On Sun, Jul 02, 2023 at 12:39:20AM +0800, Zhang Shurong wrote:
> If wIndex is 0 (and it often is), these calculations underflow and
> UBSAN complains, here resolve this by not decrementing the index when
> it is equal to 0.
>
> Similar commit 85e3990bea49 ("USB: EHCI: avoid undefined pointer
> arithmetic and placate UBSAN")
>
> The changes in this version:
> - fix some compile error
>
> Signed-off-by: Zhang Shurong <zhang_shurong@xxxxxxxxxxx>
> ---
> drivers/usb/host/r8a66597-hcd.c | 6 ++++--
> 1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/usb/host/r8a66597-hcd.c b/drivers/usb/host/r8a66597-hcd.c
> index 9f4bf8c5f8a5..6c597c668364 100644
> --- a/drivers/usb/host/r8a66597-hcd.c
> +++ b/drivers/usb/host/r8a66597-hcd.c
> @@ -2141,10 +2141,12 @@ static int r8a66597_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue,
> {
> struct r8a66597 *r8a66597 = hcd_to_r8a66597(hcd);
> int ret;
> - int port = (wIndex & 0x00FF) - 1;
> - struct r8a66597_root_hub *rh = &r8a66597->root_hub[port];
> unsigned long flags;
> + struct r8a66597_root_hub *rh;
> + u32 port = wIndex & 0xFF;
>
> + port -= (port > 0);
I have no idea about this hardware, but it seems strange to me that
calling r8a66597_hub_control with wIndex = 1 should have the same effect
as with wIndex = 0. Is you changed backed by knowledge about the
hardware, or is that just the most obvious way to get rid of the UB
warning?
Having said that, I think
port -= (port > 0);
is hard to read compared to:
if (port > 0)
port--;
Best regards
Uwe
--
Pengutronix e.K. | Uwe Kleine-König |
Industrial Linux Solutions | https://www.pengutronix.de/ |
Attachment:
signature.asc
Description: PGP signature
