-debug.c:689 dma_debug_device_change+0x15a/0x1b0()
Hardware name: HP xw8600 Workstation
pci 0000:80:00.0: DMA-API: device driver has pending DMA allocations while released from device [count=240]
Modules linked in: iwlagn(-) aes_x86_64 aes_generic fuse cpufreq_ondemand acpi_cpufreq freq_table mperf xt_physdev ipt_REJECT nf_conntrack_ipv4 nf_defrag_ipv4 iptable_filter ip_tables ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables ipv6 ext3 jbd dm_mirror dm_region_hash dm_log dm_mod uinput hp_wmi sparse_keymap sg wmi microcode serio_raw tg3 arc4 ecb shpchp mac80211 cfg80211 rfkill ext4 mbcache jbd2 sr_mod cdrom sd_mod crc_t10dif firewire_ohci firewire_core crc_itu_t mptsas mptscsih mptbase scsi_transport_sas pata_acpi ata_generic ata_piix ahci libahci floppy nouveau ttm drm_kms_helper drm i2c_algo_bit i2c_core video [last unloaded: iwlagn]
Pid: 9131, comm: rmmod Tainted: G        W   2.6.38-rc6-wl+ #33
Call Trace:
 [<ffffffff810649ef>] ? warn_slowpath_common+0x7f/0xc0
 [<ffffffff81064ae6>] ? warn_slowpath_fmt+0x46/0x50
 [<ffffffff812320ab>] ? dma_debug_device_change+0xdb/0x1b0
 [<ffffffff8123212a>] ? dma_debug_device_change+0x15a/0x1b0
 [<ffffffff8149dc18>] ? notifier_call_chain+0x58/0xb0
 [<ffffffff8108e370>] ? __blocking_notifier_call_chain+0x60/0x90
 [<ffffffff8108e3b6>] ? blocking_notifier_call_chain+0x16/0x20
 [<ffffffff812f570c>] ? __device_release_driver+0xbc/0xe0
 [<ffffffff812f5808>] ? driver_detach+0xd8/0xe0
 [<ffffffff812f45d1>] ? bus_remove_driver+0x91/0x100
 [<ffffffff812f6022>] ? driver_unregister+0x62/0xa0
 [<ffffffff8123d5d4>] ? pci_unregister_driver+0x44/0xa0
 [<ffffffffa05632d1>] ? iwl_exit+0x15/0x1c [iwlagn]
 [<ffffffff810ab492>] ? sys_delete_module+0x1a2/0x270
 [<ffffffff81498da9>] ? trace_hardirqs_on_thunk+0x3a/0x3f
 [<ffffffff8100bf42>] ? system_call_fastpath+0x16/0x1b

I still can observe above warning after apply patch, but it is very
hard to reproduce it, and have count=1. Whereas that one is easy to
reproduce using debugfs force_reset while transmitting data, and have
very big counts eg. 240, like quoted here. So count=1 WARNING seems
to be different issue that need to be resolved separately.

v1 -> v2: fix infinity loop bug I made during "for" to "while" loop transition.
v2 -> v3: remove unneeded EXPORT_SYMBOL

Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

3 files changed, 59 insertions, 26 deletions

diff --git a/drivers/net/wireless/iwlwifi/iwl-agn-tx.c b/drivers/net/wireless/iwlwifi/iwl-agn-tx.c
index 266490d8a397..a709d05c5868 100644
--- a/drivers/net/wireless/iwlwifi/iwl-agn-tx.c
+++ b/drivers/net/wireless/iwlwifi/iwl-agn-tx.c
@@ -947,7 +947,7 @@ void iwlagn_txq_ctx_reset(struct iwl_priv *priv)
  */
 void iwlagn_txq_ctx_stop(struct iwl_priv *priv)
 {
-	int ch;
+	int ch, txq_id;
 	unsigned long flags;
 
 	/* Turn off all Tx DMA fifos */
@@ -966,6 +966,16 @@ void iwlagn_txq_ctx_stop(struct iwl_priv *priv)
 			    iwl_read_direct32(priv, FH_TSSR_TX_STATUS_REG));
 	}
 	spin_unlock_irqrestore(&priv->lock, flags);
+
+	if (!priv->txq)
+		return;
+
+	/* Unmap DMA from host system and free skb's */
+	for (txq_id = 0; txq_id < priv->hw_params.max_txq_num; txq_id++)
+		if (txq_id == priv->cmd_queue)
+			iwl_cmd_queue_unmap(priv);
+		else
+			iwl_tx_queue_unmap(priv, txq_id);
 }
 
 /*
diff --git a/drivers/net/wireless/iwlwifi/iwl-core.h b/drivers/net/wireless/iwlwifi/iwl-core.h
index 909b42d5d9c0..ce368d8d402b 100644
--- a/drivers/net/wireless/iwlwifi/iwl-core.h
+++ b/drivers/net/wireless/iwlwifi/iwl-core.h
@@ -510,6 +510,7 @@ void iwl_rx_reply_error(struct iwl_priv *priv,
 * RX
 ******************************************************/
 void iwl_cmd_queue_free(struct iwl_priv *priv);
+void iwl_cmd_queue_unmap(struct iwl_priv *priv);
 int iwl_rx_queue_alloc(struct iwl_priv *priv);
 void iwl_rx_queue_update_write_ptr(struct iwl_priv *priv,
 				  struct iwl_rx_queue *q);
@@ -534,6 +535,7 @@ int iwl_tx_queue_init(struct iwl_priv *priv, struct iwl_tx_queue *txq,
 void iwl_tx_queue_reset(struct iwl_priv *priv, struct iwl_tx_queue *txq,
 			int slots_num, u32 txq_id);
 void iwl_tx_queue_free(struct iwl_priv *priv, int txq_id);
+void iwl_tx_queue_unmap(struct iwl_priv *priv, int txq_id);
 void iwl_setup_watchdog(struct iwl_priv *priv);
 /*****************************************************
  * TX power
diff --git a/drivers/net/wireless/iwlwifi/iwl-tx.c b/drivers/net/wireless/iwlwifi/iwl-tx.c
index 7e607d39da1c..277c9175dcf6 100644
--- a/drivers/net/wireless/iwlwifi/iwl-tx.c
+++ b/drivers/net/wireless/iwlwifi/iwl-tx.c
@@ -86,6 +86,23 @@ void iwl_txq_update_write_ptr(struct iwl_priv *priv, struct iwl_tx_queue *txq)
 }
 
 /**
+ * iwl_tx_queue_unmap -  Unmap any remaining DMA mappings and free skb's
+ */
+void iwl_tx_queue_unmap(struct iwl_priv *priv, int txq_id)
+{
+	struct iwl_tx_queue *txq = &priv->txq[txq_id];
+	struct iwl_queue *q = &txq->q;
+
+	if (q->n_bd == 0)
+		return;
+
+	 while (q->write_ptr != q->read_ptr) {
+		priv->cfg->ops->lib->txq_free_tfd(priv, txq);
+		q->read_ptr = iwl_queue_inc_wrap(q->read_ptr, q->n_bd);
+	}
+}
+
+/**
  * iwl_tx_queue_free - Deallocate DMA queue.
  * @txq: Transmit queue to deallocate.
  *
@@ -96,17 +113,10 @@ void iwl_txq_update_write_ptr(struct iwl_priv *priv, struct iwl_tx_queue *txq)
 void iwl_tx_queue_free(struct iwl_priv *priv, int txq_id)
 {
 	struct iwl_tx_queue *txq = &priv->txq[txq_id];
-	struct iwl_queue *q = &txq->q;
 	struct device *dev = &priv->pci_dev->dev;
 	int i;
 
-	if (q->n_bd == 0)
-		return;
-
-	/* first, empty all BD's */
-	for (; q->write_ptr != q->read_ptr;
-	     q->read_ptr = iwl_queue_inc_wrap(q->read_ptr, q->n_bd))
-		priv->cfg->ops->lib->txq_free_tfd(priv, txq);
+	iwl_tx_queue_unmap(priv, txq_id);
 
 	/* De-alloc array of command/tx buffers */
 	for (i = 0; i < TFD_TX_CMD_SLOTS; i++)
@@ -132,39 +142,33 @@ void iwl_tx_queue_free(struct iwl_priv *priv, int txq_id)
 }
 
 /**
- * iwl_cmd_queue_free - Deallocate DMA queue.
- * @txq: Transmit queue to deallocate.
- *
- * Empty queue by removing and destroying all BD's.
- * Free all buffers.
- * 0-fill, but do not free "txq" descriptor structure.
+ * iwl_cmd_queue_unmap - Unmap any remaining DMA mappings from command queue
  */
-void iwl_cmd_queue_free(struct iwl_priv *priv)
+void iwl_cmd_queue_unmap(struct iwl_priv *priv)
 {
 	struct iwl_tx_queue *txq = &priv->txq[priv->cmd_queue];
 	struct iwl_queue *q = &txq->q;
-	struct device *dev = &priv->pci_dev->dev;
 	int i;
 	bool huge = false;
 
 	if (q->n_bd == 0)
 		return;
 
-	for (; q->read_ptr != q->write_ptr;
-	     q->read_ptr = iwl_queue_inc_wrap(q->read_ptr, q->n_bd)) {
+	while (q->read_ptr != q->write_ptr) {
 		/* we have no way to tell if it is a huge cmd ATM */
 		i = get_cmd_index(q, q->read_ptr, 0);
 
-		if (txq->meta[i].flags & CMD_SIZE_HUGE) {
+		if (txq->meta[i].flags & CMD_SIZE_HUGE)
 			huge = true;
-			continue;
-		}
+		else
+			pci_unmap_single(priv->pci_dev,
+					 dma_unmap_addr(&txq->meta[i], mapping),
+					 dma_unmap_len(&txq->meta[i], len),
+					 PCI_DMA_BIDIRECTIONAL);
 
-		pci_unmap_single(priv->pci_dev,
-				 dma_unmap_addr(&txq->meta[i], mapping),
-				 dma_unmap_len(&txq->meta[i], len),
-				 PCI_DMA_BIDIRECTIONAL);
+	     q->read_ptr = iwl_queue_inc_wrap(q->read_ptr, q->n_bd);
 	}
+
 	if (huge) {
 		i = q->n_window;
 		pci_unmap_single(priv->pci_dev,
@@ -172,6 +176,23 @@ void iwl_cmd_queue_free(struct iwl_priv *priv)
 				 dma_unmap_len(&txq->meta[i], len),
 				 PCI_DMA_BIDIRECTIONAL);
 	}
+}
+
+/**
+ * iwl_cmd_queue_free - Deallocate DMA queue.
+ * @txq: Transmit queue to deallocate.
+ *
+ * Empty queue by removing and destroying all BD's.
+ * Free all buffers.
+ * 0-fill, but do not free "txq" descriptor structure.
+ */
+void iwl_cmd_queue_free(struct iwl_priv *priv)
+{
+	struct iwl_tx_queue *txq = &priv->txq[priv->cmd_queue];
+	struct device *dev = &priv->pci_dev->dev;
+	int i;
+
+	iwl_cmd_queue_unmap(priv);
 
 	/* De-alloc array of command/tx buffers */
 	for (i = 0; i <= TFD_CMD_SLOTS; i++)