Skip to contentSkip to conversation

net: nixge: fix skb leak and missing bail on DMA mapping error

PATCH In review kernelcoredev 2 participants 2 messages Apr 27

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net: nixge: fix skb leak and missing bail on DMA mapping error

When dma_mapping_error() fires during RX buffer refill in nixge_recv(), the skb allocated by netdev_alloc_skb_ip_align() is never freed, and execution continues writing a corrupt physical address into the hardware descriptor ring.

Fix by freeing the skb with dev_kfree_skb() and returning early.

This was tested via code inspection, as I do not have access to the hardware. If dma_mapping_error() occurs, the current code continues execution, which can leak the skb and program an invalid DMA address into the descriptor ring. This patch ensures proper cleanup and early return on error.

Fixes 492caffa8a1a ("net: ethernet: nixge: Add support for National Instruments XGE netdev")
---
 drivers/net/ethernet/ni/nixge.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/net/ethernet/ni/nixge.c +3 -2
index 230d5ff99..b64e2f355 100644
@@ -645,8 +645,9 @@ static int nixge_recv(struct net_device *ndev, int budget)
 					  NIXGE_MAX_JUMBO_FRAME_SIZE,
 					  DMA_FROM_DEVICE);
 		if (dma_mapping_error(ndev->dev.parent, cur_phys)) {
-			/* FIXME: bail out and clean up */
-			netdev_err(ndev, "Failed to map ...\n");
+			netdev_err(ndev, "Failed to map RX buffer\n");
+			dev_kfree_skb(new_skb);
+			return packets;
 		}
 		nixge_hw_dma_bd_set_phys(cur_p, cur_phys);
 		cur_p->cntrl = NIXGE_MAX_JUMBO_FRAME_SIZE;
Signature · 1 line
2.54.0
#1 lore raw
Quoted 5 lines · kernelcoredev

When dma_mapping_error() fires during RX buffer refill in nixge_recv(), the skb allocated by netdev_alloc_skb_ip_align() is never freed, and execution continues writing a corrupt physical address into the hardware descriptor ring.

Quoted 7 lines

Fix by freeing the skb with dev_kfree_skb() and returning early.

This was tested via code inspection, as I do not have access to the hardware. If dma_mapping_error() occurs, the current code continues execution, which can leak the skb and program an invalid DMA address into the descriptor ring. This patch ensures proper cleanup and early return on error.

Please continue your code inspection. What happens to the ring descriptors if you return early?

If the fix was so simple, why do you think there is a FIXME? This is the problem of not testing on hardware. Do you know if your change makes the code better or worse? Please argue why it is better, including an explanation of what happens to the ring descriptors with your fix in place.

Andrew