IOGear firmware versions >= 1.4.12224 fail to be downloaded because of a
spurious (and harmless) RCEB received after the download notification. This
patch handles this RCEB and keeps compatibility with future versions that might
not emit this RCEB.

i1480_rceb_check() is reused to check for the RCEB. It is also refactored with
improved comments and reused in another place in mac.c where the checking was
being duplicated.

This patch was tested on both i1480 and GUWA100U HWAs, with all firmware
versions currently available.
Signed-off-by: Anderson Lizardo <anderson.lizardo@indt.org.br>

Some hardware/firmware combinations (most notably an IOGear HWA using the i1480
firmware) kill the host controller after issuing a GET_MAC_PHY_INFO command.
Removing this check seems harmless otherwise.

The patch fixes the issue where the HC is killed, showing the message:

ehci_hcd 0000:00:1d.7: HC died; cleaning up

After this error, USB comes back only after reloading the ehci_hcd module.
Signed-off-by: Anderson Lizardo <anderson.lizardo@indt.org.br>
Signed-off-by: David Vrabel <david.vrabel@csr.com>

The Intel i1480 HWA uses WHCI commands/events even though reporting itself as
WUSB compliant. This patch fixes this by marking it with the
WUSB_QUIRK_WHCI_CMD_EVT flag, which disables WUSB command/event filtering.
Signed-off-by: Anderson Lizardo <anderson.lizardo@indt.org.br>
Signed-off-by: David Vrabel <david.vrabel@csr.com>

The D-Link DUB-1210 HWA uses commands and events from the WHCI specification,
although reporting itself as WUSB compliant. Therefore, we disable WUSB
command/event filtering for it.

USB_DEVICE_AND_INTERFACE_INFO is used for matching only the RC interface.
Signed-off-by: Anderson Lizardo <anderson.lizardo@indt.org.br>
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Call uwb_dbg_init() so the debugfs files are accessible.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

A DRP notification may sometimes have empty IE data. This patch fixes
uwb_est_get_size() so that this case is handled properly.
Signed-off-by: Anderson Lizardo <anderson.lizardo@indt.org.br>
Signed-off-by: David Vrabel <david.vrabel@csr.com>

WUSB 1.0 (Table 8-4) mentions that Abort RPipe requests must have bmRequestType
equal to 0x25, although current implementation sets bmRequestType to 0xa5. This
patch fixes this typo.
Signed-off-by: Anderson Lizardo <anderson.lizardo@indt.org.br>
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Error path for wusb_set_dev_addr() was handled incorrectly. Fix it by
considering error only when return value is negative.
Signed-off-by: Anderson Lizardo <anderson.lizardo@indt.org.br>
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add a driver for Wireless USB host controllers connected via USB (a
Host Wire Adapter or HWA).
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Common code for supporting Host Wire Adapters and Device Wire Adapters.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

A driver for Wireless USB host controllers that comply with the
Wireless Host Controller Interface (HCI) specification as published by
Intel.

The latest publically available version of the specification (0.95) is
supported (except for isochronous transfers).

Build fixes by Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add a driver for cable based associated of (Wireless) USB devices.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the WUSB build system (Kconfig and Kbuild) files.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the WUSB security (authentication) code.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the WUSB protocol (MMC management and device connection) code.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add support for Ceritified Wireless USB 1.0 to the USB stack.

This has been split into several patches for easier review.

core (this patch):
  - host controller infrastructure
  - cluster reservation
  - UWB PAL registration
  - fake root hub

protocol:
  - MMC management (start/stop, managing IEs)
  - device connection

security:
  - device authentication and authorization

build-system:
  - Kconfig and Kbuild files
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Common header files derived from the WUSB 1.0 specification.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the driver for the WLP capability of the Intel i1480 device.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the driver for downloading the firmware to an Intel i1480 device.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the WLP build system (Kconfig and Kbuild files).
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the Wireless Service Set (WSS) code.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the WLP message formatting/decoding code.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the generic code for the WiMedia Logical Link Control Protocol (WLP).

This has been split into several patches for easier review.

core (this patch):
  - everything else

messages:
  - WLP message construction/decode

wss:
  - Wireless Service Set support

build-system:
  - Kconfig and Kbuild files
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add a driver for USB-connected UWB radio controllers (HWAs).
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the driver for WHCI radio controllers.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

This enumerates the capabilties of a WHCI device, adding a umc device for
each one.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

The UMC bus is used for the capabilities exposed by a UWB Multi-interface
Controller as described in the WHCI specification.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

The Kbuild and Kconfig files.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add various debugfs files, principaly for the reservation manager.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

DRP and reservation management.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Most of the MAC Layer Management Entity (MLME) support: address, beacon, IE
and scan management.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Add the UWB radio controller interface (URCI) support.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

UWB device and radio controller device and event management.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

Signed-off-by: David Vrabel <david.vrabel@csr.com>

Documentation (and example utilities) for the UWB (and WUSB) stacks.

Some of the documentation may be out-of-date.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

bitmap_copy_le() copies a bitmap, putting the bits into little-endian
order (i.e., each unsigned long word in the bitmap is put into
little-endian order).

The UWB stack used bitmaps to manage Medium Access Slot availability,
and these bitmaps need to be written to the hardware in LE order.
Signed-off-by: David Vrabel <david.vrabel@csr.com>

this patch turns the netdev timeout WARN_ON_ONCE() into a WARN_ONCE(),
so that the device and driver names are inside the warning message.
This helps automated tools like kerneloops.org to collect the data
and do statistics, as well as making it more likely that humans
cut-n-paste the important message as part of a bugreport.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/slab-2.6:
  slub: fixed uninitialized counter in struct kmem_cache_node

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  [MIPS] VR41xx: unsigned irq cannot be negative

This fill fix the following regression list entry:

Bug-Entry	: http://bugzilla.kernel.org/show_bug.cgi?id=11276
Subject		: build error: CONFIG_OPTIMIZE_INLINING=y causes gcc 4.2 to do stupid things
Submitter	: Randy Dunlap <randy.dunlap@oracle.com>
Date		: 2008-08-06 17:18 (38 days old)
References	: http://marc.info/?l=linux-kernel&m=121804329014332&w=4
		  http://lkml.org/lkml/2008/7/22/353
Handled-By	: Bjorn Helgaas <bjorn.helgaas@hp.com>
Patch		: http://lkml.org/lkml/2008/7/22/364

with what I believe is a better fix than the one referenced
in the regression entry above.

These PNP header interfaces try to work in such a way that
you can reference some of them even if PNP is not enabled,
and the compiler was expected to optimize everything away.

Which is mostly fine, except that there was one interface
for which there was not provided an inline "NOP" implementation.

Once we add that, all of these compile failures cannot handle
any more.

pnp: Provide NOP inline implementation of pnp_get_resource() when !PNP

Fixes kernel bugzilla #11276.
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>