RE: [PATCH] drive/realtek/rtlwifi: fix possible memory leak

[Ping-Ke Shih]

The subject should be unique and with rtw-next tag, such as

"[PATCH rtw-next] wifi: rtlwifi: fix possible skb memory leak in
_rtl_pci_init_one_rxdesc()."

Thomas Fourier <fourier.thomas@xxxxxxxxx> wrote:
> When `dma_mapping_error()` is true, if a new `skb` has been allocated,
> then it must be de-allocated.

Add a "Compile tested only". (I guess you even didn't)

> 
> Signed-off-by: Thomas Fourier <fourier.thomas@xxxxxxxxx>
> ---
>  drivers/net/wireless/realtek/rtlwifi/pci.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/wireless/realtek/rtlwifi/pci.c b/drivers/net/wireless/realtek/rtlwifi/pci.c
> index 898f597f70a9..f754f1c3f783 100644
> --- a/drivers/net/wireless/realtek/rtlwifi/pci.c
> +++ b/drivers/net/wireless/realtek/rtlwifi/pci.c
> @@ -572,8 +572,11 @@ static int _rtl_pci_init_one_rxdesc(struct ieee80211_hw *hw,
>                 dma_map_single(&rtlpci->pdev->dev, skb_tail_pointer(skb),
>                                rtlpci->rxbuffersize, DMA_FROM_DEVICE);
>         bufferaddress = *((dma_addr_t *)skb->cb);
> -       if (dma_mapping_error(&rtlpci->pdev->dev, bufferaddress))
> +       if (dma_mapping_error(&rtlpci->pdev->dev, bufferaddress)) {
> +               if (!new_skb)
> +                       kfree_skb(skb)

kernel test robot reported error here. 

>                 return 0;
> +       }
>         rtlpci->rx_ring[rxring_idx].rx_buf[desc_idx] = skb;
>         if (rtlpriv->use_new_trx_flow) {
>                 /* skb->cb may be 64 bit address */
> --
> 2.43.0