On Mon, 2001-10-01 at 13:41, Greg KH wrote:
> On Mon, Oct 01, 2001 at 12:29:55PM -0700, Dmitri wrote:
> > I have one minor patch outstanding, though (hotplugging fix). It should
> > be one line change in ibmcam.c and ultracam.c files - when it is ready.
>
> Ok, then I will leave the video files alone in both trees unless you let
> me know differently.
5000+ line patch must not be delayed by one-liner :-) If you have a
moment then I would appreciate if you merge -ac2 changes into Linus's
tree, and since we are early into prereleases I should have plenty of
time to test the result (I always keep the latest official kernel on a
test box).
If that big patch works then I will update hotplugging - it is an API
misunderstanding. I haven't looked into this yet, but here is most of
the bug report, please feel free to patch that up too if you want:
> But the driver does not load properly with hotplug. It appears to be
> because of the way the version is specified in the driver. In the
> driver source the bcdDevice_lo and bcdDevice_hi fields look as though
> they are to be interpretted as the high and low bytes of the version,
> but looking at the usbutils source (called from hotplug) it appears
> that these two fields are interpretted as a range of versions
> of hardware supported by the driver.
>
> For example (original source):
>
> { USB_DEVICE_VER(IBMCAM_VENDOR_ID, NETCAM_PRODUCT_ID, 0x03, 0x0a)
> }, /* Model 4 */
>
> looks like the 0x03, 0x0a is supposed to mean version 3.0a but in
> the context of usbutils this would mean version 0.03 to 0.0a.
> Changing this line to:
>
> { USB_DEVICE_VER(IBMCAM_VENDOR_ID, NETCAM_PRODUCT_ID, 0x030a, 0x030a)
> }, /* Model 4 */
>
> causes hotplug to recognize the version correctly and insert the
> driver.
I am wading knee deep in deadlines at work, so it would take me a while
to do anything about it :-( I am already working 14/7/365... can't do
much more than that.
Thanks,
Dmitri
PGP signature
