iwlwifi: pcie: add infrastructure for multi-queue rx
The 9000 series devices will support multi rx queues.
Current code has one static rx queue - change it to allocate
a number of queues per the device capability (pre-9000 devices
have the number of rx queues set to one).
Subsequent generalizations are:
Change the code to access an explicit numbered rx queue only
when the queue number is known - when handling interrupt, when
accessing the default queue and when iterating the queues.
The rest of the functions will receive the rx queue as a pointer.
Generalize the warning in allocation failure to consider the
allocator status instead of a single rx queue status.
Move the rx initial pool of memory buffers to be shared among
all the queues and allocated to the default queue on init.
Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
diff --git a/drivers/net/wireless/intel/iwlwifi/pcie/rx.c b/drivers/net/wireless/intel/iwlwifi/pcie/rx.c
index ccafbd8..f557f3d 100644
--- a/drivers/net/wireless/intel/iwlwifi/pcie/rx.c
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/rx.c
@@ -173,10 +173,9 @@
/*
* iwl_pcie_rxq_inc_wr_ptr - Update the write pointer for the RX queue
*/
-static void iwl_pcie_rxq_inc_wr_ptr(struct iwl_trans *trans)
+static void iwl_pcie_rxq_inc_wr_ptr(struct iwl_trans *trans,
+ struct iwl_rxq *rxq)
{
- struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
u32 reg;
lockdep_assert_held(&rxq->lock);
@@ -207,18 +206,18 @@
static void iwl_pcie_rxq_check_wrptr(struct iwl_trans *trans)
{
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
+ int i;
- spin_lock(&rxq->lock);
+ for (i = 0; i < trans->num_rx_queues; i++) {
+ struct iwl_rxq *rxq = &trans_pcie->rxq[i];
- if (!rxq->need_update)
- goto exit_unlock;
-
- iwl_pcie_rxq_inc_wr_ptr(trans);
- rxq->need_update = false;
-
- exit_unlock:
- spin_unlock(&rxq->lock);
+ if (!rxq->need_update)
+ continue;
+ spin_lock(&rxq->lock);
+ iwl_pcie_rxq_inc_wr_ptr(trans, rxq);
+ rxq->need_update = false;
+ spin_unlock(&rxq->lock);
+ }
}
/*
@@ -232,10 +231,8 @@
* also updates the memory address in the firmware to reference the new
* target buffer.
*/
-static void iwl_pcie_rxq_restock(struct iwl_trans *trans)
+static void iwl_pcie_rxq_restock(struct iwl_trans *trans, struct iwl_rxq *rxq)
{
- struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
struct iwl_rx_mem_buffer *rxb;
/*
@@ -272,7 +269,7 @@
* Increment device's write pointer in multiples of 8. */
if (rxq->write_actual != (rxq->write & ~0x7)) {
spin_lock(&rxq->lock);
- iwl_pcie_rxq_inc_wr_ptr(trans);
+ iwl_pcie_rxq_inc_wr_ptr(trans, rxq);
spin_unlock(&rxq->lock);
}
}
@@ -285,13 +282,9 @@
gfp_t priority)
{
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
struct page *page;
gfp_t gfp_mask = priority;
- if (rxq->free_count > RX_LOW_WATERMARK)
- gfp_mask |= __GFP_NOWARN;
-
if (trans_pcie->rx_page_order > 0)
gfp_mask |= __GFP_COMP;
@@ -301,16 +294,13 @@
if (net_ratelimit())
IWL_DEBUG_INFO(trans, "alloc_pages failed, order: %d\n",
trans_pcie->rx_page_order);
- /* Issue an error if the hardware has consumed more than half
- * of its free buffer list and we don't have enough
- * pre-allocated buffers.
+ /*
+ * Issue an error if we don't have enough pre-allocated
+ * buffers.
` */
- if (rxq->free_count <= RX_LOW_WATERMARK &&
- iwl_rxq_space(rxq) > (RX_QUEUE_SIZE / 2) &&
- net_ratelimit())
+ if (!(gfp_mask & __GFP_NOWARN) && net_ratelimit())
IWL_CRIT(trans,
- "Failed to alloc_pages with GFP_KERNEL. Only %u free buffers remaining.\n",
- rxq->free_count);
+ "Failed to alloc_pages\n");
return NULL;
}
return page;
@@ -325,10 +315,10 @@
* iwl_pcie_rxq_restock. The latter function will update the HW to use the newly
* allocated buffers.
*/
-static void iwl_pcie_rxq_alloc_rbs(struct iwl_trans *trans, gfp_t priority)
+static void iwl_pcie_rxq_alloc_rbs(struct iwl_trans *trans, gfp_t priority,
+ struct iwl_rxq *rxq)
{
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
struct iwl_rx_mem_buffer *rxb;
struct page *page;
@@ -386,41 +376,24 @@
}
}
-static void iwl_pcie_rxq_free_rbs(struct iwl_trans *trans)
+static void iwl_pcie_free_rbs_pool(struct iwl_trans *trans)
{
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
int i;
- lockdep_assert_held(&rxq->lock);
-
for (i = 0; i < RX_QUEUE_SIZE; i++) {
- if (!rxq->pool[i].page)
+ if (!trans_pcie->rx_pool[i].page)
continue;
- dma_unmap_page(trans->dev, rxq->pool[i].page_dma,
+ dma_unmap_page(trans->dev, trans_pcie->rx_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;
+ __free_pages(trans_pcie->rx_pool[i].page,
+ trans_pcie->rx_page_order);
+ trans_pcie->rx_pool[i].page = NULL;
}
}
/*
- * iwl_pcie_rx_replenish - Move all used buffers from rx_used to rx_free
- *
- * When moving to rx_free an page is allocated for the slot.
- *
- * Also restock the Rx queue via iwl_pcie_rxq_restock.
- * This is called only during initialization
- */
-static void iwl_pcie_rx_replenish(struct iwl_trans *trans)
-{
- iwl_pcie_rxq_alloc_rbs(trans, GFP_KERNEL);
-
- iwl_pcie_rxq_restock(trans);
-}
-
-/*
* iwl_pcie_rx_allocator - Allocates pages in the background for RX queues
*
* Allocates for each received request 8 pages
@@ -444,6 +417,11 @@
while (pending) {
int i;
struct list_head local_allocated;
+ gfp_t gfp_mask = GFP_KERNEL;
+
+ /* Do not post a warning if there are only a few requests */
+ if (pending < RX_PENDING_WATERMARK)
+ gfp_mask |= __GFP_NOWARN;
INIT_LIST_HEAD(&local_allocated);
@@ -463,7 +441,7 @@
BUG_ON(rxb->page);
/* Alloc a new receive buffer */
- page = iwl_pcie_rx_alloc_page(trans, GFP_KERNEL);
+ page = iwl_pcie_rx_alloc_page(trans, gfp_mask);
if (!page)
continue;
rxb->page = page;
@@ -561,38 +539,60 @@
static int iwl_pcie_rx_alloc(struct iwl_trans *trans)
{
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
struct iwl_rb_allocator *rba = &trans_pcie->rba;
struct device *dev = trans->dev;
+ int i;
- memset(&trans_pcie->rxq, 0, sizeof(trans_pcie->rxq));
-
- spin_lock_init(&rxq->lock);
- spin_lock_init(&rba->lock);
-
- if (WARN_ON(rxq->bd || rxq->rb_stts))
+ if (WARN_ON(trans_pcie->rxq))
return -EINVAL;
- /* Allocate the circular buffer of Read Buffer Descriptors (RBDs) */
- rxq->bd = dma_zalloc_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE,
+ trans_pcie->rxq = kcalloc(trans->num_rx_queues, sizeof(struct iwl_rxq),
+ GFP_KERNEL);
+ if (!trans_pcie->rxq)
+ return -EINVAL;
+
+ spin_lock_init(&rba->lock);
+
+ for (i = 0; i < trans->num_rx_queues; i++) {
+ struct iwl_rxq *rxq = &trans_pcie->rxq[i];
+
+ spin_lock_init(&rxq->lock);
+ /*
+ * 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;
+ if (!rxq->bd)
+ goto err;
- /*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;
+ /*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;
+ }
return 0;
-err_rb_stts:
- dma_free_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE,
- rxq->bd, rxq->bd_dma);
- rxq->bd_dma = 0;
- rxq->bd = NULL;
-err_bd:
+err:
+ for (i = 0; i < trans->num_rx_queues; i++) {
+ struct iwl_rxq *rxq = &trans_pcie->rxq[i];
+
+ if (rxq->bd)
+ dma_free_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE,
+ rxq->bd, rxq->bd_dma);
+ rxq->bd_dma = 0;
+ rxq->bd = NULL;
+
+ if (rxq->rb_stts)
+ dma_free_coherent(trans->dev,
+ sizeof(struct iwl_rb_status),
+ rxq->rb_stts, rxq->rb_stts_dma);
+ }
+ kfree(trans_pcie->rxq);
return -ENOMEM;
}
@@ -661,17 +661,12 @@
static void iwl_pcie_rx_init_rxb_lists(struct iwl_rxq *rxq)
{
- int i;
-
lockdep_assert_held(&rxq->lock);
INIT_LIST_HEAD(&rxq->rx_free);
INIT_LIST_HEAD(&rxq->rx_used);
rxq->free_count = 0;
rxq->used_count = 0;
-
- for (i = 0; i < RX_QUEUE_SIZE; i++)
- list_add(&rxq->pool[i].list, &rxq->rx_used);
}
static void iwl_pcie_rx_init_rba(struct iwl_rb_allocator *rba)
@@ -709,15 +704,16 @@
int iwl_pcie_rx_init(struct iwl_trans *trans)
{
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
+ struct iwl_rxq *def_rxq;
struct iwl_rb_allocator *rba = &trans_pcie->rba;
int i, err;
- if (!rxq->bd) {
+ if (!trans_pcie->rxq) {
err = iwl_pcie_rx_alloc(trans);
if (err)
return err;
}
+ def_rxq = trans_pcie->rxq;
if (!rba->alloc_wq)
rba->alloc_wq = alloc_workqueue("rb_allocator",
WQ_HIGHPRI | WQ_UNBOUND, 1);
@@ -731,29 +727,42 @@
iwl_pcie_rx_init_rba(rba);
spin_unlock(&rba->lock);
- spin_lock(&rxq->lock);
-
/* free all first - we might be reconfigured for a different size */
- iwl_pcie_rxq_free_rbs(trans);
- iwl_pcie_rx_init_rxb_lists(rxq);
+ iwl_pcie_free_rbs_pool(trans);
for (i = 0; i < RX_QUEUE_SIZE; i++)
- rxq->queue[i] = NULL;
+ def_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;
- memset(rxq->rb_stts, 0, sizeof(*rxq->rb_stts));
- spin_unlock(&rxq->lock);
+ for (i = 0; i < trans->num_rx_queues; i++) {
+ struct iwl_rxq *rxq = &trans_pcie->rxq[i];
- iwl_pcie_rx_replenish(trans);
+ spin_lock(&rxq->lock);
+ /*
+ * Set read write pointer to reflect that we have processed
+ * and used all buffers, but have not restocked the Rx queue
+ * with fresh buffers
+ */
+ rxq->read = 0;
+ rxq->write = 0;
+ rxq->write_actual = 0;
+ memset(rxq->rb_stts, 0, sizeof(*rxq->rb_stts));
- iwl_pcie_rx_hw_init(trans, rxq);
+ iwl_pcie_rx_init_rxb_lists(rxq);
- spin_lock(&rxq->lock);
- iwl_pcie_rxq_inc_wr_ptr(trans);
- spin_unlock(&rxq->lock);
+ spin_unlock(&rxq->lock);
+ }
+
+ /* move the entire pool to the default queue ownership */
+ for (i = 0; i < RX_QUEUE_SIZE; i++)
+ list_add(&trans_pcie->rx_pool[i].list, &def_rxq->rx_used);
+
+ iwl_pcie_rxq_alloc_rbs(trans, GFP_KERNEL, def_rxq);
+ iwl_pcie_rxq_restock(trans, def_rxq);
+ iwl_pcie_rx_hw_init(trans, def_rxq);
+
+ spin_lock(&def_rxq->lock);
+ iwl_pcie_rxq_inc_wr_ptr(trans, def_rxq);
+ spin_unlock(&def_rxq->lock);
return 0;
}
@@ -761,12 +770,14 @@
void iwl_pcie_rx_free(struct iwl_trans *trans)
{
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
struct iwl_rb_allocator *rba = &trans_pcie->rba;
+ int i;
- /*if rxq->bd is NULL, it means that nothing has been allocated,
- * exit now */
- if (!rxq->bd) {
+ /*
+ * if rxq is NULL, it means that nothing has been allocated,
+ * exit now
+ */
+ if (!trans_pcie->rxq) {
IWL_DEBUG_INFO(trans, "Free NULL rx context\n");
return;
}
@@ -781,23 +792,28 @@
iwl_pcie_rx_free_rba(trans);
spin_unlock(&rba->lock);
- spin_lock(&rxq->lock);
- iwl_pcie_rxq_free_rbs(trans);
- spin_unlock(&rxq->lock);
+ iwl_pcie_free_rbs_pool(trans);
- dma_free_coherent(trans->dev, sizeof(__le32) * RX_QUEUE_SIZE,
- rxq->bd, rxq->bd_dma);
- rxq->bd_dma = 0;
- rxq->bd = NULL;
+ for (i = 0; i < trans->num_rx_queues; i++) {
+ struct iwl_rxq *rxq = &trans_pcie->rxq[i];
- 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");
- rxq->rb_stts_dma = 0;
- rxq->rb_stts = NULL;
+ if (rxq->bd)
+ dma_free_coherent(trans->dev,
+ sizeof(__le32) * RX_QUEUE_SIZE,
+ rxq->bd, rxq->bd_dma);
+ rxq->bd_dma = 0;
+ 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");
+ }
+
+ kfree(trans_pcie->rxq);
}
/*
@@ -841,11 +857,11 @@
}
static void iwl_pcie_rx_handle_rb(struct iwl_trans *trans,
+ struct iwl_rxq *rxq,
struct iwl_rx_mem_buffer *rxb,
bool emergency)
{
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
struct iwl_txq *txq = &trans_pcie->txq[trans_pcie->cmd_queue];
bool page_stolen = false;
int max_len = PAGE_SIZE << trans_pcie->rx_page_order;
@@ -975,7 +991,7 @@
static void iwl_pcie_rx_handle(struct iwl_trans *trans)
{
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
- struct iwl_rxq *rxq = &trans_pcie->rxq;
+ struct iwl_rxq *rxq = &trans_pcie->rxq[0];
u32 r, i, j, count = 0;
bool emergency = false;
@@ -1000,7 +1016,7 @@
rxq->queue[i] = NULL;
IWL_DEBUG_RX(trans, "rxbuf: HW = %d, SW = %d\n", r, i);
- iwl_pcie_rx_handle_rb(trans, rxb, emergency);
+ iwl_pcie_rx_handle_rb(trans, rxq, rxb, emergency);
i = (i + 1) & RX_QUEUE_MASK;
@@ -1043,7 +1059,7 @@
if (rxq->used_count < RX_QUEUE_SIZE / 3)
emergency = false;
spin_unlock(&rxq->lock);
- iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC);
+ iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC, rxq);
spin_lock(&rxq->lock);
}
}
@@ -1055,7 +1071,7 @@
if (rxq->free_count >= RX_CLAIM_REQ_ALLOC) {
rxq->read = i;
spin_unlock(&rxq->lock);
- iwl_pcie_rxq_restock(trans);
+ iwl_pcie_rxq_restock(trans, rxq);
goto restart;
}
}
@@ -1077,7 +1093,7 @@
* will be restocked by the next call of iwl_pcie_rxq_restock.
*/
if (unlikely(emergency && count))
- iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC);
+ iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC, rxq);
if (trans_pcie->napi.poll)
napi_gro_flush(&trans_pcie->napi, false);