ibmveth: Fix DMA unmap error
authorBrian King <brking@linux.vnet.ibm.com>
Wed, 7 Sep 2011 14:41:03 +0000 (14:41 +0000)
committerGreg Kroah-Hartman <gregkh@suse.de>
Mon, 3 Oct 2011 18:40:32 +0000 (11:40 -0700)
commit 33a48ab105a75d37021e422a0a3283241099b142 upstream.

Commit 6e8ab30ec677 (ibmveth: Add scatter-gather support) introduced a
DMA mapping API inconsistency resulting in dma_unmap_page getting
called on memory mapped via dma_map_single. This was seen when
CONFIG_DMA_API_DEBUG was enabled. Fix up this API usage inconsistency.

Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Acked-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
drivers/net/ibmveth.c

index f9d49cc..dc9bac0 100644 (file)
@@ -1025,7 +1025,12 @@ retry_bounce:
                netdev->stats.tx_bytes += skb->len;
        }
 
-       for (i = 0; i < skb_shinfo(skb)->nr_frags + 1; i++)
+       dma_unmap_single(&adapter->vdev->dev,
+                        descs[0].fields.address,
+                        descs[0].fields.flags_len & IBMVETH_BUF_LEN_MASK,
+                        DMA_TO_DEVICE);
+
+       for (i = 1; i < skb_shinfo(skb)->nr_frags + 1; i++)
                dma_unmap_page(&adapter->vdev->dev, descs[i].fields.address,
                               descs[i].fields.flags_len & IBMVETH_BUF_LEN_MASK,
                               DMA_TO_DEVICE);