On Fri, 23 Mar 2012, Oliver Neukum wrote:
> Hi,
>
> it turns out that there are a lot of broken HP printers.
> The key to make them work well should be to disable polling
> while printing in hp-lib. (It is switched on only by udev in the
> first place)
>
> Nevertheless the harmful commands may come from user space,
> so we need to do something about that.
>
> Given the number of affected devices I refrained from bloating the
> quirks table and coded a heuristics. What do you think?
Have you tested this? I'm doubtful about how well it will work in
practice. Many mass-storage devices don't implement the class-specific
reset properly.
> diff --git a/drivers/usb/storage/usb.c b/drivers/usb/storage/usb.c
> index db51ba1..8a53a78 100644
> --- a/drivers/usb/storage/usb.c
> +++ b/drivers/usb/storage/usb.c
> @@ -585,6 +585,16 @@ static int get_device_info(struct us_data *us, const struct usb_device_id *id,
> msgs[msg],
> utsname()->release);
> }
> + /* a collective workaround for buggy HP printer+storage combos */
> + if (id->idVendor == 0x03f0)
> + if (dev->actconfig->desc.bNumInterfaces > 1)
> + for (int i = 0; i < udev->actconfig->desc.bNumInterfaces; i++) {
> + struct usb_interface *intf = config->interface[i];
> + if (intf->cur_altsetting->desc->bInterfaceClass == 7) {
> + us->fflags |= USB_QUIRK_RESET_MORPHS;
> + break;
> + }
> + }
How about using symbolic constants instead of 0x03f0 and 7? And
including a blank line between the declaration and the following
executable statement?
Also, why use udev->actconfig in one place and config in another?
Alan Stern
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
