Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking changes from David Miller:

1) Allow to dump, monitor, and change the bridge multicast database
   using netlink.  From Cong Wang.

2) RFC 5961 TCP blind data injection attack mitigation, from Eric
   Dumazet.

3) Networking user namespace support from Eric W. Biederman.

4) tuntap/virtio-net multiqueue support by Jason Wang.

5) Support for checksum offload of encapsulated packets (basically,
   tunneled traffic can still be checksummed by HW).  From Joseph
   Gasparakis.

6) Allow BPF filter access to VLAN tags, from Eric Dumazet and
   Daniel Borkmann.

7) Bridge port parameters over netlink and BPDU blocking support
   from Stephen Hemminger.

8) Improve data access patterns during inet socket demux by rearranging
   socket layout, from Eric Dumazet.

9) TIPC protocol updates and cleanups from Ying Xue, Paul Gortmaker, and
   Jon Maloy.

10) Update TCP socket hash sizing to be more in line with current day
    realities.  The existing heurstics were choosen a decade ago.
    From Eric Dumazet.

11) Fix races, queue bloat, and excessive wakeups in ATM and
    associated drivers, from Krzysztof Mazur and David Woodhouse.

12) Support DOVE (Distributed Overlay Virtual Ethernet) extensions
    in VXLAN driver, from David Stevens.

13) Add "oops_only" mode to netconsole, from Amerigo Wang.

14) Support set and query of VEB/VEPA bridge mode via PF_BRIDGE, also
    allow DCB netlink to work on namespaces other than the initial
    namespace.  From John Fastabend.

15) Support PTP in the Tigon3 driver, from Matt Carlson.

16) tun/vhost zero copy fixes and improvements, plus turn it on
    by default, from Michael S. Tsirkin.

17) Support per-association statistics in SCTP, from Michele
    Baldessari.

And many, many, driver updates, cleanups, and improvements.  Too
numerous to mention individually.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1722 commits)
  net/mlx4_en: Add support for destination MAC in steering rules
  net/mlx4_en: Use generic etherdevice.h functions.
  net: ethtool: Add destination MAC address to flow steering API
  bridge: add support of adding and deleting mdb entries
  bridge: notify mdb changes via netlink
  ndisc: Unexport ndisc_{build,send}_skb().
  uapi: add missing netconf.h to export list
  pkt_sched: avoid requeues if possible
  solos-pci: fix double-free of TX skb in DMA mode
  bnx2: Fix accidental reversions.
  bna: Driver Version Updated to 3.1.2.1
  bna: Firmware update
  bna: Add RX State
  bna: Rx Page Based Allocation
  bna: TX Intr Coalescing Fix
  bna: Tx and Rx Optimizations
  bna: Code Cleanup and Enhancements
  ath9k: check pdata variable before dereferencing it
  ath5k: RX timestamp is reported at end of frame
  ath9k_htc: RX timestamp is reported at end of frame
  ...

1 files changed, 95 insertions, 969 deletions

diff --git a/drivers/net/wireless/iwlwifi/pcie/trans.c b/drivers/net/wireless/iwlwifi/pcie/trans.c
index fe0fffd04304..d66cad4a7d6a 100644
--- a/drivers/net/wireless/iwlwifi/pcie/trans.c
+++ b/drivers/net/wireless/iwlwifi/pcie/trans.c
@@ -74,584 +74,8 @@
 #include "iwl-prph.h"
 #include "iwl-agn-hw.h"
 #include "internal.h"
-/* FIXME: need to abstract out TX command (once we know what it looks like) */
-#include "dvm/commands.h"
 
-#define SCD_QUEUECHAIN_SEL_ALL(trans, trans_pcie)	\
-	(((1<<trans->cfg->base_params->num_of_queues) - 1) &\
-	(~(1<<(trans_pcie)->cmd_queue)))
-
-static int iwl_trans_rx_alloc(struct iwl_trans *trans)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_rx_queue *rxq = &trans_pcie->rxq;
-	struct device *dev = trans->dev;
-
-	memset(&trans_pcie->rxq, 0, sizeof(trans_pcie->rxq));
-
-	spin_lock_init(&rxq->lock);
-
-	if (WARN_ON(rxq->bd || rxq->rb_stts))
-		return -EINVAL;
-
-	/* Allocate the circular buffer of Read Buffer Descriptors (RBDs) */
-	rxq->bd = dma_zalloc_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE,
-				      &rxq->bd_dma, GFP_KERNEL);
-	if (!rxq->bd)
-		goto err_bd;
-
-	/*Allocate the driver's pointer to receive buffer status */
-	rxq->rb_stts = dma_zalloc_coherent(dev, sizeof(*rxq->rb_stts),
-					   &rxq->rb_stts_dma, GFP_KERNEL);
-	if (!rxq->rb_stts)
-		goto err_rb_stts;
-
-	return 0;
-
-err_rb_stts:
-	dma_free_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE,
-			  rxq->bd, rxq->bd_dma);
-	memset(&rxq->bd_dma, 0, sizeof(rxq->bd_dma));
-	rxq->bd = NULL;
-err_bd:
-	return -ENOMEM;
-}
-
-static void iwl_trans_rxq_free_rx_bufs(struct iwl_trans *trans)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_rx_queue *rxq = &trans_pcie->rxq;
-	int i;
-
-	/* Fill the rx_used queue with _all_ of the Rx buffers */
-	for (i = 0; i < RX_FREE_BUFFERS + RX_QUEUE_SIZE; i++) {
-		/* In the reset function, these buffers may have been allocated
-		 * to an SKB, so we need to unmap and free potential storage */
-		if (rxq->pool[i].page != NULL) {
-			dma_unmap_page(trans->dev, rxq->pool[i].page_dma,
-				       PAGE_SIZE << trans_pcie->rx_page_order,
-				       DMA_FROM_DEVICE);
-			__free_pages(rxq->pool[i].page,
-				     trans_pcie->rx_page_order);
-			rxq->pool[i].page = NULL;
-		}
-		list_add_tail(&rxq->pool[i].list, &rxq->rx_used);
-	}
-}
-
-static void iwl_trans_rx_hw_init(struct iwl_trans *trans,
-				 struct iwl_rx_queue *rxq)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	u32 rb_size;
-	const u32 rfdnlog = RX_QUEUE_SIZE_LOG; /* 256 RBDs */
-	u32 rb_timeout = RX_RB_TIMEOUT; /* FIXME: RX_RB_TIMEOUT for all devices? */
-
-	if (trans_pcie->rx_buf_size_8k)
-		rb_size = FH_RCSR_RX_CONFIG_REG_VAL_RB_SIZE_8K;
-	else
-		rb_size = FH_RCSR_RX_CONFIG_REG_VAL_RB_SIZE_4K;
-
-	/* Stop Rx DMA */
-	iwl_write_direct32(trans, FH_MEM_RCSR_CHNL0_CONFIG_REG, 0);
-
-	/* Reset driver's Rx queue write index */
-	iwl_write_direct32(trans, FH_RSCSR_CHNL0_RBDCB_WPTR_REG, 0);
-
-	/* Tell device where to find RBD circular buffer in DRAM */
-	iwl_write_direct32(trans, FH_RSCSR_CHNL0_RBDCB_BASE_REG,
-			   (u32)(rxq->bd_dma >> 8));
-
-	/* Tell device where in DRAM to update its Rx status */
-	iwl_write_direct32(trans, FH_RSCSR_CHNL0_STTS_WPTR_REG,
-			   rxq->rb_stts_dma >> 4);
-
-	/* Enable Rx DMA
-	 * FH_RCSR_CHNL0_RX_IGNORE_RXF_EMPTY is set because of HW bug in
-	 *      the credit mechanism in 5000 HW RX FIFO
-	 * Direct rx interrupts to hosts
-	 * Rx buffer size 4 or 8k
-	 * RB timeout 0x10
-	 * 256 RBDs
-	 */
-	iwl_write_direct32(trans, FH_MEM_RCSR_CHNL0_CONFIG_REG,
-			   FH_RCSR_RX_CONFIG_CHNL_EN_ENABLE_VAL |
-			   FH_RCSR_CHNL0_RX_IGNORE_RXF_EMPTY |
-			   FH_RCSR_CHNL0_RX_CONFIG_IRQ_DEST_INT_HOST_VAL |
-			   rb_size|
-			   (rb_timeout << FH_RCSR_RX_CONFIG_REG_IRQ_RBTH_POS)|
-			   (rfdnlog << FH_RCSR_RX_CONFIG_RBDCB_SIZE_POS));
-
-	/* Set interrupt coalescing timer to default (2048 usecs) */
-	iwl_write8(trans, CSR_INT_COALESCING, IWL_HOST_INT_TIMEOUT_DEF);
-}
-
-static int iwl_rx_init(struct iwl_trans *trans)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_rx_queue *rxq = &trans_pcie->rxq;
-
-	int i, err;
-	unsigned long flags;
-
-	if (!rxq->bd) {
-		err = iwl_trans_rx_alloc(trans);
-		if (err)
-			return err;
-	}
-
-	spin_lock_irqsave(&rxq->lock, flags);
-	INIT_LIST_HEAD(&rxq->rx_free);
-	INIT_LIST_HEAD(&rxq->rx_used);
-
-	iwl_trans_rxq_free_rx_bufs(trans);
-
-	for (i = 0; i < RX_QUEUE_SIZE; i++)
-		rxq->queue[i] = NULL;
-
-	/* Set us so that we have processed and used all buffers, but have
-	 * not restocked the Rx queue with fresh buffers */
-	rxq->read = rxq->write = 0;
-	rxq->write_actual = 0;
-	rxq->free_count = 0;
-	spin_unlock_irqrestore(&rxq->lock, flags);
-
-	iwl_rx_replenish(trans);
-
-	iwl_trans_rx_hw_init(trans, rxq);
-
-	spin_lock_irqsave(&trans_pcie->irq_lock, flags);
-	rxq->need_update = 1;
-	iwl_rx_queue_update_write_ptr(trans, rxq);
-	spin_unlock_irqrestore(&trans_pcie->irq_lock, flags);
-
-	return 0;
-}
-
-static void iwl_trans_pcie_rx_free(struct iwl_trans *trans)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_rx_queue *rxq = &trans_pcie->rxq;
-	unsigned long flags;
-
-	/*if rxq->bd is NULL, it means that nothing has been allocated,
-	 * exit now */
-	if (!rxq->bd) {
-		IWL_DEBUG_INFO(trans, "Free NULL rx context\n");
-		return;
-	}
-
-	spin_lock_irqsave(&rxq->lock, flags);
-	iwl_trans_rxq_free_rx_bufs(trans);
-	spin_unlock_irqrestore(&rxq->lock, flags);
-
-	dma_free_coherent(trans->dev, sizeof(__le32) * RX_QUEUE_SIZE,
-			  rxq->bd, rxq->bd_dma);
-	memset(&rxq->bd_dma, 0, sizeof(rxq->bd_dma));
-	rxq->bd = NULL;
-
-	if (rxq->rb_stts)
-		dma_free_coherent(trans->dev,
-				  sizeof(struct iwl_rb_status),
-				  rxq->rb_stts, rxq->rb_stts_dma);
-	else
-		IWL_DEBUG_INFO(trans, "Free rxq->rb_stts which is NULL\n");
-	memset(&rxq->rb_stts_dma, 0, sizeof(rxq->rb_stts_dma));
-	rxq->rb_stts = NULL;
-}
-
-static int iwl_trans_rx_stop(struct iwl_trans *trans)
-{
-
-	/* stop Rx DMA */
-	iwl_write_direct32(trans, FH_MEM_RCSR_CHNL0_CONFIG_REG, 0);
-	return iwl_poll_direct_bit(trans, FH_MEM_RSSR_RX_STATUS_REG,
-				   FH_RSSR_CHNL0_RX_STATUS_CHNL_IDLE, 1000);
-}
-
-static int iwlagn_alloc_dma_ptr(struct iwl_trans *trans,
-				struct iwl_dma_ptr *ptr, size_t size)
-{
-	if (WARN_ON(ptr->addr))
-		return -EINVAL;
-
-	ptr->addr = dma_alloc_coherent(trans->dev, size,
-				       &ptr->dma, GFP_KERNEL);
-	if (!ptr->addr)
-		return -ENOMEM;
-	ptr->size = size;
-	return 0;
-}
-
-static void iwlagn_free_dma_ptr(struct iwl_trans *trans,
-				struct iwl_dma_ptr *ptr)
-{
-	if (unlikely(!ptr->addr))
-		return;
-
-	dma_free_coherent(trans->dev, ptr->size, ptr->addr, ptr->dma);
-	memset(ptr, 0, sizeof(*ptr));
-}
-
-static void iwl_trans_pcie_queue_stuck_timer(unsigned long data)
-{
-	struct iwl_tx_queue *txq = (void *)data;
-	struct iwl_queue *q = &txq->q;
-	struct iwl_trans_pcie *trans_pcie = txq->trans_pcie;
-	struct iwl_trans *trans = iwl_trans_pcie_get_trans(trans_pcie);
-	u32 scd_sram_addr = trans_pcie->scd_base_addr +
-		SCD_TX_STTS_MEM_LOWER_BOUND + (16 * txq->q.id);
-	u8 buf[16];
-	int i;
-
-	spin_lock(&txq->lock);
-	/* check if triggered erroneously */
-	if (txq->q.read_ptr == txq->q.write_ptr) {
-		spin_unlock(&txq->lock);
-		return;
-	}
-	spin_unlock(&txq->lock);
-
-	IWL_ERR(trans, "Queue %d stuck for %u ms.\n", txq->q.id,
-		jiffies_to_msecs(trans_pcie->wd_timeout));
-	IWL_ERR(trans, "Current SW read_ptr %d write_ptr %d\n",
-		txq->q.read_ptr, txq->q.write_ptr);
-
-	iwl_read_targ_mem_bytes(trans, scd_sram_addr, buf, sizeof(buf));
-
-	iwl_print_hex_error(trans, buf, sizeof(buf));
-
-	for (i = 0; i < FH_TCSR_CHNL_NUM; i++)
-		IWL_ERR(trans, "FH TRBs(%d) = 0x%08x\n", i,
-			iwl_read_direct32(trans, FH_TX_TRB_REG(i)));
-
-	for (i = 0; i < trans->cfg->base_params->num_of_queues; i++) {
-		u32 status = iwl_read_prph(trans, SCD_QUEUE_STATUS_BITS(i));
-		u8 fifo = (status >> SCD_QUEUE_STTS_REG_POS_TXF) & 0x7;
-		bool active = !!(status & BIT(SCD_QUEUE_STTS_REG_POS_ACTIVE));
-		u32 tbl_dw =
-			iwl_read_targ_mem(trans,
-					  trans_pcie->scd_base_addr +
-					  SCD_TRANS_TBL_OFFSET_QUEUE(i));
-
-		if (i & 0x1)
-			tbl_dw = (tbl_dw & 0xFFFF0000) >> 16;
-		else
-			tbl_dw = tbl_dw & 0x0000FFFF;
-
-		IWL_ERR(trans,
-			"Q %d is %sactive and mapped to fifo %d ra_tid 0x%04x [%d,%d]\n",
-			i, active ? "" : "in", fifo, tbl_dw,
-			iwl_read_prph(trans,
-				      SCD_QUEUE_RDPTR(i)) & (txq->q.n_bd - 1),
-			iwl_read_prph(trans, SCD_QUEUE_WRPTR(i)));
-	}
-
-	for (i = q->read_ptr; i != q->write_ptr;
-	     i = iwl_queue_inc_wrap(i, q->n_bd)) {
-		struct iwl_tx_cmd *tx_cmd =
-			(struct iwl_tx_cmd *)txq->entries[i].cmd->payload;
-		IWL_ERR(trans, "scratch %d = 0x%08x\n", i,
-			get_unaligned_le32(&tx_cmd->scratch));
-	}
-
-	iwl_op_mode_nic_error(trans->op_mode);
-}
-
-static int iwl_trans_txq_alloc(struct iwl_trans *trans,
-			       struct iwl_tx_queue *txq, int slots_num,
-			       u32 txq_id)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	size_t tfd_sz = sizeof(struct iwl_tfd) * TFD_QUEUE_SIZE_MAX;
-	int i;
-
-	if (WARN_ON(txq->entries || txq->tfds))
-		return -EINVAL;
-
-	setup_timer(&txq->stuck_timer, iwl_trans_pcie_queue_stuck_timer,
-		    (unsigned long)txq);
-	txq->trans_pcie = trans_pcie;
-
-	txq->q.n_window = slots_num;
-
-	txq->entries = kcalloc(slots_num,
-			       sizeof(struct iwl_pcie_tx_queue_entry),
-			       GFP_KERNEL);
-
-	if (!txq->entries)
-		goto error;
-
-	if (txq_id == trans_pcie->cmd_queue)
-		for (i = 0; i < slots_num; i++) {
-			txq->entries[i].cmd =
-				kmalloc(sizeof(struct iwl_device_cmd),
-					GFP_KERNEL);
-			if (!txq->entries[i].cmd)
-				goto error;
-		}
-
-	/* Circular buffer of transmit frame descriptors (TFDs),
-	 * shared with device */
-	txq->tfds = dma_alloc_coherent(trans->dev, tfd_sz,
-				       &txq->q.dma_addr, GFP_KERNEL);
-	if (!txq->tfds) {
-		IWL_ERR(trans, "dma_alloc_coherent(%zd) failed\n", tfd_sz);
-		goto error;
-	}
-	txq->q.id = txq_id;
-
-	return 0;
-error:
-	if (txq->entries && txq_id == trans_pcie->cmd_queue)
-		for (i = 0; i < slots_num; i++)
-			kfree(txq->entries[i].cmd);
-	kfree(txq->entries);
-	txq->entries = NULL;
-
-	return -ENOMEM;
-
-}
-
-static int iwl_trans_txq_init(struct iwl_trans *trans, struct iwl_tx_queue *txq,
-			      int slots_num, u32 txq_id)
-{
-	int ret;
-
-	txq->need_update = 0;
-
-	/* TFD_QUEUE_SIZE_MAX must be power-of-two size, otherwise
-	 * iwl_queue_inc_wrap and iwl_queue_dec_wrap are broken. */
-	BUILD_BUG_ON(TFD_QUEUE_SIZE_MAX & (TFD_QUEUE_SIZE_MAX - 1));
-
-	/* Initialize queue's high/low-water marks, and head/tail indexes */
-	ret = iwl_queue_init(&txq->q, TFD_QUEUE_SIZE_MAX, slots_num,
-			txq_id);
-	if (ret)
-		return ret;
-
-	spin_lock_init(&txq->lock);
-
-	/*
-	 * Tell nic where to find circular buffer of Tx Frame Descriptors for
-	 * given Tx queue, and enable the DMA channel used for that queue.
-	 * Circular buffer (TFD queue in DRAM) physical base address */
-	iwl_write_direct32(trans, FH_MEM_CBBC_QUEUE(txq_id),
-			     txq->q.dma_addr >> 8);
-
-	return 0;
-}
-
-/**
- * iwl_tx_queue_unmap -  Unmap any remaining DMA mappings and free skb's
- */
-static void iwl_tx_queue_unmap(struct iwl_trans *trans, int txq_id)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_tx_queue *txq = &trans_pcie->txq[txq_id];
-	struct iwl_queue *q = &txq->q;
-	enum dma_data_direction dma_dir;
-
-	if (!q->n_bd)
-		return;
-
-	/* In the command queue, all the TBs are mapped as BIDI
-	 * so unmap them as such.
-	 */
-	if (txq_id == trans_pcie->cmd_queue)
-		dma_dir = DMA_BIDIRECTIONAL;
-	else
-		dma_dir = DMA_TO_DEVICE;
-
-	spin_lock_bh(&txq->lock);
-	while (q->write_ptr != q->read_ptr) {
-		iwl_txq_free_tfd(trans, txq, dma_dir);
-		q->read_ptr = iwl_queue_inc_wrap(q->read_ptr, q->n_bd);
-	}
-	spin_unlock_bh(&txq->lock);
-}
-
-/**
- * iwl_tx_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.
- */
-static void iwl_tx_queue_free(struct iwl_trans *trans, int txq_id)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_tx_queue *txq = &trans_pcie->txq[txq_id];
-	struct device *dev = trans->dev;
-	int i;
-
-	if (WARN_ON(!txq))
-		return;
-
-	iwl_tx_queue_unmap(trans, txq_id);
-
-	/* De-alloc array of command/tx buffers */
-	if (txq_id == trans_pcie->cmd_queue)
-		for (i = 0; i < txq->q.n_window; i++) {
-			kfree(txq->entries[i].cmd);
-			kfree(txq->entries[i].copy_cmd);
-		}
-
-	/* De-alloc circular buffer of TFDs */
-	if (txq->q.n_bd) {
-		dma_free_coherent(dev, sizeof(struct iwl_tfd) *
-				  txq->q.n_bd, txq->tfds, txq->q.dma_addr);
-		memset(&txq->q.dma_addr, 0, sizeof(txq->q.dma_addr));
-	}
-
-	kfree(txq->entries);
-	txq->entries = NULL;
-
-	del_timer_sync(&txq->stuck_timer);
-
-	/* 0-fill queue descriptor structure */
-	memset(txq, 0, sizeof(*txq));
-}
-
-/**
- * iwl_trans_tx_free - Free TXQ Context
- *
- * Destroy all TX DMA queues and structures
- */
-static void iwl_trans_pcie_tx_free(struct iwl_trans *trans)
-{
-	int txq_id;
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-
-	/* Tx queues */
-	if (trans_pcie->txq) {
-		for (txq_id = 0;
-		     txq_id < trans->cfg->base_params->num_of_queues; txq_id++)
-			iwl_tx_queue_free(trans, txq_id);
-	}
-
-	kfree(trans_pcie->txq);
-	trans_pcie->txq = NULL;
-
-	iwlagn_free_dma_ptr(trans, &trans_pcie->kw);
-
-	iwlagn_free_dma_ptr(trans, &trans_pcie->scd_bc_tbls);
-}
-
-/**
- * iwl_trans_tx_alloc - allocate TX context
- * Allocate all Tx DMA structures and initialize them
- *
- * @param priv
- * @return error code
- */
-static int iwl_trans_tx_alloc(struct iwl_trans *trans)
-{
-	int ret;
-	int txq_id, slots_num;
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-
-	u16 scd_bc_tbls_size = trans->cfg->base_params->num_of_queues *
-			sizeof(struct iwlagn_scd_bc_tbl);
-
-	/*It is not allowed to alloc twice, so warn when this happens.
-	 * We cannot rely on the previous allocation, so free and fail */
-	if (WARN_ON(trans_pcie->txq)) {
-		ret = -EINVAL;
-		goto error;
-	}
-
-	ret = iwlagn_alloc_dma_ptr(trans, &trans_pcie->scd_bc_tbls,
-				   scd_bc_tbls_size);
-	if (ret) {
-		IWL_ERR(trans, "Scheduler BC Table allocation failed\n");
-		goto error;
-	}
-
-	/* Alloc keep-warm buffer */
-	ret = iwlagn_alloc_dma_ptr(trans, &trans_pcie->kw, IWL_KW_SIZE);
-	if (ret) {
-		IWL_ERR(trans, "Keep Warm allocation failed\n");
-		goto error;
-	}
-
-	trans_pcie->txq = kcalloc(trans->cfg->base_params->num_of_queues,
-				  sizeof(struct iwl_tx_queue), GFP_KERNEL);
-	if (!trans_pcie->txq) {
-		IWL_ERR(trans, "Not enough memory for txq\n");
-		ret = ENOMEM;
-		goto error;
-	}
-
-	/* Alloc and init all Tx queues, including the command queue (#4/#9) */
-	for (txq_id = 0; txq_id < trans->cfg->base_params->num_of_queues;
-	     txq_id++) {
-		slots_num = (txq_id == trans_pcie->cmd_queue) ?
-					TFD_CMD_SLOTS : TFD_TX_CMD_SLOTS;
-		ret = iwl_trans_txq_alloc(trans, &trans_pcie->txq[txq_id],
-					  slots_num, txq_id);
-		if (ret) {
-			IWL_ERR(trans, "Tx %d queue alloc failed\n", txq_id);
-			goto error;
-		}
-	}
-
-	return 0;
-
-error:
-	iwl_trans_pcie_tx_free(trans);
-
-	return ret;
-}
-static int iwl_tx_init(struct iwl_trans *trans)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	int ret;
-	int txq_id, slots_num;
-	unsigned long flags;
-	bool alloc = false;
-
-	if (!trans_pcie->txq) {
-		ret = iwl_trans_tx_alloc(trans);
-		if (ret)
-			goto error;
-		alloc = true;
-	}
-
-	spin_lock_irqsave(&trans_pcie->irq_lock, flags);
-
-	/* Turn off all Tx DMA fifos */
-	iwl_write_prph(trans, SCD_TXFACT, 0);
-
-	/* Tell NIC where to find the "keep warm" buffer */
-	iwl_write_direct32(trans, FH_KW_MEM_ADDR_REG,
-			   trans_pcie->kw.dma >> 4);
-
-	spin_unlock_irqrestore(&trans_pcie->irq_lock, flags);
-
-	/* Alloc and init all Tx queues, including the command queue (#4/#9) */
-	for (txq_id = 0; txq_id < trans->cfg->base_params->num_of_queues;
-	     txq_id++) {
-		slots_num = (txq_id == trans_pcie->cmd_queue) ?
-					TFD_CMD_SLOTS : TFD_TX_CMD_SLOTS;
-		ret = iwl_trans_txq_init(trans, &trans_pcie->txq[txq_id],
-					 slots_num, txq_id);
-		if (ret) {
-			IWL_ERR(trans, "Tx %d queue init failed\n", txq_id);
-			goto error;
-		}
-	}
-
-	return 0;
-error:
-	/*Upon error, free only if we allocated something */
-	if (alloc)
-		iwl_trans_pcie_tx_free(trans);
-	return ret;
-}
-
-static void iwl_set_pwr_vmain(struct iwl_trans *trans)
+static void iwl_pcie_set_pwr_vmain(struct iwl_trans *trans)
 {
 /*
  * (for documentation purposes)
@@ -673,18 +97,11 @@ static void iwl_set_pwr_vmain(struct iwl_trans *trans)
 #define PCI_CFG_LINK_CTRL_VAL_L0S_EN	0x01
 #define PCI_CFG_LINK_CTRL_VAL_L1_EN	0x02
 
-static u16 iwl_pciexp_link_ctrl(struct iwl_trans *trans)
+static void iwl_pcie_apm_config(struct iwl_trans *trans)
 {
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	u16 pci_lnk_ctl;
+	u16 lctl;
 
-	pcie_capability_read_word(trans_pcie->pci_dev, PCI_EXP_LNKCTL,
-				  &pci_lnk_ctl);
-	return pci_lnk_ctl;
-}
-
-static void iwl_apm_config(struct iwl_trans *trans)
-{
 	/*
 	 * HW bug W/A for instability in PCIe bus L0S->L1 transition.
 	 * Check if BIOS (or OS) enabled L1-ASPM on this device.
@@ -693,29 +110,27 @@ static void iwl_apm_config(struct iwl_trans *trans)
 	 * If not (unlikely), enable L0S, so there is at least some
 	 *    power savings, even without L1.
 	 */
-	u16 lctl = iwl_pciexp_link_ctrl(trans);
+	pcie_capability_read_word(trans_pcie->pci_dev, PCI_EXP_LNKCTL, &lctl);
 
 	if ((lctl & PCI_CFG_LINK_CTRL_VAL_L1_EN) ==
 				PCI_CFG_LINK_CTRL_VAL_L1_EN) {
 		/* L1-ASPM enabled; disable(!) L0S */
 		iwl_set_bit(trans, CSR_GIO_REG, CSR_GIO_REG_VAL_L0S_ENABLED);
-		dev_printk(KERN_INFO, trans->dev,
-			   "L1 Enabled; Disabling L0S\n");
+		dev_info(trans->dev, "L1 Enabled; Disabling L0S\n");
 	} else {
 		/* L1-ASPM disabled; enable(!) L0S */
 		iwl_clear_bit(trans, CSR_GIO_REG, CSR_GIO_REG_VAL_L0S_ENABLED);
-		dev_printk(KERN_INFO, trans->dev,
-			   "L1 Disabled; Enabling L0S\n");
+		dev_info(trans->dev, "L1 Disabled; Enabling L0S\n");
 	}
 	trans->pm_support = !(lctl & PCI_CFG_LINK_CTRL_VAL_L0S_EN);
 }
 
 /*
  * Start up NIC's basic functionality after it has been reset
- * (e.g. after platform boot, or shutdown via iwl_apm_stop())
+ * (e.g. after platform boot, or shutdown via iwl_pcie_apm_stop())
  * NOTE:  This does not load uCode nor start the embedded processor
  */
-static int iwl_apm_init(struct iwl_trans *trans)
+static int iwl_pcie_apm_init(struct iwl_trans *trans)
 {
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
 	int ret = 0;
@@ -747,7 +162,7 @@ static int iwl_apm_init(struct iwl_trans *trans)
 	iwl_set_bit(trans, CSR_HW_IF_CONFIG_REG,
 		    CSR_HW_IF_CONFIG_REG_BIT_HAP_WAKE_L1A);
 
-	iwl_apm_config(trans);
+	iwl_pcie_apm_config(trans);
 
 	/* Configure analog phase-lock-loop before activating to D0A */
 	if (trans->cfg->base_params->pll_cfg_val)
@@ -793,7 +208,7 @@ out:
 	return ret;
 }
 
-static int iwl_apm_stop_master(struct iwl_trans *trans)
+static int iwl_pcie_apm_stop_master(struct iwl_trans *trans)
 {
 	int ret = 0;
 
@@ -811,7 +226,7 @@ static int iwl_apm_stop_master(struct iwl_trans *trans)
 	return ret;
 }
 
-static void iwl_apm_stop(struct iwl_trans *trans)
+static void iwl_pcie_apm_stop(struct iwl_trans *trans)
 {
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
 	IWL_DEBUG_INFO(trans, "Stop card, put in low power state\n");
@@ -819,7 +234,7 @@ static void iwl_apm_stop(struct iwl_trans *trans)
 	clear_bit(STATUS_DEVICE_ENABLED, &trans_pcie->status);
 
 	/* Stop device's DMA activity */
-	iwl_apm_stop_master(trans);
+	iwl_pcie_apm_stop_master(trans);
 
 	/* Reset the entire device */
 	iwl_set_bit(trans, CSR_RESET, CSR_RESET_REG_FLAG_SW_RESET);
@@ -834,29 +249,29 @@ static void iwl_apm_stop(struct iwl_trans *trans)
 		      CSR_GP_CNTRL_REG_FLAG_INIT_DONE);
 }
 
-static int iwl_nic_init(struct iwl_trans *trans)
+static int iwl_pcie_nic_init(struct iwl_trans *trans)
 {
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
 	unsigned long flags;
 
 	/* nic_init */
 	spin_lock_irqsave(&trans_pcie->irq_lock, flags);
-	iwl_apm_init(trans);
+	iwl_pcie_apm_init(trans);
 
 	/* Set interrupt coalescing calibration timer to default (512 usecs) */
 	iwl_write8(trans, CSR_INT_COALESCING, IWL_HOST_INT_CALIB_TIMEOUT_DEF);
 
 	spin_unlock_irqrestore(&trans_pcie->irq_lock, flags);
 
-	iwl_set_pwr_vmain(trans);
+	iwl_pcie_set_pwr_vmain(trans);
 
 	iwl_op_mode_nic_config(trans->op_mode);
 
 	/* Allocate the RX queue, or reset if it is already allocated */
-	iwl_rx_init(trans);
+	iwl_pcie_rx_init(trans);
 
 	/* Allocate or reset and init all Tx and Command queues */
-	if (iwl_tx_init(trans))
+	if (iwl_pcie_tx_init(trans))
 		return -ENOMEM;
 
 	if (trans->cfg->base_params->shadow_reg_enable) {
@@ -871,7 +286,7 @@ static int iwl_nic_init(struct iwl_trans *trans)
 #define HW_READY_TIMEOUT (50)
 
 /* Note: returns poll_bit return value, which is >= 0 if success */
-static int iwl_set_hw_ready(struct iwl_trans *trans)
+static int iwl_pcie_set_hw_ready(struct iwl_trans *trans)
 {
 	int ret;
 
@@ -889,14 +304,14 @@ static int iwl_set_hw_ready(struct iwl_trans *trans)
 }
 
 /* Note: returns standard 0/-ERROR code */
-static int iwl_prepare_card_hw(struct iwl_trans *trans)
+static int iwl_pcie_prepare_card_hw(struct iwl_trans *trans)
 {
 	int ret;
 	int t = 0;
 
 	IWL_DEBUG_INFO(trans, "iwl_trans_prepare_card_hw enter\n");
 
-	ret = iwl_set_hw_ready(trans);
+	ret = iwl_pcie_set_hw_ready(trans);
 	/* If the card is ready, exit 0 */
 	if (ret >= 0)
 		return 0;
@@ -906,7 +321,7 @@ static int iwl_prepare_card_hw(struct iwl_trans *trans)
 		    CSR_HW_IF_CONFIG_REG_PREPARE);
 
 	do {
-		ret = iwl_set_hw_ready(trans);
+		ret = iwl_pcie_set_hw_ready(trans);
 		if (ret >= 0)
 			return 0;
 
@@ -920,7 +335,7 @@ static int iwl_prepare_card_hw(struct iwl_trans *trans)
 /*
  * ucode
  */
-static int iwl_load_firmware_chunk(struct iwl_trans *trans, u32 dst_addr,
+static int iwl_pcie_load_firmware_chunk(struct iwl_trans *trans, u32 dst_addr,
 				   dma_addr_t phy_addr, u32 byte_cnt)
 {
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
@@ -967,7 +382,7 @@ static int iwl_load_firmware_chunk(struct iwl_trans *trans, u32 dst_addr,
 	return 0;
 }
 
-static int iwl_load_section(struct iwl_trans *trans, u8 section_num,
+static int iwl_pcie_load_section(struct iwl_trans *trans, u8 section_num,
 			    const struct fw_desc *section)
 {
 	u8 *v_addr;
@@ -988,8 +403,9 @@ static int iwl_load_section(struct iwl_trans *trans, u8 section_num,
 		copy_size = min_t(u32, PAGE_SIZE, section->len - offset);
 
 		memcpy(v_addr, (u8 *)section->data + offset, copy_size);
-		ret = iwl_load_firmware_chunk(trans, section->offset + offset,
-					      p_addr, copy_size);
+		ret = iwl_pcie_load_firmware_chunk(trans,
+						   section->offset + offset,
+						   p_addr, copy_size);
 		if (ret) {
 			IWL_ERR(trans,
 				"Could not load the [%d] uCode section\n",
@@ -1002,7 +418,7 @@ static int iwl_load_section(struct iwl_trans *trans, u8 section_num,
 	return ret;
 }
 
-static int iwl_load_given_ucode(struct iwl_trans *trans,
+static int iwl_pcie_load_given_ucode(struct iwl_trans *trans,
 				const struct fw_img *image)
 {
 	int i, ret = 0;
@@ -1011,7 +427,7 @@ static int iwl_load_given_ucode(struct iwl_trans *trans,
 		if (!image->sec[i].data)
 			break;
 
-		ret = iwl_load_section(trans, i, &image->sec[i]);
+		ret = iwl_pcie_load_section(trans, i, &image->sec[i]);
 		if (ret)
 			return ret;
 	}
@@ -1025,15 +441,18 @@ static int iwl_load_given_ucode(struct iwl_trans *trans,
 static int iwl_trans_pcie_start_fw(struct iwl_trans *trans,
 				   const struct fw_img *fw)
 {
+	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
 	int ret;
 	bool hw_rfkill;
 
 	/* This may fail if AMT took ownership of the device */
-	if (iwl_prepare_card_hw(trans)) {
+	if (iwl_pcie_prepare_card_hw(trans)) {
 		IWL_WARN(trans, "Exit HW not ready\n");
 		return -EIO;
 	}
 
+	clear_bit(STATUS_FW_ERROR, &trans_pcie->status);
+
 	iwl_enable_rfkill_int(trans);
 
 	/* If platform's RF_KILL switch is NOT set to KILL */
@@ -1044,7 +463,7 @@ static int iwl_trans_pcie_start_fw(struct iwl_trans *trans,
 
 	iwl_write32(trans, CSR_INT, 0xFFFFFFFF);
 
-	ret = iwl_nic_init(trans);
+	ret = iwl_pcie_nic_init(trans);
 	if (ret) {
 		IWL_ERR(trans, "Unable to init nic\n");
 		return ret;
@@ -1064,125 +483,13 @@ static int iwl_trans_pcie_start_fw(struct iwl_trans *trans,
 	iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL);
 
 	/* Load the given image to the HW */
-	return iwl_load_given_ucode(trans, fw);
-}
-
-/*
- * Activate/Deactivate Tx DMA/FIFO channels according tx fifos mask
- */
-static void iwl_trans_txq_set_sched(struct iwl_trans *trans, u32 mask)
-{
-	struct iwl_trans_pcie __maybe_unused *trans_pcie =
-		IWL_TRANS_GET_PCIE_TRANS(trans);
-
-	iwl_write_prph(trans, SCD_TXFACT, mask);
-}
-
-static void iwl_tx_start(struct iwl_trans *trans)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	u32 a;
-	int chan;
-	u32 reg_val;
-
-	/* make sure all queue are not stopped/used */
-	memset(trans_pcie->queue_stopped, 0, sizeof(trans_pcie->queue_stopped));
-	memset(trans_pcie->queue_used, 0, sizeof(trans_pcie->queue_used));
-
-	trans_pcie->scd_base_addr =
-		iwl_read_prph(trans, SCD_SRAM_BASE_ADDR);
-	a = trans_pcie->scd_base_addr + SCD_CONTEXT_MEM_LOWER_BOUND;
-	/* reset conext data memory */
-	for (; a < trans_pcie->scd_base_addr + SCD_CONTEXT_MEM_UPPER_BOUND;
-		a += 4)
-		iwl_write_targ_mem(trans, a, 0);
-	/* reset tx status memory */
-	for (; a < trans_pcie->scd_base_addr + SCD_TX_STTS_MEM_UPPER_BOUND;
-		a += 4)
-		iwl_write_targ_mem(trans, a, 0);
-	for (; a < trans_pcie->scd_base_addr +
-	       SCD_TRANS_TBL_OFFSET_QUEUE(
-				trans->cfg->base_params->num_of_queues);
-	       a += 4)
-		iwl_write_targ_mem(trans, a, 0);
-
-	iwl_write_prph(trans, SCD_DRAM_BASE_ADDR,
-		       trans_pcie->scd_bc_tbls.dma >> 10);
-
-	/* The chain extension of the SCD doesn't work well. This feature is
-	 * enabled by default by the HW, so we need to disable it manually.
-	 */
-	iwl_write_prph(trans, SCD_CHAINEXT_EN, 0);
-
-	iwl_trans_ac_txq_enable(trans, trans_pcie->cmd_queue,
-				trans_pcie->cmd_fifo);
-
-	/* Activate all Tx DMA/FIFO channels */
-	iwl_trans_txq_set_sched(trans, IWL_MASK(0, 7));
-
-	/* Enable DMA channel */
-	for (chan = 0; chan < FH_TCSR_CHNL_NUM ; chan++)
-		iwl_write_direct32(trans, FH_TCSR_CHNL_TX_CONFIG_REG(chan),
-				   FH_TCSR_TX_CONFIG_REG_VAL_DMA_CHNL_ENABLE |
-				   FH_TCSR_TX_CONFIG_REG_VAL_DMA_CREDIT_ENABLE);
-
-	/* Update FH chicken bits */
-	reg_val = iwl_read_direct32(trans, FH_TX_CHICKEN_BITS_REG);
-	iwl_write_direct32(trans, FH_TX_CHICKEN_BITS_REG,
-			   reg_val | FH_TX_CHICKEN_BITS_SCD_AUTO_RETRY_EN);
-
-	/* Enable L1-Active */
-	iwl_clear_bits_prph(trans, APMG_PCIDEV_STT_REG,
-			    APMG_PCIDEV_STT_VAL_L1_ACT_DIS);
+	return iwl_pcie_load_given_ucode(trans, fw);
 }
 
-static void iwl_trans_pcie_fw_alive(struct iwl_trans *trans)
-{
-	iwl_reset_ict(trans);
-	iwl_tx_start(trans);
-}
-
-/**
- * iwlagn_txq_ctx_stop - Stop all Tx DMA channels
- */
-static int iwl_trans_tx_stop(struct iwl_trans *trans)
+static void iwl_trans_pcie_fw_alive(struct iwl_trans *trans, u32 scd_addr)
 {
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	int ch, txq_id, ret;
-	unsigned long flags;
-
-	/* Turn off all Tx DMA fifos */
-	spin_lock_irqsave(&trans_pcie->irq_lock, flags);
-
-	iwl_trans_txq_set_sched(trans, 0);
-
-	/* Stop each Tx DMA channel, and wait for it to be idle */
-	for (ch = 0; ch < FH_TCSR_CHNL_NUM; ch++) {
-		iwl_write_direct32(trans,
-				   FH_TCSR_CHNL_TX_CONFIG_REG(ch), 0x0);
-		ret = iwl_poll_direct_bit(trans, FH_TSSR_TX_STATUS_REG,
-			FH_TSSR_TX_STATUS_REG_MSK_CHNL_IDLE(ch), 1000);
-		if (ret < 0)
-			IWL_ERR(trans,
-				"Failing on timeout while stopping DMA channel %d [0x%08x]\n",
-				ch,
-				iwl_read_direct32(trans,
-						  FH_TSSR_TX_STATUS_REG));
-	}
-	spin_unlock_irqrestore(&trans_pcie->irq_lock, flags);
-
-	if (!trans_pcie->txq) {
-		IWL_WARN(trans,
-			 "Stopping tx queues that aren't allocated...\n");
-		return 0;
-	}
-
-	/* Unmap DMA from host system and free skb's */
-	for (txq_id = 0; txq_id < trans->cfg->base_params->num_of_queues;
-	     txq_id++)
-		iwl_tx_queue_unmap(trans, txq_id);
-
-	return 0;
+	iwl_pcie_reset_ict(trans);
+	iwl_pcie_tx_start(trans, scd_addr);
 }
 
 static void iwl_trans_pcie_stop_device(struct iwl_trans *trans)
@@ -1196,7 +503,7 @@ static void iwl_trans_pcie_stop_device(struct iwl_trans *trans)
 	spin_unlock_irqrestore(&trans_pcie->irq_lock, flags);
 
 	/* device going down, Stop using ICT table */
-	iwl_disable_ict(trans);
+	iwl_pcie_disable_ict(trans);
 
 	/*
 	 * If a HW restart happens during firmware loading,
@@ -1206,8 +513,8 @@ static void iwl_trans_pcie_stop_device(struct iwl_trans *trans)
 	 * already dead.
 	 */
 	if (test_bit(STATUS_DEVICE_ENABLED, &trans_pcie->status)) {
-		iwl_trans_tx_stop(trans);
-		iwl_trans_rx_stop(trans);
+		iwl_pcie_tx_stop(trans);
+		iwl_pcie_rx_stop(trans);
 
 		/* Power-down device's busmaster DMA clocks */
 		iwl_write_prph(trans, APMG_CLK_DIS_REG,
@@ -1220,7 +527,7 @@ static void iwl_trans_pcie_stop_device(struct iwl_trans *trans)
 		      CSR_GP_CNTRL_REG_FLAG_MAC_ACCESS_REQ);
 
 	/* Stop the device, and put it in low power state */
-	iwl_apm_stop(trans);
+	iwl_pcie_apm_stop(trans);
 
 	/* Upon stop, the APM issues an interrupt if HW RF kill is set.
 	 * Clean again the interrupt here
@@ -1245,6 +552,7 @@ static void iwl_trans_pcie_stop_device(struct iwl_trans *trans)
 	clear_bit(STATUS_INT_ENABLED, &trans_pcie->status);
 	clear_bit(STATUS_DEVICE_ENABLED, &trans_pcie->status);
 	clear_bit(STATUS_TPOWER_PMI, &trans_pcie->status);
+	clear_bit(STATUS_RFKILL, &trans_pcie->status);
 }
 
 static void iwl_trans_pcie_wowlan_suspend(struct iwl_trans *trans)
@@ -1258,169 +566,6 @@ static void iwl_trans_pcie_wowlan_suspend(struct iwl_trans *trans)
 		      CSR_GP_CNTRL_REG_FLAG_MAC_ACCESS_REQ);
 }
 
-static int iwl_trans_pcie_tx(struct iwl_trans *trans, struct sk_buff *skb,
-			     struct iwl_device_cmd *dev_cmd, int txq_id)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data;
-	struct iwl_tx_cmd *tx_cmd = (struct iwl_tx_cmd *) dev_cmd->payload;
-	struct iwl_cmd_meta *out_meta;
-	struct iwl_tx_queue *txq;
-	struct iwl_queue *q;
-	dma_addr_t phys_addr = 0;
-	dma_addr_t txcmd_phys;
-	dma_addr_t scratch_phys;
-	u16 len, firstlen, secondlen;
-	u8 wait_write_ptr = 0;
-	__le16 fc = hdr->frame_control;
-	u8 hdr_len = ieee80211_hdrlen(fc);
-	u16 __maybe_unused wifi_seq;
-
-	txq = &trans_pcie->txq[txq_id];
-	q = &txq->q;
-
-	if (unlikely(!test_bit(txq_id, trans_pcie->queue_used))) {
-		WARN_ON_ONCE(1);
-		return -EINVAL;
-	}
-
-	spin_lock(&txq->lock);
-
-	/* In AGG mode, the index in the ring must correspond to the WiFi
-	 * sequence number. This is a HW requirements to help the SCD to parse
-	 * the BA.
-	 * Check here that the packets are in the right place on the ring.
-	 */
-#ifdef CONFIG_IWLWIFI_DEBUG
-	wifi_seq = SEQ_TO_SN(le16_to_cpu(hdr->seq_ctrl));
-	WARN_ONCE((iwl_read_prph(trans, SCD_AGGR_SEL) & BIT(txq_id)) &&
-		  ((wifi_seq & 0xff) != q->write_ptr),
-		  "Q: %d WiFi Seq %d tfdNum %d",
-		  txq_id, wifi_seq, q->write_ptr);
-#endif
-
-	/* Set up driver data for this TFD */
-	txq->entries[q->write_ptr].skb = skb;
-	txq->entries[q->write_ptr].cmd = dev_cmd;
-
-	dev_cmd->hdr.cmd = REPLY_TX;
-	dev_cmd->hdr.sequence =
-		cpu_to_le16((u16)(QUEUE_TO_SEQ(txq_id) |
-			    INDEX_TO_SEQ(q->write_ptr)));
-
-	/* Set up first empty entry in queue's array of Tx/cmd buffers */
-	out_meta = &txq->entries[q->write_ptr].meta;
-
-	/*
-	 * Use the first empty entry in this queue's command buffer array
-	 * to contain the Tx command and MAC header concatenated together
-	 * (payload data will be in another buffer).
-	 * Size of this varies, due to varying MAC header length.
-	 * If end is not dword aligned, we'll have 2 extra bytes at the end
-	 * of the MAC header (device reads on dword boundaries).
-	 * We'll tell device about this padding later.
-	 */
-	len = sizeof(struct iwl_tx_cmd) +
-		sizeof(struct iwl_cmd_header) + hdr_len;
-	firstlen = (len + 3) & ~3;
-
-	/* Tell NIC about any 2-byte padding after MAC header */
-	if (firstlen != len)
-		tx_cmd->tx_flags |= TX_CMD_FLG_MH_PAD_MSK;
-
-	/* Physical address of this Tx command's header (not MAC header!),
-	 * within command buffer array. */
-	txcmd_phys = dma_map_single(trans->dev,
-				    &dev_cmd->hdr, firstlen,
-				    DMA_BIDIRECTIONAL);
-	if (unlikely(dma_mapping_error(trans->dev, txcmd_phys)))
-		goto out_err;
-	dma_unmap_addr_set(out_meta, mapping, txcmd_phys);
-	dma_unmap_len_set(out_meta, len, firstlen);
-
-	if (!ieee80211_has_morefrags(fc)) {
-		txq->need_update = 1;
-	} else {
-		wait_write_ptr = 1;
-		txq->need_update = 0;
-	}
-
-	/* Set up TFD's 2nd entry to point directly to remainder of skb,
-	 * if any (802.11 null frames have no payload). */
-	secondlen = skb->len - hdr_len;
-	if (secondlen > 0) {
-		phys_addr = dma_map_single(trans->dev, skb->data + hdr_len,
-					   secondlen, DMA_TO_DEVICE);
-		if (unlikely(dma_mapping_error(trans->dev, phys_addr))) {
-			dma_unmap_single(trans->dev,
-					 dma_unmap_addr(out_meta, mapping),
-					 dma_unmap_len(out_meta, len),
-					 DMA_BIDIRECTIONAL);
-			goto out_err;
-		}
-	}
-
-	/* Attach buffers to TFD */
-	iwlagn_txq_attach_buf_to_tfd(trans, txq, txcmd_phys, firstlen, 1);
-	if (secondlen > 0)
-		iwlagn_txq_attach_buf_to_tfd(trans, txq, phys_addr,
-					     secondlen, 0);
-
-	scratch_phys = txcmd_phys + sizeof(struct iwl_cmd_header) +
-				offsetof(struct iwl_tx_cmd, scratch);
-
-	/* take back ownership of DMA buffer to enable update */
-	dma_sync_single_for_cpu(trans->dev, txcmd_phys, firstlen,
-				DMA_BIDIRECTIONAL);
-	tx_cmd->dram_lsb_ptr = cpu_to_le32(scratch_phys);
-	tx_cmd->dram_msb_ptr = iwl_get_dma_hi_addr(scratch_phys);
-
-	IWL_DEBUG_TX(trans, "sequence nr = 0X%x\n",
-		     le16_to_cpu(dev_cmd->hdr.sequence));
-	IWL_DEBUG_TX(trans, "tx_flags = 0X%x\n", le32_to_cpu(tx_cmd->tx_flags));
-
-	/* Set up entry for this TFD in Tx byte-count array */
-	iwl_trans_txq_update_byte_cnt_tbl(trans, txq, le16_to_cpu(tx_cmd->len));
-
-	dma_sync_single_for_device(trans->dev, txcmd_phys, firstlen,
-				   DMA_BIDIRECTIONAL);
-
-	trace_iwlwifi_dev_tx(trans->dev,
-			     &txq->tfds[txq->q.write_ptr],
-			     sizeof(struct iwl_tfd),
-			     &dev_cmd->hdr, firstlen,
-			     skb->data + hdr_len, secondlen);
-
-	/* start timer if queue currently empty */
-	if (txq->need_update && q->read_ptr == q->write_ptr &&
-	    trans_pcie->wd_timeout)
-		mod_timer(&txq->stuck_timer, jiffies + trans_pcie->wd_timeout);
-
-	/* Tell device the write index *just past* this latest filled TFD */
-	q->write_ptr = iwl_queue_inc_wrap(q->write_ptr, q->n_bd);
-	iwl_txq_update_write_ptr(trans, txq);
-
-	/*
-	 * At this point the frame is "transmitted" successfully
-	 * and we will get a TX status notification eventually,
-	 * regardless of the value of ret. "ret" only indicates
-	 * whether or not we should update the write pointer.
-	 */
-	if (iwl_queue_space(q) < q->high_mark) {
-		if (wait_write_ptr) {
-			txq->need_update = 1;
-			iwl_txq_update_write_ptr(trans, txq);
-		} else {
-			iwl_stop_queue(trans, txq);
-		}
-	}
-	spin_unlock(&txq->lock);
-	return 0;
- out_err:
-	spin_unlock(&txq->lock);
-	return -1;
-}
-
 static int iwl_trans_pcie_start_hw(struct iwl_trans *trans)
 {
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
@@ -1431,29 +576,28 @@ static int iwl_trans_pcie_start_hw(struct iwl_trans *trans)
 
 	if (!trans_pcie->irq_requested) {
 		tasklet_init(&trans_pcie->irq_tasklet, (void (*)(unsigned long))
-			iwl_irq_tasklet, (unsigned long)trans);
+			iwl_pcie_tasklet, (unsigned long)trans);
 
-		iwl_alloc_isr_ict(trans);
+		iwl_pcie_alloc_ict(trans);
 
-		err = request_irq(trans_pcie->irq, iwl_isr_ict, IRQF_SHARED,
-				  DRV_NAME, trans);
+		err = request_irq(trans_pcie->irq, iwl_pcie_isr_ict,
+				  IRQF_SHARED, DRV_NAME, trans);
 		if (err) {
 			IWL_ERR(trans, "Error allocating IRQ %d\n",
 				trans_pcie->irq);
 			goto error;
 		}
 
-		INIT_WORK(&trans_pcie->rx_replenish, iwl_bg_rx_replenish);
 		trans_pcie->irq_requested = true;
 	}
 
-	err = iwl_prepare_card_hw(trans);
+	err = iwl_pcie_prepare_card_hw(trans);
 	if (err) {
 		IWL_ERR(trans, "Error while preparing HW: %d\n", err);
 		goto err_free_irq;
 	}
 
-	iwl_apm_init(trans);
+	iwl_pcie_apm_init(trans);
 
 	/* From now on, the op_mode will be kept updated about RF kill state */
 	iwl_enable_rfkill_int(trans);
@@ -1467,7 +611,7 @@ err_free_irq:
 	trans_pcie->irq_requested = false;
 	free_irq(trans_pcie->irq, trans);
 error:
-	iwl_free_isr_ict(trans);
+	iwl_pcie_free_ict(trans);
 	tasklet_kill(&trans_pcie->irq_tasklet);
 	return err;
 }
@@ -1483,12 +627,14 @@ static void iwl_trans_pcie_stop_hw(struct iwl_trans *trans,
 	iwl_disable_interrupts(trans);
 	spin_unlock_irqrestore(&trans_pcie->irq_lock, flags);
 
-	iwl_apm_stop(trans);
+	iwl_pcie_apm_stop(trans);
 
 	spin_lock_irqsave(&trans_pcie->irq_lock, flags);
 	iwl_disable_interrupts(trans);
 	spin_unlock_irqrestore(&trans_pcie->irq_lock, flags);
 
+	iwl_pcie_disable_ict(trans);
+
 	if (!op_mode_leaving) {
 		/*
 		 * Even if we stop the HW, we still want the RF kill
@@ -1507,28 +653,6 @@ static void iwl_trans_pcie_stop_hw(struct iwl_trans *trans,
 	}
 }
 
-static void iwl_trans_pcie_reclaim(struct iwl_trans *trans, int txq_id, int ssn,
-				   struct sk_buff_head *skbs)
-{
-	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_tx_queue *txq = &trans_pcie->txq[txq_id];
-	/* n_bd is usually 256 => n_bd - 1 = 0xff */
-	int tfd_num = ssn & (txq->q.n_bd - 1);
-	int freed = 0;
-
-	spin_lock(&txq->lock);
-
-	if (txq->q.read_ptr != tfd_num) {
-		IWL_DEBUG_TX_REPLY(trans, "[Q %d] %d -> %d (%d)\n",
-				   txq_id, txq->q.read_ptr, tfd_num, ssn);
-		freed = iwl_tx_queue_reclaim(trans, txq_id, tfd_num, skbs);
-		if (iwl_queue_space(&txq->q) > txq->q.low_mark)
-			iwl_wake_queue(trans, txq);
-	}
-
-	spin_unlock(&txq->lock);
-}
-
 static void iwl_trans_pcie_write8(struct iwl_trans *trans, u32 ofs, u8 val)
 {
 	writeb(val, IWL_TRANS_GET_PCIE_TRANS(trans)->hw_base + ofs);
@@ -1544,6 +668,20 @@ static u32 iwl_trans_pcie_read32(struct iwl_trans *trans, u32 ofs)
 	return readl(IWL_TRANS_GET_PCIE_TRANS(trans)->hw_base + ofs);
 }
 
+static u32 iwl_trans_pcie_read_prph(struct iwl_trans *trans, u32 reg)
+{
+	iwl_trans_pcie_write32(trans, HBUS_TARG_PRPH_RADDR, reg | (3 << 24));
+	return iwl_trans_pcie_read32(trans, HBUS_TARG_PRPH_RDAT);
+}
+
+static void iwl_trans_pcie_write_prph(struct iwl_trans *trans, u32 addr,
+				      u32 val)
+{
+	iwl_trans_pcie_write32(trans, HBUS_TARG_PRPH_WADDR,
+			       ((addr & 0x0000FFFF) | (3 << 24)));
+	iwl_trans_pcie_write32(trans, HBUS_TARG_PRPH_WDAT, val);
+}
+
 static void iwl_trans_pcie_configure(struct iwl_trans *trans,
 				     const struct iwl_trans_config *trans_cfg)
 {
@@ -1575,12 +713,12 @@ void iwl_trans_pcie_free(struct iwl_trans *trans)
 {
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
 
-	iwl_trans_pcie_tx_free(trans);
-	iwl_trans_pcie_rx_free(trans);
+	iwl_pcie_tx_free(trans);
+	iwl_pcie_rx_free(trans);
 
 	if (trans_pcie->irq_requested == true) {
 		free_irq(trans_pcie->irq, trans);
-		iwl_free_isr_ict(trans);
+		iwl_pcie_free_ict(trans);
 	}
 
 	pci_disable_msi(trans_pcie->pci_dev);
@@ -1626,10 +764,10 @@ static int iwl_trans_pcie_resume(struct iwl_trans *trans)
 
 #define IWL_FLUSH_WAIT_MS	2000
 
-static int iwl_trans_pcie_wait_tx_queue_empty(struct iwl_trans *trans)
+static int iwl_trans_pcie_wait_txq_empty(struct iwl_trans *trans)
 {
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_tx_queue *txq;
+	struct iwl_txq *txq;
 	struct iwl_queue *q;
 	int cnt;
 	unsigned long now = jiffies;
@@ -1673,7 +811,7 @@ static const char *get_fh_string(int cmd)
 #undef IWL_CMD
 }
 
-int iwl_dump_fh(struct iwl_trans *trans, char **buf)
+int iwl_pcie_dump_fh(struct iwl_trans *trans, char **buf)
 {
 	int i;
 	static const u32 fh_tbl[] = {
@@ -1752,7 +890,7 @@ static const char *get_csr_string(int cmd)
 #undef IWL_CMD
 }
 
-void iwl_dump_csr(struct iwl_trans *trans)
+void iwl_pcie_dump_csr(struct iwl_trans *trans)
 {
 	int i;
 	static const u32 csr_tbl[] = {
@@ -1809,7 +947,6 @@ static ssize_t iwl_dbgfs_##name##_write(struct file *file,              \
 					const char __user *user_buf,    \
 					size_t count, loff_t *ppos);
 
-
 #define DEBUGFS_READ_FILE_OPS(name)					\
 	DEBUGFS_READ_FUNC(name);					\
 static const struct file_operations iwl_dbgfs_##name##_ops = {		\
@@ -1842,7 +979,7 @@ static ssize_t iwl_dbgfs_tx_queue_read(struct file *file,
 {
 	struct iwl_trans *trans = file->private_data;
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_tx_queue *txq;
+	struct iwl_txq *txq;
 	struct iwl_queue *q;
 	char *buf;
 	int pos = 0;
@@ -1879,7 +1016,7 @@ static ssize_t iwl_dbgfs_rx_queue_read(struct file *file,
 {
 	struct iwl_trans *trans = file->private_data;
 	struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
-	struct iwl_rx_queue *rxq = &trans_pcie->rxq;
+	struct iwl_rxq *rxq = &trans_pcie->rxq;
 	char buf[256];
 	int pos = 0;
 	const size_t bufsz = sizeof(buf);
@@ -1998,7 +1135,7 @@ static ssize_t iwl_dbgfs_csr_write(struct file *file,
 	if (sscanf(buf, "%d", &csr) != 1)
 		return -EFAULT;
 
-	iwl_dump_csr(trans);
+	iwl_pcie_dump_csr(trans);
 
 	return count;
 }
@@ -2012,7 +1149,7 @@ static ssize_t iwl_dbgfs_fh_reg_read(struct file *file,
 	int pos = 0;
 	ssize_t ret = -EFAULT;
 
-	ret = pos = iwl_dump_fh(trans, &buf);
+	ret = pos = iwl_pcie_dump_fh(trans, &buf);
 	if (buf) {
 		ret = simple_read_from_buffer(user_buf,
 					      count, ppos, buf, pos);
@@ -2081,7 +1218,7 @@ static const struct iwl_trans_ops trans_ops_pcie = {
 
 	.wowlan_suspend = iwl_trans_pcie_wowlan_suspend,
 
-	.send_cmd = iwl_trans_pcie_send_cmd,
+	.send_cmd = iwl_trans_pcie_send_hcmd,
 
 	.tx = iwl_trans_pcie_tx,
 	.reclaim = iwl_trans_pcie_reclaim,
@@ -2091,7 +1228,7 @@ static const struct iwl_trans_ops trans_ops_pcie = {
 
 	.dbgfs_register = iwl_trans_pcie_dbgfs_register,
 
-	.wait_tx_queue_empty = iwl_trans_pcie_wait_tx_queue_empty,
+	.wait_tx_queue_empty = iwl_trans_pcie_wait_txq_empty,
 
 #ifdef CONFIG_PM_SLEEP
 	.suspend = iwl_trans_pcie_suspend,
@@ -2100,6 +1237,8 @@ static const struct iwl_trans_ops trans_ops_pcie = {
 	.write8 = iwl_trans_pcie_write8,
 	.write32 = iwl_trans_pcie_write32,
 	.read32 = iwl_trans_pcie_read32,
+	.read_prph = iwl_trans_pcie_read_prph,
+	.write_prph = iwl_trans_pcie_write_prph,
 	.configure = iwl_trans_pcie_configure,
 	.set_pmi = iwl_trans_pcie_set_pmi,
 };
@@ -2116,7 +1255,7 @@ struct iwl_trans *iwl_trans_pcie_alloc(struct pci_dev *pdev,
 	trans = kzalloc(sizeof(struct iwl_trans) +
 			sizeof(struct iwl_trans_pcie), GFP_KERNEL);
 
-	if (WARN_ON(!trans))
+	if (!trans)
 		return NULL;
 
 	trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
@@ -2149,43 +1288,38 @@ struct iwl_trans *iwl_trans_pcie_alloc(struct pci_dev *pdev,
 							  DMA_BIT_MASK(32));
 		/* both attempts failed: */
 		if (err) {
-			dev_printk(KERN_ERR, &pdev->dev,
-				   "No suitable DMA available.\n");
+			dev_err(&pdev->dev, "No suitable DMA available\n");
 			goto out_pci_disable_device;
 		}
 	}
 
 	err = pci_request_regions(pdev, DRV_NAME);
 	if (err) {
-		dev_printk(KERN_ERR, &pdev->dev,
-			   "pci_request_regions failed\n");
+		dev_err(&pdev->dev, "pci_request_regions failed\n");
 		goto out_pci_disable_device;
 	}
 
 	trans_pcie->hw_base = pci_ioremap_bar(pdev, 0);
 	if (!trans_pcie->hw_base) {
-		dev_printk(KERN_ERR, &pdev->dev, "pci_ioremap_bar failed\n");
+		dev_err(&pdev->dev, "pci_ioremap_bar failed\n");
 		err = -ENODEV;
 		goto out_pci_release_regions;
 	}
 
-	dev_printk(KERN_INFO, &pdev->dev,
-		   "pci_resource_len = 0x%08llx\n",
-		   (unsigned long long) pci_resource_len(pdev, 0));
-	dev_printk(KERN_INFO, &pdev->dev,
-		   "pci_resource_base = %p\n", trans_pcie->hw_base);
-
-	dev_printk(KERN_INFO, &pdev->dev,
-		   "HW Revision ID = 0x%X\n", pdev->revision);
-
 	/* We disable the RETRY_TIMEOUT register (0x41) to keep
 	 * PCI Tx retries from interfering with C3 CPU state */
 	pci_write_config_byte(pdev, PCI_CFG_RETRY_TIMEOUT, 0x00);
 
 	err = pci_enable_msi(pdev);
-	if (err)
-		dev_printk(KERN_ERR, &pdev->dev,
-			   "pci_enable_msi failed(0X%x)\n", err);
+	if (err) {
+		dev_err(&pdev->dev, "pci_enable_msi failed(0X%x)\n", err);
+		/* enable rfkill interrupt: hw bug w/a */
+		pci_read_config_word(pdev, PCI_COMMAND, &pci_cmd);
+		if (pci_cmd & PCI_COMMAND_INTX_DISABLE) {
+			pci_cmd &= ~PCI_COMMAND_INTX_DISABLE;
+			pci_write_config_word(pdev, PCI_COMMAND, pci_cmd);
+		}
+	}
 
 	trans->dev = &pdev->dev;
 	trans_pcie->irq = pdev->irq;
@@ -2195,16 +1329,8 @@ struct iwl_trans *iwl_trans_pcie_alloc(struct pci_dev *pdev,
 	snprintf(trans->hw_id_str, sizeof(trans->hw_id_str),
 		 "PCI ID: 0x%04X:0x%04X", pdev->device, pdev->subsystem_device);
 
-	/* TODO: Move this away, not needed if not MSI */
-	/* enable rfkill interrupt: hw bug w/a */
-	pci_read_config_word(pdev, PCI_COMMAND, &pci_cmd);
-	if (pci_cmd & PCI_COMMAND_INTX_DISABLE) {
-		pci_cmd &= ~PCI_COMMAND_INTX_DISABLE;
-		pci_write_config_word(pdev, PCI_COMMAND, pci_cmd);
-	}
-
 	/* Initialize the wait queue for commands */
-	init_waitqueue_head(&trans->wait_command_queue);
+	init_waitqueue_head(&trans_pcie->wait_command_queue);
 	spin_lock_init(&trans->reg_lock);
 
 	snprintf(trans->dev_cmd_pool_name, sizeof(trans->dev_cmd_pool_name),