In the error handling of 'offset > adapter->ring_size', the
tx_ring->tx_buffer allocated by kzalloc should be freed,
instead of 'goto failed' instantly.
Fixes: a6a5325239c2 ("atl1e: Atheros L1E Gigabit Ethernet driver")
Signed-off-by: Zhipeng Lu <[email protected]>
---
drivers/net/ethernet/atheros/atl1e/atl1e_main.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
index 5935be190b9e..deb5a3f207cc 100644
--- a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
+++ b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
@@ -866,6 +866,7 @@ static int atl1e_setup_ring_resources(struct atl1e_adapter *adapter)
netdev_err(adapter->netdev, "offset(%d) > ring size(%d) !!\n",
offset, adapter->ring_size);
err = -1;
+ kfree(tx_ring->tx_buffer);
goto failed;
}
--
2.34.1
>diff --git a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
>b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
>index 5935be190b9e..deb5a3f207cc 100644
>--- a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
>+++ b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
>@@ -866,6 +866,7 @@ static int atl1e_setup_ring_resources(struct
>atl1e_adapter *adapter)
> netdev_err(adapter->netdev, "offset(%d) > ring size(%d) !!\n",
> offset, adapter->ring_size);
> err = -1;
>+ kfree(tx_ring->tx_buffer);
[Suman] I think we should do tx_ring->tx_buffer = NULL also, to avoid use after free?
> goto failed;
> }
>
>--
>2.34.1
>
On Thu, 7 Dec 2023 17:08:15 +0000 Suman Ghosh wrote:
> >+ kfree(tx_ring->tx_buffer);
>
> [Suman] I think we should do tx_ring->tx_buffer = NULL also, to avoid use after free?
It's up to the driver. Some may call that defensive programming.
>On Thu, 7 Dec 2023 17:08:15 +0000 Suman Ghosh wrote:
>> >+ kfree(tx_ring->tx_buffer);
>>
>> [Suman] I think we should do tx_ring->tx_buffer = NULL also, to avoid
>use after free?
>
>It's up to the driver. Some may call that defensive programming.
[Suman] Agree. I pointed it out since this driver is using this approach at other places. But sure, it is up to Zhipeng.
> >On Thu, 7 Dec 2023 17:08:15 +0000 Suman Ghosh wrote:
> >> >+ kfree(tx_ring->tx_buffer);
> >>
> >> [Suman] I think we should do tx_ring->tx_buffer = NULL also, to avoid
> >use after free?
> >
> >It's up to the driver. Some may call that defensive programming.
> [Suman] Agree. I pointed it out since this driver is using this approach at other places. But sure, it is up to Zhipeng.
[Zhipeng] I think Suman's suggestion is valuable, it prevents potiential use-after-free and is consistent with other free operations in the same module.