iwlwifi: pcie: New RBD allocation model

As a preperation for multiple RX queues change the RBD
allocation model.

The new model includes a background allocator. The allocator is
called by the interrupt handler when there are two released
buffers by the queue, and the allocator starts allocating eight
pages per request.
When the queue has released 8 pages it tries claiming the
request. If the pages are not ready - it keeps claiming.
This new model should make sure that RBDs are always available
across the multiple queues.

The RBDs are transferred between the allocator and the queue.
The queue moves the free RBDs upon freeing them to the allocator.
The allocator moves them back to the queue's possession when the
request is claimed.
The allocator has an initial pool to make sure there are always RBDs
available for the request completion.
Release of the buffers at exit is done per pools - the allocator
frees its own initial pool and the queue frees its own pool.

Existing code refactor -
-Queue's initial pool is the size of the queue only as the allocation
of the new buffers no longer uses this pool.
-Removal of replenish background work, and replenish calls in the
interrupt handler and restock().
-The replenish() and the rxq used_list are used only during
initialization.
-Moved page allocation to a new function for code reuse.

New code -
Allocator code - new structure and functions.
Interrupt handler uses the allocator functions for replenishing buffers.
Reuse of the restock() method.
Signed-off-by: NSara Sharon <sara.sharon@intel.com>
Signed-off-by: NEmmanuel Grumbach <emmanuel.grumbach@intel.com>

drivers/net/wireless/iwlwifi/iwl-fh.h

@@ -438,12 +438,6 @@ static inline unsigned int FH_MEM_CBBC_QUEUE(unsigned int chnl)
 #define RX_QUEUE_MASK                         255
 #define RX_QUEUE_SIZE_LOG                     8
 
-/*
- * RX related structures and functions
- */
-#define RX_FREE_BUFFERS 64
-#define RX_LOW_WATERMARK 8
-
 /**
  * struct iwl_rb_status - reserve buffer status
  * 	host memory mapped FH registers

drivers/net/wireless/iwlwifi/pcie/internal.h

@@ -50,6 +50,15 @@
  */
 #define IWL_PCIE_MAX_FRAGS (IWL_NUM_OF_TBS - 3)
 
+/*
+ * RX related structures and functions
+ */
+#define RX_NUM_QUEUES 1
+#define RX_POST_REQ_ALLOC 2
+#define RX_CLAIM_REQ_ALLOC 8
+#define RX_POOL_SIZE ((RX_CLAIM_REQ_ALLOC - RX_POST_REQ_ALLOC) * RX_NUM_QUEUES)
+#define RX_LOW_WATERMARK 8
+
 struct iwl_host_cmd;
 
 /*This file includes the declaration that are internal to the
@@ -83,29 +92,29 @@ struct isr_statistics {
  * struct iwl_rxq - Rx queue
  * @bd: driver's pointer to buffer of receive buffer descriptors (rbd)
  * @bd_dma: bus address of buffer of receive buffer descriptors (rbd)
- * @pool:
- * @queue:
  * @read: Shared index to newest available Rx buffer
  * @write: Shared index to oldest written Rx packet
  * @free_count: Number of pre-allocated buffers in rx_free
+ * @used_count: Number of RBDs handled to allocator to use for allocation
  * @write_actual:
- * @rx_free: list of free SKBs for use
- * @rx_used: List of Rx buffers with no SKB
+ * @rx_free: list of RBDs with allocated RB ready for use
+ * @rx_used: list of RBDs with no RB attached
  * @need_update: flag to indicate we need to update read/write index
  * @rb_stts: driver's pointer to receive buffer status
  * @rb_stts_dma: bus address of receive buffer status
  * @lock:
+ * @pool: initial pool of iwl_rx_mem_buffer for the queue
+ * @queue: actual rx queue
  *
  * NOTE:  rx_free and rx_used are used as a FIFO for iwl_rx_mem_buffers
  */
 struct iwl_rxq {
 	__le32 *bd;
 	dma_addr_t bd_dma;
-	struct iwl_rx_mem_buffer pool[RX_QUEUE_SIZE + RX_FREE_BUFFERS];
-	struct iwl_rx_mem_buffer *queue[RX_QUEUE_SIZE];
 	u32 read;
 	u32 write;
 	u32 free_count;
+	u32 used_count;
 	u32 write_actual;
 	struct list_head rx_free;
 	struct list_head rx_used;
@@ -113,6 +122,32 @@ struct iwl_rxq {
 	struct iwl_rb_status *rb_stts;
 	dma_addr_t rb_stts_dma;
 	spinlock_t lock;
+	struct iwl_rx_mem_buffer pool[RX_QUEUE_SIZE];
+	struct iwl_rx_mem_buffer *queue[RX_QUEUE_SIZE];
+};
+
+/**
+ * struct iwl_rb_allocator - Rx allocator
+ * @pool: initial pool of allocator
+ * @req_pending: number of requests the allcator had not processed yet
+ * @req_ready: number of requests honored and ready for claiming
+ * @rbd_allocated: RBDs with pages allocated and ready to be handled to
+ *	the queue. This is a list of &struct iwl_rx_mem_buffer
+ * @rbd_empty: RBDs with no page attached for allocator use. This is a list
+ *	of &struct iwl_rx_mem_buffer
+ * @lock: protects the rbd_allocated and rbd_empty lists
+ * @alloc_wq: work queue for background calls
+ * @rx_alloc: work struct for background calls
+ */
+struct iwl_rb_allocator {
+	struct iwl_rx_mem_buffer pool[RX_POOL_SIZE];
+	atomic_t req_pending;
+	atomic_t req_ready;
+	struct list_head rbd_allocated;
+	struct list_head rbd_empty;
+	spinlock_t lock;
+	struct workqueue_struct *alloc_wq;
+	struct work_struct rx_alloc;
 };
 
 struct iwl_dma_ptr {
@@ -256,7 +291,7 @@ iwl_pcie_get_scratchbuf_dma(struct iwl_txq *txq, int idx)
 /**
  * struct iwl_trans_pcie - PCIe transport specific data
  * @rxq: all the RX queue data
- * @rx_replenish: work that will be called when buffers need to be allocated
+ * @rba: allocator for RX replenishing
  * @drv - pointer to iwl_drv
  * @trans: pointer to the generic transport area
  * @scd_base_addr: scheduler sram base address in SRAM
@@ -281,7 +316,7 @@ iwl_pcie_get_scratchbuf_dma(struct iwl_txq *txq, int idx)
  */
 struct iwl_trans_pcie {
 	struct iwl_rxq rxq;
-	struct work_struct rx_replenish;
+	struct iwl_rb_allocator rba;
 	struct iwl_trans *trans;
 	struct iwl_drv *drv;