Re: Certain cameras no longer working with uvcvideo on recent (openSUSE) kernels

[Greg KH <gregkh@linuxfoundation.org>]

On Wed, Jan 01, 2020 at 08:35:59PM +0200, Laurent Pinchart wrote:
> Hi Roger,
> 
> (CCin'g Alan Stern and linux-usb)
> 
> On Wed, Jan 01, 2020 at 05:52:27PM +0000, Roger Whittaker wrote:
> > On Wed, Jan 01, 2020 at 07:24:49PM +0200, Laurent Pinchart wrote:
> > 
> > > The last message is worse. Could you send me the output of lsusb -v (you
> > > can restrict it to the camera with -d), if possible running as root, for
> > > both the working and non-working kernels ?
> > 
> > Thanks very much for your reply.
> > 
> > The lsusb outputs are attached - they are in fact identical to each
> > other.
> > 
> > Also attached, the dmesg lines when replugging the camera on both
> > kernels.
> 
> Thank you for the information.
> 
> I had missed the following message:
> 
> [  470.351700] usb 1-1.4.3.1: config 1 interface 2 altsetting 0 endpoint 0x82 has wMaxPacketSize 0, skipping
> 
> This seems to be the culprit, and it points to the USB core. One
> interface is ignored due to its wMaxPacketSize value, and the uvcvideo
> driver then fails to find it.
> 
> The wMaxPacketSize check was added in
> 
> commit d482c7bb0541d19dea8bff437a9f3c5563b5b2d2
> Author: Alan Stern <stern@rowland.harvard.edu>
> Date:   Mon Oct 28 10:52:35 2019 -0400
> 
>     USB: Skip endpoints with 0 maxpacket length
> 
>     Endpoints with a maxpacket length of 0 are probably useless.  They
>     can't transfer any data, and it's not at all unlikely that an HCD will
>     crash or hang when trying to handle an URB for such an endpoint.
> 
>     Currently the USB core does not check for endpoints having a maxpacket
>     value of 0.  This patch adds a check, printing a warning and skipping
>     over any endpoints it catches.
> 
>     Now, the USB spec does not rule out endpoints having maxpacket = 0.
>     But since they wouldn't have any practical use, there doesn't seem to
>     be any good reason for us to accept them.
> 
>     Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
> 
>     Link: https://lore.kernel.org/r/Pine.LNX.4.44L0.1910281050420.1485-100000@iolanthe.rowland.org
>     Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> 
> The commit was merged in v5.4 and backported to v5.3.11 in
> 47aaab6377204cdbcd16f52a23c584f994fd0d15.
> 
> For reference for Alan and linux-usb, the issue being discussed is
> described in https://bugzilla.suse.com/show_bug.cgi?id=1159811. The
> above commit seems to cause a regression with several cameras. I've
> attached to this e-mail the lsusb output provided by Roger.

How can a device work with an endpoint of 0 length?

What does the driver expect to do with those endpoints?  Does it expect
it to be present but just ignore it?

thanks,

greg k-h