Skip to contentSkip to conversation

sfc: allocate a big enough SKB for loopback selftest packet

PATCH Applied [email protected] 3 participants 3 messages Aug 21, 2023 – Aug 22, 2023

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 sfc: allocate a big enough SKB for loopback selftest packet

From: Edward Cree <[email protected]>

Cited commits passed a size to alloc_skb that was only big enough for
the actual packet contents, but the following skb_put + memcpy writes the whole struct efx_loopback_payload including leading and trailing padding bytes (which are then stripped off with skb_pull/skb_trim).
This could cause an skb_over_panic, although in practice we get saved
by kmalloc_size_roundup.
Pass the entire size we use, instead of the size of the final packet.

Fixes cf60ed469629 ("sfc: use padding to fix alignment in loopback test")
Fixes 30c24dd87f3f ("sfc: siena: use padding to fix alignment in loopback test")
Fixes 1186c6b31ee1 ("sfc: falcon: use padding to fix alignment in loopback test")
---
 drivers/net/ethernet/sfc/falcon/selftest.c | 2 +-
 drivers/net/ethernet/sfc/selftest.c        | 2 +-
 drivers/net/ethernet/sfc/siena/selftest.c  | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)
drivers/net/ethernet/sfc/falcon/selftest.c +1 -1
index cf1d67b6d86d..c3dc88e6c26c 100644
@@ -428,7 +428,7 @@ static int ef4_begin_loopback(struct ef4_tx_queue *tx_queue)
 	for (i = 0; i < state->packet_count; i++) {
 		/* Allocate an skb, holding an extra reference for
 		 * transmit completion counting */
-		skb = alloc_skb(EF4_LOOPBACK_PAYLOAD_LEN, GFP_KERNEL);
+		skb = alloc_skb(sizeof(state->payload), GFP_KERNEL);
 		if (!skb)
 			return -ENOMEM;
 		state->skbs[i] = skb;
drivers/net/ethernet/sfc/selftest.c +1 -1
index 19a0b8584afb..563c1e317ce9 100644
@@ -426,7 +426,7 @@ static int efx_begin_loopback(struct efx_tx_queue *tx_queue)
 	for (i = 0; i < state->packet_count; i++) {
 		/* Allocate an skb, holding an extra reference for
 		 * transmit completion counting */
-		skb = alloc_skb(EFX_LOOPBACK_PAYLOAD_LEN, GFP_KERNEL);
+		skb = alloc_skb(sizeof(state->payload), GFP_KERNEL);
 		if (!skb)
 			return -ENOMEM;
 		state->skbs[i] = skb;
drivers/net/ethernet/sfc/siena/selftest.c +1 -1
index b55fd3346972..526da43d4b61 100644
@@ -426,7 +426,7 @@ static int efx_begin_loopback(struct efx_tx_queue *tx_queue)
 	for (i = 0; i < state->packet_count; i++) {
 		/* Allocate an skb, holding an extra reference for
 		 * transmit completion counting */
-		skb = alloc_skb(EFX_LOOPBACK_PAYLOAD_LEN, GFP_KERNEL);
+		skb = alloc_skb(sizeof(state->payload), GFP_KERNEL);
 		if (!skb)
 			return -ENOMEM;
 		state->skbs[i] = skb;
reviewer Reviewed-by on v1 1/1
#1 lore raw
Quoted 10 lines · On Mon, Aug 21, 2023 at 07:01:53PM +0100, [email protected] wrote:

From: Edward Cree <[email protected]>

Cited commits passed a size to alloc_skb that was only big enough for
the actual packet contents, but the following skb_put + memcpy writes the whole struct efx_loopback_payload including leading and trailing padding bytes (which are then stripped off with skb_pull/skb_trim).
This could cause an skb_over_panic, although in practice we get saved
by kmalloc_size_roundup.
Pass the entire size we use, instead of the size of the final packet.

Quoted 5 lines
Fixes cf60ed469629 ("sfc: use padding to fix alignment in loopback test")
Fixes 30c24dd87f3f ("sfc: siena: use padding to fix alignment in loopback test")
Fixes 1186c6b31ee1 ("sfc: falcon: use padding to fix alignment in loopback test")
#2 lore raw
In reply to [email protected]
This patch was applied to netdev/net.git (main) by Jakub Kicinski <[email protected]>:

Hello:

This patch was applied to netdev/net.git (main)
by Jakub Kicinski <[email protected]>:

Quoted 11 lines · On Mon, 21 Aug 2023 19:01:53 +0100 you wrote:

From: Edward Cree <[email protected]>

Cited commits passed a size to alloc_skb that was only big enough for
the actual packet contents, but the following skb_put + memcpy writes the whole struct efx_loopback_payload including leading and trailing padding bytes (which are then stripped off with skb_pull/skb_trim).
This could cause an skb_over_panic, although in practice we get saved
by kmalloc_size_roundup.
Pass the entire size we use, instead of the size of the final packet.

[...]

Here is the summary with links:
- [net] sfc: allocate a big enough SKB for loopback selftest packet

    https://git.kernel.org/netdev/net/c/6dc5774deefe

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html