sfc: Use pci_map_single() to map the skb header when doing TSO
authorBen Hutchings <bhutchings@solarflare.com>
Mon, 1 Sep 2008 11:46:40 +0000 (12:46 +0100)
committerJeff Garzik <jgarzik@redhat.com>
Wed, 3 Sep 2008 13:53:44 +0000 (09:53 -0400)
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
drivers/net/sfc/tx.c

index 0452ea6..1112775 100644 (file)
@@ -287,9 +287,14 @@ static inline int efx_enqueue_skb(struct efx_tx_queue *tx_queue,
        }
 
        /* Free the fragment we were mid-way through pushing */
-       if (unmap_len)
-               pci_unmap_page(pci_dev, unmap_addr, unmap_len,
-                              PCI_DMA_TODEVICE);
+       if (unmap_len) {
+               if (unmap_single)
+                       pci_unmap_single(pci_dev, unmap_addr, unmap_len,
+                                        PCI_DMA_TODEVICE);
+               else
+                       pci_unmap_page(pci_dev, unmap_addr, unmap_len,
+                                      PCI_DMA_TODEVICE);
+       }
 
        return rc;
 }
@@ -561,8 +566,7 @@ struct tso_state {
                /* DMA address and length of the whole fragment */
                unsigned int unmap_len;
                dma_addr_t unmap_addr;
-               struct page *page;
-               unsigned page_off;
+               unsigned int unmap_single;
        } ifc;
 
        struct {
@@ -686,18 +690,14 @@ efx_tsoh_heap_free(struct efx_tx_queue *tx_queue, struct efx_tso_header *tsoh)
  * @tx_queue:          Efx TX queue
  * @dma_addr:          DMA address of fragment
  * @len:               Length of fragment
- * @skb:               Only non-null for end of last segment
- * @end_of_packet:     True if last fragment in a packet
- * @unmap_addr:                DMA address of fragment for unmapping
- * @unmap_len:         Only set this in last segment of a fragment
+ * @final_buffer:      The final buffer inserted into the queue
  *
  * Push descriptors onto the TX queue.  Return 0 on success or 1 if
  * @tx_queue full.
  */
 static int efx_tx_queue_insert(struct efx_tx_queue *tx_queue,
                               dma_addr_t dma_addr, unsigned len,
-                              const struct sk_buff *skb, int end_of_packet,
-                              dma_addr_t unmap_addr, unsigned unmap_len)
+                              struct efx_tx_buffer **final_buffer)
 {
        struct efx_tx_buffer *buffer;
        struct efx_nic *efx = tx_queue->efx;
@@ -725,8 +725,10 @@ static int efx_tx_queue_insert(struct efx_tx_queue *tx_queue,
                        fill_level = (tx_queue->insert_count
                                      - tx_queue->old_read_count);
                        q_space = efx->type->txd_ring_mask - 1 - fill_level;
-                       if (unlikely(q_space-- <= 0))
+                       if (unlikely(q_space-- <= 0)) {
+                               *final_buffer = NULL;
                                return 1;
+                       }
                        smp_mb();
                        --tx_queue->stopped;
                }
@@ -766,10 +768,7 @@ static int efx_tx_queue_insert(struct efx_tx_queue *tx_queue,
 
        EFX_BUG_ON_PARANOID(!len);
        buffer->len = len;
-       buffer->skb = skb;
-       buffer->continuation = !end_of_packet;
-       buffer->unmap_addr = unmap_addr;
-       buffer->unmap_len = unmap_len;
+       *final_buffer = buffer;
        return 0;
 }
 
@@ -817,9 +816,16 @@ static void efx_enqueue_unwind(struct efx_tx_queue *tx_queue)
                buffer->len = 0;
                buffer->continuation = 1;
                if (buffer->unmap_len) {
-                       pci_unmap_page(tx_queue->efx->pci_dev,
-                                      buffer->unmap_addr,
-                                      buffer->unmap_len, PCI_DMA_TODEVICE);
+                       if (buffer->unmap_single)
+                               pci_unmap_single(tx_queue->efx->pci_dev,
+                                                buffer->unmap_addr,
+                                                buffer->unmap_len,
+                                                PCI_DMA_TODEVICE);
+                       else
+                               pci_unmap_page(tx_queue->efx->pci_dev,
+                                              buffer->unmap_addr,
+                                              buffer->unmap_len,
+                                              PCI_DMA_TODEVICE);
                        buffer->unmap_len = 0;
                }
        }
@@ -846,31 +852,40 @@ static inline void tso_start(struct tso_state *st, const struct sk_buff *skb)
 
        st->packet_space = st->p.full_packet_size;
        st->remaining_len = skb->len - st->p.header_length;
+       st->ifc.unmap_len = 0;
+       st->ifc.unmap_single = 0;
 }
 
-
-/**
- * tso_get_fragment - record fragment details and map for DMA
- * @st:                        TSO state
- * @efx:               Efx NIC
- * @data:              Pointer to fragment data
- * @len:               Length of fragment
- *
- * Record fragment details and map for DMA.  Return 0 on success, or
- * -%ENOMEM if DMA mapping fails.
- */
 static inline int tso_get_fragment(struct tso_state *st, struct efx_nic *efx,
-                                  int len, struct page *page, int page_off)
+                                  skb_frag_t *frag)
 {
+       st->ifc.unmap_addr = pci_map_page(efx->pci_dev, frag->page,
+                                         frag->page_offset, frag->size,
+                                         PCI_DMA_TODEVICE);
+       if (likely(!pci_dma_mapping_error(efx->pci_dev, st->ifc.unmap_addr))) {
+               st->ifc.unmap_single = 0;
+               st->ifc.unmap_len = frag->size;
+               st->ifc.len = frag->size;
+               st->ifc.dma_addr = st->ifc.unmap_addr;
+               return 0;
+       }
+       return -ENOMEM;
+}
+
+static inline int
+tso_get_head_fragment(struct tso_state *st, struct efx_nic *efx,
+                     const struct sk_buff *skb)
+{
+       int hl = st->p.header_length;
+       int len = skb_headlen(skb) - hl;
 
-       st->ifc.unmap_addr = pci_map_page(efx->pci_dev, page, page_off,
-                                         len, PCI_DMA_TODEVICE);
+       st->ifc.unmap_addr = pci_map_single(efx->pci_dev, skb->data + hl,
+                                           len, PCI_DMA_TODEVICE);
        if (likely(!pci_dma_mapping_error(efx->pci_dev, st->ifc.unmap_addr))) {
+               st->ifc.unmap_single = 1;
                st->ifc.unmap_len = len;
                st->ifc.len = len;
                st->ifc.dma_addr = st->ifc.unmap_addr;
-               st->ifc.page = page;
-               st->ifc.page_off = page_off;
                return 0;
        }
        return -ENOMEM;
@@ -891,7 +906,7 @@ static inline int tso_fill_packet_with_fragment(struct efx_tx_queue *tx_queue,
                                                const struct sk_buff *skb,
                                                struct tso_state *st)
 {
-
+       struct efx_tx_buffer *buffer;
        int n, end_of_packet, rc;
 
        if (st->ifc.len == 0)
@@ -907,16 +922,25 @@ static inline int tso_fill_packet_with_fragment(struct efx_tx_queue *tx_queue,
        st->packet_space -= n;
        st->remaining_len -= n;
        st->ifc.len -= n;
-       st->ifc.page_off += n;
-       end_of_packet = st->remaining_len == 0 || st->packet_space == 0;
 
-       rc = efx_tx_queue_insert(tx_queue, st->ifc.dma_addr, n,
-                                st->remaining_len ? NULL : skb,
-                                end_of_packet, st->ifc.unmap_addr,
-                                st->ifc.len ? 0 : st->ifc.unmap_len);
+       rc = efx_tx_queue_insert(tx_queue, st->ifc.dma_addr, n, &buffer);
+       if (likely(rc == 0)) {
+               if (st->remaining_len == 0)
+                       /* Transfer ownership of the skb */
+                       buffer->skb = skb;
 
-       st->ifc.dma_addr += n;
+               end_of_packet = st->remaining_len == 0 || st->packet_space == 0;
+               buffer->continuation = !end_of_packet;
 
+               if (st->ifc.len == 0) {
+                       /* Transfer ownership of the pci mapping */
+                       buffer->unmap_len = st->ifc.unmap_len;
+                       buffer->unmap_single = st->ifc.unmap_single;
+                       st->ifc.unmap_len = 0;
+               }
+       }
+
+       st->ifc.dma_addr += n;
        return rc;
 }
 
@@ -1008,9 +1032,9 @@ static inline int tso_start_new_packet(struct efx_tx_queue *tx_queue,
 static int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue,
                               const struct sk_buff *skb)
 {
+       struct efx_nic *efx = tx_queue->efx;
        int frag_i, rc, rc2 = NETDEV_TX_OK;
        struct tso_state state;
-       skb_frag_t *f;
 
        /* Verify TSO is safe - these checks should never fail. */
        efx_tso_check_safe(skb);
@@ -1026,25 +1050,12 @@ static int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue,
                /* Grab the first payload fragment. */
                EFX_BUG_ON_PARANOID(skb_shinfo(skb)->nr_frags < 1);
                frag_i = 0;
-               f = &skb_shinfo(skb)->frags[frag_i];
-               rc = tso_get_fragment(&state, tx_queue->efx,
-                                     f->size, f->page, f->page_offset);
+               rc = tso_get_fragment(&state, efx,
+                                     skb_shinfo(skb)->frags + frag_i);
                if (rc)
                        goto mem_err;
        } else {
-               /* It may look like this code fragment assumes that the
-                * skb->data portion does not cross a page boundary, but
-                * that is not the case.  It is guaranteed to be direct
-                * mapped memory, and therefore is physically contiguous,
-                * and so DMA will work fine.  kmap_atomic() on this region
-                * will just return the direct mapping, so that will work
-                * too.
-                */
-               int page_off = (unsigned long)skb->data & (PAGE_SIZE - 1);
-               int hl = state.p.header_length;
-               rc = tso_get_fragment(&state, tx_queue->efx,
-                                     skb_headlen(skb) - hl,
-                                     virt_to_page(skb->data), page_off + hl);
+               rc = tso_get_head_fragment(&state, efx, skb);
                if (rc)
                        goto mem_err;
                frag_i = -1;
@@ -1063,9 +1074,8 @@ static int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue,
                        if (++frag_i >= skb_shinfo(skb)->nr_frags)
                                /* End of payload reached. */
                                break;
-                       f = &skb_shinfo(skb)->frags[frag_i];
-                       rc = tso_get_fragment(&state, tx_queue->efx,
-                                             f->size, f->page, f->page_offset);
+                       rc = tso_get_fragment(&state, efx,
+                                             skb_shinfo(skb)->frags + frag_i);
                        if (rc)
                                goto mem_err;
                }
@@ -1083,8 +1093,7 @@ static int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue,
        return NETDEV_TX_OK;
 
  mem_err:
-       EFX_ERR(tx_queue->efx, "Out of memory for TSO headers, or PCI mapping"
-               " error\n");
+       EFX_ERR(efx, "Out of memory for TSO headers, or PCI mapping error\n");
        dev_kfree_skb_any((struct sk_buff *)skb);
        goto unwind;
 
@@ -1093,13 +1102,18 @@ static int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue,
 
        /* Stop the queue if it wasn't stopped before. */
        if (tx_queue->stopped == 1)
-               efx_stop_queue(tx_queue->efx);
+               efx_stop_queue(efx);
 
  unwind:
        /* Free the DMA mapping we were in the process of writing out */
-       if (state.ifc.unmap_len)
-               pci_unmap_page(tx_queue->efx->pci_dev, state.ifc.unmap_addr,
-                              state.ifc.unmap_len, PCI_DMA_TODEVICE);
+       if (state.ifc.unmap_len) {
+               if (state.ifc.unmap_single)
+                       pci_unmap_single(efx->pci_dev, state.ifc.unmap_addr,
+                                        state.ifc.unmap_len, PCI_DMA_TODEVICE);
+               else
+                       pci_unmap_page(efx->pci_dev, state.ifc.unmap_addr,
+                                      state.ifc.unmap_len, PCI_DMA_TODEVICE);
+       }
 
        efx_enqueue_unwind(tx_queue);
        return rc2;