Here are some USB fixes for 2.6.20-rc5
They include a small number of fixes for some USB bugs, and some new
device ids, all of the details are below. I've also disabled the USB
multithreaded probe option, as it broke a number of people's machines.
Most of them have been in the -mm releases, but due to LCA, some of them
have not.
The gadget driver update looks big, but David says it fixes a number of
problems with it, and it's self-contained.
Please pull from:
master.kernel.org:/pub/scm/linux/kernel/git/gregkh/usb-2.6.git/
The full patches will be sent to the linux-usb-devel mailing list, if
anyone wants to see them.
thanks,
greg k-h
drivers/usb/core/Kconfig | 16 ------
drivers/usb/core/hub.c | 9 +---
drivers/usb/host/ohci-ep93xx.c | 2 +-
drivers/usb/input/hid-core.c | 4 ++
drivers/usb/input/usbtouchscreen.c | 98 +++++++++++++++++++++--------------
drivers/usb/net/asix.c | 18 +++++--
drivers/usb/net/rndis_host.c | 23 +++++---
drivers/usb/serial/option.c | 3 +
drivers/usb/storage/unusual_devs.h | 19 +++++++
9 files changed, 114 insertions(+), 78 deletions(-)
---------------
Andres Salomon (1):
USB: asix: Detect internal PHY and enable/use accordingly
Daniel Gollub (1):
USB: rndis_host: fix crash while probing a Nokia S60 mobile
Daniel Ritz (1):
usbtouchscreen: make ITM screens report BTN_TOUCH as zero when not touched
[email protected] (1):
USB: add vendor/device id for Option GT Max 3.6 cards
Greg Kroah-Hartman (1):
USB: disable USB_MULTITHREAD_PROBE
Manuel Osdoba (1):
USB: unusual_devs.h entry for nokia 6233
Oliver Neukum (1):
USB: make usbhid ignore Imation Disc Stakka
Pete Zaitcev (1):
USB: unusual_devs.h for 0x046b:ff40
Petr Stetiar (1):
USB: Fix for typo in ohci-ep93xx.c
Hi Greg,
> They include a small number of fixes for some USB bugs, and some new
> device ids, all of the details are below. I've also disabled the USB
> multithreaded probe option, as it broke a number of people's machines.
what about the two pending patches to make device_move() working as
expected. The current implementation is broken and basically useless.
Regards
Marcel
On Mon, Jan 22, 2007 at 09:15:04PM +0100, Marcel Holtmann wrote:
> Hi Greg,
>
> > They include a small number of fixes for some USB bugs, and some new
> > device ids, all of the details are below. I've also disabled the USB
> > multithreaded probe option, as it broke a number of people's machines.
>
> what about the two pending patches to make device_move() working as
> expected. The current implementation is broken and basically useless.
I don't feel comfortable adding them to the 2.6.20 kernel at this late
period, for a variety of reasons. I really want them to be tested out
in the -mm tree and then in the full -rc series due to the fact that
they are non-trivial, and touch the driver core code that everyone uses,
not just that function.
Also, it's a new feature (reparent back to the original parent), not
necessarily a bugfix.
And yes, I realize that the functions don't work fully right now, but as
long as you don't want to re-parent the devices, they should work,
right?
thanks,
greg k-h
Hi Greg,
> > > They include a small number of fixes for some USB bugs, and some new
> > > device ids, all of the details are below. I've also disabled the USB
> > > multithreaded probe option, as it broke a number of people's machines.
> >
> > what about the two pending patches to make device_move() working as
> > expected. The current implementation is broken and basically useless.
>
> I don't feel comfortable adding them to the 2.6.20 kernel at this late
> period, for a variety of reasons. I really want them to be tested out
> in the -mm tree and then in the full -rc series due to the fact that
> they are non-trivial, and touch the driver core code that everyone uses,
> not just that function.
fair enough.
> Also, it's a new feature (reparent back to the original parent), not
> necessarily a bugfix.
>
> And yes, I realize that the functions don't work fully right now, but as
> long as you don't want to re-parent the devices, they should work,
> right?
Actually even that was broken for me.
Regards
Marcel
On Mon, 22 Jan 2007 21:28:24 +0100,
Marcel Holtmann <[email protected]> wrote:
> > I don't feel comfortable adding them to the 2.6.20 kernel at this late
> > period, for a variety of reasons. I really want them to be tested out
> > in the -mm tree and then in the full -rc series due to the fact that
> > they are non-trivial, and touch the driver core code that everyone uses,
> > not just that function.
I agree that the patch allowing device_move(dev, NULL) is non-trivial,
but the one removing the useless device_is_registered() check should be
straightforward enough.
>
> fair enough.
>
> > Also, it's a new feature (reparent back to the original parent), not
> > necessarily a bugfix.
> >
> > And yes, I realize that the functions don't work fully right now, but as
> > long as you don't want to re-parent the devices, they should work,
> > right?
>
> Actually even that was broken for me.
If removing the device_is_registered() check helps you, I'd lobby for
the inclusion of that part (and leave device_move(dev, NULL) for
post-2.6.20). (It should help for devices with no bus; for my use case,
everything worked fine.)
--
Cornelia Huck
Linux for zSeries Developer
Tel.: +49-7031-16-4837, Mail: [email protected]