[lvc-project] [PATCH] usb: c67x00 - fix potential division by zero in c67x00_end_of_data()

[Alan Stern]

On Mon, Oct 06, 2025 at 05:06:30PM +0300, Pavel Zhigulin wrote:
> The function 'usb_maxpacket' relies on the value of 'epd->wMaxPacketSize',
> which can be zero if the device reports itself as an eUSB2 device
> (see 'usb_parse_endpoint' in drivers/usb/core/config.c). Under normal
> conditions everything works correctly, but if a broken or malformed
> device is handled by this module, a division by zero may occur.
> 
> Fix the division by zero by checking the result of the 'usb_maxpacket'
> call.
> 
> Found by Linux Verification Center (linuxtesting.org) with SVACE.

This should not be necessary.  usb_submit_urb() already checks whether 
the endpoint's maxpacket size is 0, and returns -EMSGSIZE if it is.

Alan Stern

> 
> Fixes: e9b29ffc519b ("USB: add Cypress c67x00 OTG controller HCD driver")
> Signed-off-by: Pavel Zhigulin <Pavel.Zhigulin at kaspersky.com>
> ---
>  drivers/usb/c67x00/c67x00-sched.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/usb/c67x00/c67x00-sched.c b/drivers/usb/c67x00/c67x00-sched.c
> index a09fa68a6ce7..3211843497cc 100644
> --- a/drivers/usb/c67x00/c67x00-sched.c
> +++ b/drivers/usb/c67x00/c67x00-sched.c
> @@ -868,6 +868,9 @@ static inline int c67x00_end_of_data(struct c67x00_td *td)
> 
>  	maxps = usb_maxpacket(td_udev(td), td->pipe);
> 
> +	if (unlikely(!maxps))
> +		return 1;
> +
>  	if (unlikely(act_bytes < maxps))
>  		return 1;	/* Smaller then full packet */
> 
> --
> 2.43.0
> 
>