[PATCH 6/9] staging: fsl-dpaa2/eth: Fix double DMA unmap

[Ioana Radulescu]

In case we fail to allocate a skb for a fragmented ingress
frame, the cleanup function will attempt to unmap again the
first frame fragment, which had already been unmapped during
early Rx processing.

Avoid this by freeing the first buffer immediately in case
we hit an error, leaving the cleanup function to free only
the subsequent buffers.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@xxxxxxx>
---
 drivers/staging/fsl-dpaa2/ethernet/dpaa2-eth.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/drivers/staging/fsl-dpaa2/ethernet/dpaa2-eth.c b/drivers/staging/fsl-dpaa2/ethernet/dpaa2-eth.c
index a0be9ab..6540ab0 100644
--- a/drivers/staging/fsl-dpaa2/ethernet/dpaa2-eth.c
+++ b/drivers/staging/fsl-dpaa2/ethernet/dpaa2-eth.c
@@ -104,9 +104,11 @@ static void free_rx_fd(struct dpaa2_eth_priv *priv,
 		/* We don't support any other format */
 		return;
 
-	/* For S/G frames, we first need to free all SG entries */
+	/* For S/G frames, we first need to free all SG entries
+	 * except the first one, which was taken care of already
+	 */
 	sgt = vaddr + dpaa2_fd_get_offset(fd);
-	for (i = 0; i < DPAA2_ETH_MAX_SG_ENTRIES; i++) {
+	for (i = 1; i < DPAA2_ETH_MAX_SG_ENTRIES; i++) {
 		addr = dpaa2_sg_get_addr(&sgt[i]);
 		sg_vaddr = dpaa2_iova_to_virt(priv->iommu_domain, addr);
 		dma_unmap_single(dev, addr, DPAA2_ETH_RX_BUF_SIZE,
@@ -179,6 +181,11 @@ static struct sk_buff *build_frag_skb(struct dpaa2_eth_priv *priv,
 			skb = build_skb(sg_vaddr, DPAA2_ETH_RX_BUF_SIZE +
 				SKB_DATA_ALIGN(sizeof(struct skb_shared_info)));
 			if (unlikely(!skb)) {
+				/* Free the first SG entry now, since we already
+				 * unmapped it and obtained the virtual address
+				 */
+				skb_free_frag(sg_vaddr);
+
 				/* We still need to subtract the buffers used
 				 * by this FD from our software counter
 				 */
-- 
2.7.4