From: Sonia Sharma <[email protected]>
The switch statement in netvsc_send_completion() is incorrectly validating
the length of incoming network packets by falling through to the next case.
Avoid the fallthrough. Instead break after a case match and then process
the complete() call.
The current code has not caused any known failures. But nonetheless, the
code should be corrected as a different ordering of the switch cases might
cause a length check to fail when it should not.
Signed-off-by: Sonia Sharma <[email protected]>
---
Changes in v3:
* added return statement in default case as pointed by Michael Kelley.
Changes in v4:
* added fixes tag
* modified commit message to explain the issue fixed by patch.
Changes in v5:
* Dropped fixes tag as suggested by Simon Horman.
* fixed indentation
Changes in v7:
* Dropped the prefix "net" from subject line.
---
drivers/net/hyperv/netvsc.c | 18 ++++++++++--------
1 file changed, 10 insertions(+), 8 deletions(-)
diff --git a/drivers/net/hyperv/netvsc.c b/drivers/net/hyperv/netvsc.c
index 82e9796c8f5e..0f7e4d377776 100644
--- a/drivers/net/hyperv/netvsc.c
+++ b/drivers/net/hyperv/netvsc.c
@@ -851,7 +851,7 @@ static void netvsc_send_completion(struct net_device *ndev,
msglen);
return;
}
- fallthrough;
+ break;
case NVSP_MSG1_TYPE_SEND_RECV_BUF_COMPLETE:
if (msglen < sizeof(struct nvsp_message_header) +
@@ -860,7 +860,7 @@ static void netvsc_send_completion(struct net_device *ndev,
msglen);
return;
}
- fallthrough;
+ break;
case NVSP_MSG1_TYPE_SEND_SEND_BUF_COMPLETE:
if (msglen < sizeof(struct nvsp_message_header) +
@@ -869,7 +869,7 @@ static void netvsc_send_completion(struct net_device *ndev,
msglen);
return;
}
- fallthrough;
+ break;
case NVSP_MSG5_TYPE_SUBCHANNEL:
if (msglen < sizeof(struct nvsp_message_header) +
@@ -878,10 +878,6 @@ static void netvsc_send_completion(struct net_device *ndev,
msglen);
return;
}
- /* Copy the response back */
- memcpy(&net_device->channel_init_pkt, nvsp_packet,
- sizeof(struct nvsp_message));
- complete(&net_device->channel_init_wait);
break;
case NVSP_MSG1_TYPE_SEND_RNDIS_PKT_COMPLETE:
@@ -904,13 +900,19 @@ static void netvsc_send_completion(struct net_device *ndev,
netvsc_send_tx_complete(ndev, net_device, incoming_channel,
desc, budget);
- break;
+ return;
default:
netdev_err(ndev,
"Unknown send completion type %d received!!\n",
nvsp_packet->hdr.msg_type);
+ return;
}
+
+ /* Copy the response back */
+ memcpy(&net_device->channel_init_pkt, nvsp_packet,
+ sizeof(struct nvsp_message));
+ complete(&net_device->channel_init_wait);
}
static u32 netvsc_get_next_send_section(struct netvsc_device *net_device)
--
2.25.1
On Mon, Oct 09, 2023 at 01:00:16AM -0700, Sonia Sharma wrote:
> From: Sonia Sharma <[email protected]>
>
> The switch statement in netvsc_send_completion() is incorrectly validating
> the length of incoming network packets by falling through to the next case.
> Avoid the fallthrough. Instead break after a case match and then process
> the complete() call.
> The current code has not caused any known failures. But nonetheless, the
> code should be corrected as a different ordering of the switch cases might
> cause a length check to fail when it should not.
>
> Signed-off-by: Sonia Sharma <[email protected]>
>
> ---
> Changes in v3:
> * added return statement in default case as pointed by Michael Kelley.
> Changes in v4:
> * added fixes tag
> * modified commit message to explain the issue fixed by patch.
> Changes in v5:
> * Dropped fixes tag as suggested by Simon Horman.
> * fixed indentation
> Changes in v7:
> * Dropped the prefix "net" from subject line.
This seems address the concerns raised by Michael Kelly in his review of v6.
It doesn't include his (or my) Reviewed-by tag(s).
That not withstanding the logic change and overall patch looks good to me.
Reviewed-by: Simon Horman <[email protected]>
Hello:
This patch was applied to netdev/net-next.git (main)
by David S. Miller <[email protected]>:
On Mon, 9 Oct 2023 01:00:16 -0700 you wrote:
> From: Sonia Sharma <[email protected]>
>
> The switch statement in netvsc_send_completion() is incorrectly validating
> the length of incoming network packets by falling through to the next case.
> Avoid the fallthrough. Instead break after a case match and then process
> the complete() call.
> The current code has not caused any known failures. But nonetheless, the
> code should be corrected as a different ordering of the switch cases might
> cause a length check to fail when it should not.
>
> [...]
Here is the summary with links:
- [net-next,v7] hv_netvsc: fix netvsc_send_completion to avoid multiple message length checks
https://git.kernel.org/netdev/net-next/c/9bae5b055022
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html