Re: [PATCH v3 1/3] staging: rtl8712: fix uninit-value in usb_read8() and friends

[Pavel Skripkin <paskripkin@gmail.com>]

[-- Attachment #1.1: Type: text/plain, Size: 2021 bytes --]

Hi Wang,

On 5/6/22 14:59, Wang Cheng wrote:
> Reported-and-tested-by: syzbot+6f5ecd144854c0d8580b@syzkaller.appspotmail.com
> Signed-off-by: Wang Cheng <wanngchenng@gmail.com>
> ---
>   drivers/staging/rtl8712/usb_ops.c | 27 ++++++++++++++++++---------
>   1 file changed, 18 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/staging/rtl8712/usb_ops.c b/drivers/staging/rtl8712/usb_ops.c
> index e64845e6adf3..af9966d03979 100644
> --- a/drivers/staging/rtl8712/usb_ops.c
> +++ b/drivers/staging/rtl8712/usb_ops.c
> @@ -29,7 +29,8 @@ static u8 usb_read8(struct intf_hdl *intfhdl, u32 addr)
>   	u16 wvalue;
>   	u16 index;
>   	u16 len;
> -	__le32 data;
> +	int status;
> +	__le32 data = 0;
>   	struct intf_priv *intfpriv = intfhdl->pintfpriv;
>   
>   	request = 0x05;
> @@ -37,8 +38,10 @@ static u8 usb_read8(struct intf_hdl *intfhdl, u32 addr)
>   	index = 0;
>   	wvalue = (u16)(addr & 0x0000ffff);
>   	len = 1;
> -	r8712_usbctrl_vendorreq(intfpriv, request, wvalue, index, &data, len,
> -				requesttype);
> +	status = r8712_usbctrl_vendorreq(intfpriv, request, wvalue, index,
> +					 &data, len, requesttype);
> +	if (status < 0)
> +		return 0;

Wait, but what about partial reads? I see that r8712_usbctrl_vendorreq() 
uses `usb_control_msg()` which can read less data than was requested.

So here you are just hiding these error by zeroing data

>   	return (u8)(le32_to_cpu(data) & 0x0ff);
>   }
>   

And as i said in previous email: caller can't see an error.

Imagine caller waits for some register to become 0. Caller sees that 
usb_read8() returned 0. Is it an error? Is register really become 0?


Generic read API prototype looks like following:

int read_something(struct my_cool_dev *dev, void *data, size_t size);

and it returns 0 on success and -errno on failure and data returned via 
passed pointer. So, if API returned an error caller should not touch 
@data, since likely it's uninitialized




With regards,
Pavel Skripkin

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 840 bytes --]