[PATCH] usb: Fix switch statement in ohci-tmio.c

Tobias Boege tobias at gambas-buch.de
Thu Aug 14 15:04:16 EDT 2014


On Thu, 14 Aug 2014, Nicholas Krause wrote:
> I am fixing the bug on at the link, https://bugzilla.kernel.org/show_bug.cgi?id=79931.
> This bug report states that in the function, tmio_hc_stop the switch has no needed breaks.
> Further more this patch fixes this bug by adding the needed breaks.
> 
> Signed-off-by: Nicholas Krause <xerofoify at gmail.com>
> ---
>  drivers/usb/host/ohci-tmio.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/drivers/usb/host/ohci-tmio.c b/drivers/usb/host/ohci-tmio.c
> index bb40958..d2b5382 100644
> --- a/drivers/usb/host/ohci-tmio.c
> +++ b/drivers/usb/host/ohci-tmio.c
> @@ -100,12 +100,16 @@ static void tmio_stop_hc(struct platform_device *dev)
>  	switch (ohci->num_ports) {
>  		default:
>  			dev_err(&dev->dev, "Unsupported amount of ports: %d\n", ohci->num_ports);
> +			break;
>  		case 3:
>  			pm |= CCR_PM_USBPW3;
> +			break;
>  		case 2:
>  			pm |= CCR_PM_USBPW2;
> +			break;
>  		case 1:
>  			pm |= CCR_PM_USBPW1;
> +			break;
>  	}
>  	tmio_iowrite8(0, tmio->ccr + CCR_INTC);
>  	tmio_iowrite8(0, tmio->ccr + CCR_ILME);
> -- 
> 1.9.1
> 

Hmm, I don't know, man. The problem might as well (or even more likely) be
the defines:

#define CCR_PM_USBPW1    0x0004
#define CCR_PM_USBPW2    0x0008
#define CCR_PM_USBPW3    0x0008

which perfectly explain the warning and CCR_PM_USBPW3 needing to be 0x0010
would make more sense if you look at the "switch" head. Alas, I have no idea
about the stuff going on there, so this remains speculation.

However, maybe you can prove me wrong? If not, keep in mind that a bug
reporter doesn't necessarily provide a correct guide to fixing the bug.

[ Also, please consider just how *many* people gave you the (still surpris-
  ingly friendly formulated) advice to be silent for a while. Don't make me
  look up the procmailrc syntax again... ]

Regards,
Tobi

-- 
"There's an old saying: Don't change anything... ever!" -- Mr. Monk




More information about the Kernelnewbies mailing list