pandora-kernel.git
12 years agousb: gadget: storage: add superspeed support
Felipe Balbi [Wed, 3 Aug 2011 11:33:27 +0000 (14:33 +0300)]
usb: gadget: storage: add superspeed support

this patch adds superspeed descriptors for the
storage gadgets.

Acked-by: Michal Nazarewicz <mina86@mina86.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: fix typo for default U1/U2 exit latencies
Felipe Balbi [Mon, 10 Oct 2011 06:43:44 +0000 (09:43 +0300)]
usb: gadget: fix typo for default U1/U2 exit latencies

s/DEFULT/DEFAULT/, no functional changes.

Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: r8a66597: convert to new style
Felipe Balbi [Mon, 10 Oct 2011 14:18:51 +0000 (17:18 +0300)]
usb: gadget: r8a66597: convert to new style

convert to new style UDC registration and remove
the global 'the_controller' pointer.

Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: renesas: convert to new style
Felipe Balbi [Mon, 10 Oct 2011 14:11:20 +0000 (17:11 +0300)]
usb: gadget: renesas: convert to new style

using udc_start()/udc_stop() as we should.

Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: net2280: convert to new style
Felipe Balbi [Mon, 10 Oct 2011 07:37:17 +0000 (10:37 +0300)]
usb: gadget: net2280: convert to new style

with this patch we can finally remove the global
"the_controller" pointer.

Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: net2272: convert to new style
Felipe Balbi [Mon, 10 Oct 2011 07:33:47 +0000 (10:33 +0300)]
usb: gadget: net2272: convert to new style

with this patch we can finally remove the global
"the controller" pointer.

Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: langwell: convert to new style
Felipe Balbi [Sat, 8 Oct 2011 10:17:34 +0000 (13:17 +0300)]
usb: gadget: langwell: convert to new style

with this patch we can finally remove the global
"the_controller" pointer.

Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: langwell: don't use the_controller when it's not needed
Felipe Balbi [Sat, 8 Oct 2011 10:08:01 +0000 (13:08 +0300)]
usb: gadget: langwell: don't use the_controller when it's not needed

in most cases, we can fetch our structure by
calling dev_get_drvdata().

Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: at91_udc: change french comments to english
Robert Schwebel [Thu, 6 Oct 2011 19:36:26 +0000 (21:36 +0200)]
usb: gadget: at91_udc: change french comments to english

While being there, change C++ style comments to /* */.

Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
Cc: Greg Kroah-Hartman <gregkh@suse.de>
Cc: linux-usb@vger.kernel.org
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: pxa25x: is_vbus_present is gone
Arnd Bergmann [Sat, 1 Oct 2011 20:03:33 +0000 (22:03 +0200)]
usb: gadget: pxa25x: is_vbus_present is gone

A recent commit obsoleted the is_vbus_present function, so
we must not use it any more.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: storage: fix mass storage gadgets to work with Synopsys UDC
Paul Zimmerman [Fri, 30 Sep 2011 22:26:06 +0000 (15:26 -0700)]
usb: gadget: storage: fix mass storage gadgets to work with Synopsys UDC

The Synopsys USB device controller requires all OUT transfer request
lengths to be aligned to max packet size. The mass storage gadgets do
not meet this requirement for Super Speed. The gadgets already have a
function which performs this alignment for CBW packets, so use it for
data packets too.

The alternative would be to implement bounce buffers in the DWC3
driver, but that could have a significant impact on performance.

This version is based upon a more-correct patch written by Alan
Stern.

Signed-off-by: Paul Zimmerman <paulz@synopsys.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agoUSB: gadget: f_midi: allow a dynamic number of input and output ports
Daniel Mack [Wed, 28 Sep 2011 14:41:34 +0000 (16:41 +0200)]
USB: gadget: f_midi: allow a dynamic number of input and output ports

The code in the MIDI gadget was already sort of prepared for multi-port
configuration, so the streaming logic itself didn't need much tweaking.
However, the descriptors change when the number of ports do, and so some
rework of the the preparation algorithms were necessary.

Successfully tested on Linux and Max OS X hosts for both input and
output streams.

Signed-off-by: Daniel Mack <zonque@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agoUSB: gadget: midi: refactor the device code
Daniel Mack [Wed, 28 Sep 2011 14:41:33 +0000 (16:41 +0200)]
USB: gadget: midi: refactor the device code

Make use of the newly added MIDI function in f_midi.c and strip down
the MIDI gadget code radically. Also use the generic framework function
to avoid code duplication and rename some symbols to bring them in sync
with other code in the gadget framework.

[ balbi@ti.com : fix Section mismatch warnings.
rebased on top of usb_speed_string() patch to
avoid conflicts. ]

Signed-off-by: Daniel Mack <zonque@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agoUSB: gadget: midi: add midi function driver
Daniel Mack [Wed, 28 Sep 2011 14:41:32 +0000 (16:41 +0200)]
USB: gadget: midi: add midi function driver

This patch adds f_midi.c to implement a USB gadget function that works
with the composite framework, so it can be combined with other USB
functions.

The code for the ALSA/MIDI logic was taken from the midi device gadget,
other parts have been rewritten to benefit from the dynamic descriptor
allocation features.

This was successfully tested on an OMAP3 board.

Signed-off-by: Daniel Mack <zonque@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb gadget: fix langwell_udc.c build error
Randy Dunlap [Fri, 7 Oct 2011 20:57:42 +0000 (13:57 -0700)]
usb gadget: fix langwell_udc.c build error

Move function to fix langwell_udc.c build error:

drivers/usb/gadget/langwell_udc.c: In function 'show_langwell_udc':
drivers/usb/gadget/langwell_udc.c:1693:3: error: implicit declaration of function 'lpm_device_speed'
drivers/usb/gadget/langwell_udc.c: At top level:
drivers/usb/gadget/langwell_udc.c:2637:37: error: conflicting types for 'lpm_device_speed'
drivers/usb/gadget/langwell_udc.c:1693:20: note: previous implicit declaration of 'lpm_device_speed' was here

Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Acked-by: Felipe Balbi <balbi@ti.com>
Cc: linux-usb@vger.kernel.org

12 years agousb: gadget: r8a66597-udc: add support for SUDMAC
Yoshihiro Shimoda [Fri, 30 Sep 2011 11:07:38 +0000 (20:07 +0900)]
usb: gadget: r8a66597-udc: add support for SUDMAC

SH7757 has a USB function with internal DMA controller (SUDMAC).
This patch supports the SUDMAC. The SUDMAC is incompatible with
general-purpose DMAC. So, it doesn't use dmaengine.

Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: r8a66597-udc: use dev_*() instead of printk()
Yoshihiro Shimoda [Fri, 30 Sep 2011 11:07:30 +0000 (20:07 +0900)]
usb: gadget: r8a66597-udc: use dev_*() instead of printk()

This patch also fix the balance of braces.

Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: r8a66597-udc: add function for external controller
Yoshihiro Shimoda [Fri, 30 Sep 2011 11:07:21 +0000 (20:07 +0900)]
usb: gadget: r8a66597-udc: add function for external controller

R8A66597 has the pin of WR0 and WR1. So, if one write-pin of CPU
connects to the pins, we have to change the setting of FIFOSEL
register in the controller. If we don't change the setting,
the controller cannot send the data of odd length.

Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agousb: gadget: r8a66597-udc: change prototype of r8a66597_write_fifo
Yoshihiro Shimoda [Fri, 30 Sep 2011 11:07:08 +0000 (20:07 +0900)]
usb: gadget: r8a66597-udc: change prototype of r8a66597_write_fifo

Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
12 years agoUSB: fix ehci alignment error
Harro Haan [Mon, 10 Oct 2011 12:38:27 +0000 (14:38 +0200)]
USB: fix ehci alignment error

The Kirkwood gave an unaligned memory access error on
line 742 of drivers/usb/host/echi-hcd.c:
"ehci->last_periodic_enable = ktime_get_real();"

Signed-off-by: Harro Haan <hrhaan@gmail.com>
Cc: stable <stable@kernel.org>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: ftdi_sio: add PID for Sony Ericsson Urban
Hakan Kvist [Mon, 3 Oct 2011 11:41:15 +0000 (13:41 +0200)]
USB: ftdi_sio: add PID for Sony Ericsson Urban

Add PID 0xfc8a, 0xfc8b for device Sony Ericsson Urban

Signed-off-by: Hakan Kvist <hakan.kvist@sonyericsson.com>
Signed-off-by: Oskar Andero <oskar.andero@sonyericsson.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: Add module.h to various dwc3 drivers
Paul Gortmaker [Fri, 30 Sep 2011 22:08:59 +0000 (18:08 -0400)]
usb: Add module.h to various dwc3 drivers

These files uses the full set of MODULE_ macros and so need to
include module.h directly.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: convert structures into bitshifts
Felipe Balbi [Fri, 30 Sep 2011 07:58:51 +0000 (10:58 +0300)]
usb: dwc3: convert structures into bitshifts

our parameter structures need to be written to
HW, so instead of assuming little endian, we
convert those into bit shifts.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: gadget: allow clock gating to work
Felipe Balbi [Fri, 30 Sep 2011 07:58:50 +0000 (10:58 +0300)]
usb: dwc3: gadget: allow clock gating to work

The dwc3 core has internal clock gating support.

Let's allow that to happen by clearing the disable
bit in GCTL register.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: core: cache GHWPARAMS* registers
Felipe Balbi [Fri, 30 Sep 2011 07:58:49 +0000 (10:58 +0300)]
usb: dwc3: core: cache GHWPARAMS* registers

cache the contents of GHWPARAMS* registers in
our device structure for easy access.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: add struct dwc3_hwparams
Felipe Balbi [Fri, 30 Sep 2011 07:58:48 +0000 (10:58 +0300)]
usb: dwc3: add struct dwc3_hwparams

That structure will hold a copy of readonly
GHWPARAMS* registers for ease accessing by
the driver.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: gadget: implement streams support
Felipe Balbi [Fri, 30 Sep 2011 07:58:47 +0000 (10:58 +0300)]
usb: dwc3: gadget: implement streams support

The following patch adds support for streams
to dwc3 driver.

While at that, also fix one small issue on
endpoint disable where we should clear all
flags not only ENABLED.

Reviewied-by: Paul Zimmerman <paulz@synopsys.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: gadget: add support for Bursts
Felipe Balbi [Fri, 30 Sep 2011 07:58:46 +0000 (10:58 +0300)]
usb: dwc3: gadget: add support for Bursts

We already have the value from gadget drivers,
just need to pass it to our controller.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: gadget: fix DMA offset calculation
Paul Zimmerman [Fri, 30 Sep 2011 07:58:45 +0000 (10:58 +0300)]
usb: dwc3: gadget: fix DMA offset calculation

Fix offset calculation in dwc3_trb_dma_offset()

Signed-off-by: Paul Zimmerman <paulz@synopsys.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: gadget: make DWC3_EP_WEDGE do the right thing
Paul Zimmerman [Fri, 30 Sep 2011 07:58:44 +0000 (10:58 +0300)]
usb: dwc3: gadget: make DWC3_EP_WEDGE do the right thing

This makes DWC3_EP_WEDGE do the right thing, which is
prevent DWC3_EP_WEDGE from ever being cleared by a
ClearFeature(HALT) command.

[ balbi@ti.com : allowed set_wedge to send SetHalt command
to controller ]

Signed-off-by: Paul Zimmerman <paulz@synopsys.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: gadget: driver should not wait for RxFIFO to drain
Paul Zimmerman [Fri, 30 Sep 2011 07:58:43 +0000 (10:58 +0300)]
usb: dwc3: gadget: driver should not wait for RxFIFO to drain

An older version of the databook said to wait for the FIFO to
drain, but that has been removed from the newer databooks.

Waiting for RxFIFO to drain caused problems when testing against
one of the host controllers available in the market.

After talking to one of the RTL engineers, he stated that we
should _not_ wait for RxFIFO to drain.

Signed-off-by: Paul Zimmerman <paulz@synopsys.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: gadget: fix DEPSTARTCFG for non-EP0 EPs
Paul Zimmerman [Fri, 30 Sep 2011 07:58:42 +0000 (10:58 +0300)]
usb: dwc3: gadget: fix DEPSTARTCFG for non-EP0 EPs

DEPSTARTCFG for non-EP0 EPs must only be sent once per config

[ balbi@ti.com : changed config_start to start_config_issued ]

Signed-off-by: Paul Zimmerman <paulz@synopsys.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc: remove "All rights reserved" statement.
Sebastian Andrzej Siewior [Fri, 30 Sep 2011 07:58:41 +0000 (10:58 +0300)]
usb: dwc: remove "All rights reserved" statement.

Some people think that this line is not compatible with the GPL. The
statement was required due to the Buenos Aires Convention and is now
deprecated. I remove it because it is said that it is pointless nowdays.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: ep0: fix debug output
Sebastian Andrzej Siewior [Fri, 30 Sep 2011 07:58:40 +0000 (10:58 +0300)]
usb: dwc3: ep0: fix debug output

Use "ep0in" and "ep0out" instead "ep1in" and "ep0out" which is confusing
and not consistent with the remaining output.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: ep0: remove second giveback in error case
Sebastian Andrzej Siewior [Fri, 30 Sep 2011 07:58:39 +0000 (10:58 +0300)]
usb: dwc3: ep0: remove second giveback in error case

We already give requests back in dwc3_ep0_stall_and_restart() so
doing it again here will most likely corrupt the list.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: ep0: fix debug message
Felipe Balbi [Fri, 30 Sep 2011 07:58:38 +0000 (10:58 +0300)]
usb: dwc3: ep0: fix debug message

The way it was before was really meaningless.
Now it looks saner.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: ep0: ignore direction on 2-stage transfer
Felipe Balbi [Fri, 30 Sep 2011 07:58:37 +0000 (10:58 +0300)]
usb: dwc3: ep0: ignore direction on 2-stage transfer

We don't need to care about direction on a two stage
transfer.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: dwc3: ep0: Make USB30CV happy with SetAddress
Felipe Balbi [Fri, 30 Sep 2011 07:58:36 +0000 (10:58 +0300)]
usb: dwc3: ep0: Make USB30CV happy with SetAddress

According to USB 3.0 Specification, a SetAddress()
while device is in Configured State has an unspecified
behavior (see Section 9.4.6). Still USB30CV wasn't
happy with my Stall reply.

To make that thing happy, just accept the SetAddress()
always. No problems have been observed thus far.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: add quirk for Logitech C300 web cam
Jon Levell [Thu, 29 Sep 2011 19:42:52 +0000 (20:42 +0100)]
USB: add quirk for Logitech C300 web cam

My webcam is a Logitech C300 and I get "chipmunk"ed squeaky sound.
The following trivial patch fixes it.

Signed-off-by: Jon Levell <linuxusb@coralbark.net>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agouser namespace: usb: make usb urbs user namespace aware (v2)
Serge Hallyn [Mon, 26 Sep 2011 15:45:18 +0000 (10:45 -0500)]
user namespace: usb: make usb urbs user namespace aware (v2)

Add to the dev_state and alloc_async structures the user namespace
corresponding to the uid and euid.  Pass these to kill_pid_info_as_uid(),
which can then implement a proper, user-namespace-aware uid check.

Changelog:
Sep 20: Per Oleg's suggestion: Instead of caching and passing user namespace,
uid, and euid each separately, pass a struct cred.
Sep 26: Address Alan Stern's comments: don't define a struct cred at
usbdev_open(), and take and put a cred at async_completed() to
ensure it lasts for the duration of kill_pid_info_as_cred().

Signed-off-by: Serge Hallyn <serge.hallyn@canonical.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: message: cleanup min_t() cast in usb_sg_init()
Dan Carpenter [Tue, 27 Sep 2011 06:25:19 +0000 (09:25 +0300)]
USB: message: cleanup min_t() cast in usb_sg_init()

"length" is type size_t so the cast to unsigned int truncates the
upper bytes.  This isn't an issue in real life (I've checked the
callers) but it's a bit messy.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: qcserial: add device ID for "HP un2430 Mobile Broadband Module"
Rigbert Hamisch [Tue, 27 Sep 2011 08:46:43 +0000 (10:46 +0200)]
USB: qcserial: add device ID for "HP un2430 Mobile Broadband Module"

add device ID for "HP un2430 Mobile Broadband Module"

Signed-off-by: Rigbert Hamisch <rigbert@gmx.de>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb/core/devio.c: Check for printer class specific request
Matthias Dellweg [Sun, 25 Sep 2011 12:26:25 +0000 (14:26 +0200)]
usb/core/devio.c: Check for printer class specific request

In the usb printer class specific request get_device_id the value of
wIndex is (interface << 8 | altsetting) instead of just interface.
This enables the detection of some printers with libusb.

Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Matthias Dellweg <2500@gmx.de>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: pid_ns: ensure pid is not freed during kill_pid_info_as_uid
Serge Hallyn [Mon, 26 Sep 2011 15:18:29 +0000 (10:18 -0500)]
USB: pid_ns: ensure pid is not freed during kill_pid_info_as_uid

Alan Stern points out that after spin_unlock(&ps->lock) there is no
guarantee that ps->pid won't be freed.  Since kill_pid_info_as_uid() is
called after the spin_unlock(), the pid passed to it must be pinned.

Reported-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Serge Hallyn <serge.hallyn@canonical.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: UHCI: improve comments and logic for root-hub suspend
Alan Stern [Mon, 26 Sep 2011 15:25:26 +0000 (11:25 -0400)]
USB: UHCI: improve comments and logic for root-hub suspend

This patch (as1488) improves the comments and logic in uhci-hcd's
suspend routine.  The existing comments are hard to understand and
don't give a good idea of what's really going on.

The question of whether EGSM (Enter Global Suspend Mode) and RD
(enable Resume Detect interrupts) can be useful when they're not both
set is difficult.  The spec doesn't give any details on how they
interact with system wakeup, although clearly they are meant to be
used together.  To be safe, the patch changes the subroutine so that
neither bit gets set unless they both do.  There shouldn't be any
functional changes from this; only systems that are designed badly or
broken in some way need to avoid using those bits.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: Update USB default wakeup settings
Alan Stern [Mon, 26 Sep 2011 15:23:38 +0000 (11:23 -0400)]
USB: Update USB default wakeup settings

This patch (as1486) implements the kernel's new wakeup policy for USB
host controllers.  Since they don't generate wakeup requests on their
but merely forward requests from their root hubs toward the CPU, they
should be enabled for wakeup by default.

Also, to be compliant with both the old and new policies, root hubs
should not be enabled for remote wakeup by default.  Userspace must
enable it explicitly if it is desired.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb/xhci: remove CONFIG_PCI in xhci.c's probe function
Sebastian Andrzej Siewior [Fri, 23 Sep 2011 21:20:02 +0000 (14:20 -0700)]
usb/xhci: remove CONFIG_PCI in xhci.c's probe function

This removes the need of ifdefs within the init function and with it the
headache about the correct clean without bus X but with bus/platform Y &
Z.
xhci-pci is only compiled if CONFIG_PCI is selected which can be
de-selected now without trouble. For now the result is kinda useless
because we have no other glue code. However, since nobody is using
USB_ARCH_HAS_XHCI then it should not be an issue :)

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb/xhci: move xhci_gen_setup() away from -pci.
Sebastian Andrzej Siewior [Fri, 23 Sep 2011 21:20:01 +0000 (14:20 -0700)]
usb/xhci: move xhci_gen_setup() away from -pci.

xhci_gen_setup() is generic so it can be used to perform the bare xhci
setup even on non-pci based platform. The typedef for the function
pointer is moved into the headerfile

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb/xhci: refactor xhci_pci_setup()
Sebastian Andrzej Siewior [Fri, 23 Sep 2011 21:20:00 +0000 (14:20 -0700)]
usb/xhci: refactor xhci_pci_setup()

xhci_pci_setup() is split into three pieces:

- xhci_gen_setup()
  The major remaining of xhci_pci_setup() is now containing the generic
  part of the xhci setup. It allocates the xhci struct, setup
  hcs_params? and friends, performs xhci_halt(), xhci_init and so one.
  It also obtains the quirks via a callback
- xhci_pci_quirks()
  It checks the origin of the xhci core and sets core specific quirks.
- xhci_pci_setup()
  PCI specific setup functions. Besides calling xhci_gen_setup() with
  xhci_pci_quirks() as an argument it performs PCI specific setup like
  obtaining the address of sbrn via a PCI config space.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb/xhci: replace pci_*_consistent() with dma_*_coherent()
Sebastian Andrzej Siewior [Fri, 23 Sep 2011 21:19:59 +0000 (14:19 -0700)]
usb/xhci: replace pci_*_consistent() with dma_*_coherent()

pci_*_consistent() calls dma_*_coherent() with GFP_ATOMIC and requires
pci_dev struct. This is a preparion for later where we no longer have
the pci struct around.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb/xhci: hide MSI code behind PCI bars
Sebastian Andrzej Siewior [Fri, 23 Sep 2011 21:19:58 +0000 (14:19 -0700)]
usb/xhci: hide MSI code behind PCI bars

The MSI related fuctionality requires a few structs which are not
available if CONFIG_PCI is not enabled. This is a prepartion to allow
xhci be built without CONFIG_PCI set.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb/xhci: group MSI interrupt registration into its own function
Sebastian Andrzej Siewior [Fri, 23 Sep 2011 21:19:57 +0000 (14:19 -0700)]
usb/xhci: group MSI interrupt registration into its own function

This patch moves the complete MSI/MSI-X/Legacy dance into its own
function. There is however one difference: If the XHCI_BROKEN_MSI flag
is set then we don't free and register the irq, we simply return.
This is preparation for later PCI decouple.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb/host: introduce USB_ARCH_HAS_XHCI
Felipe Balbi [Fri, 23 Sep 2011 21:19:56 +0000 (14:19 -0700)]
usb/host: introduce USB_ARCH_HAS_XHCI

to make it look like OHCI and EHCI, we introduce
that symbol and USB_XHCI_HCD depend on that
instead of PCI.

[bigeasy@linutronix.de: wire up USB_ARCH_HAS_HCD]

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxHCI: fix debug message
Andiry Xu [Fri, 23 Sep 2011 21:19:55 +0000 (14:19 -0700)]
xHCI: fix debug message

Fix the debug message in xhci_address_device().

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxHCI: AMD isoc link TRB chain bit quirk
Andiry Xu [Fri, 23 Sep 2011 21:19:54 +0000 (14:19 -0700)]
xHCI: AMD isoc link TRB chain bit quirk

Setting the chain (CH) bit in the link TRB of isochronous transfer rings
is required by AMD 0.96 xHCI host controller to successfully transverse
multi-TRB TD that span through different memory segments.

When a Missed Service Error event occurs, if the chain bit is not set in
the link TRB and the host skips TDs which just across a link TRB, the
host may falsely recognize the link TRB as a normal TRB. You can see
this may cause big trouble - the host does not jump to the right address
which is pointed by the link TRB, but continue fetching the memory which
is after the link TRB address, which may not even belong to the host,
and the result cannot be predicted.

This causes some big problems. Without the former patch I sent: "xHCI:
prevent infinite loop when processing MSE event", the system may hang.
With that patch applied, system does not hang, but the host still access
wrong memory address and isoc transfer will fail. With this patch,
isochronous transfer works as expected.

This patch should be applied to kernels as old as 2.6.36, which was when
the first isochronous support was added for the xHCI host controller.

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Cc: stable@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousbcore: add sysfs support to xHCI usb2 hardware LPM
Andiry Xu [Fri, 23 Sep 2011 21:19:53 +0000 (14:19 -0700)]
usbcore: add sysfs support to xHCI usb2 hardware LPM

This patch adds sysfs support to xHCI usb2 hardware LPM, so developer can
enable and disable usb2 hardware LPM manually for test purpose.

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxHCI: set USB2 hardware LPM
Andiry Xu [Fri, 23 Sep 2011 21:19:52 +0000 (14:19 -0700)]
xHCI: set USB2 hardware LPM

If the device pass the USB2 software LPM and the host supports hardware
LPM, enable hardware LPM for the device to let the host decide when to
put the link into lower power state.

If hardware LPM is enabled for a port and driver wants to put it into
suspend, it must first disable hardware LPM, resume the port into U0,
and then suspend the port.

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxHCI: test USB2 software LPM
Andiry Xu [Fri, 23 Sep 2011 21:19:51 +0000 (14:19 -0700)]
xHCI: test USB2 software LPM

This patch tests USB2 software LPM for a USB2 LPM-capable device.

When a lpm-capable device is addressed, if the host also supports software
LPM, apply a test by putting the device into L1 state and resume it to see
if the device can do L1 suspend/resume successfully.

If the device fails to enter L1 or resume from L1 state, it may not
function normally and usbcore may disconnect and re-enumerate it. In this
case, store the device's Vid and Pid information, make sure the host will
not test LPM for it twice.

The test result is per device/host. Some devices claim to be lpm-capable,
but fail to enter L1 or resume. So the test is necessary.

The xHCI 1.0 errata has modified the USB2.0 LPM implementation. It redefines
the HIRD field to BESL, and adds another register Port Hardware LPM Control
(PORTHLPMC). However, this should not affect the LPM behavior on xHC which
does not implement 1.0 errata.

USB2.0 LPM errata defines a new bit BESL in the device's USB 2.0 extension
descriptor. If the device reports it uses BESL, driver should use BESL
instead of HIRD for it.

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxHCI: Check host USB2 LPM capability
Andiry Xu [Fri, 23 Sep 2011 21:19:51 +0000 (14:19 -0700)]
xHCI: Check host USB2 LPM capability

Check the host's USB2 LPM capability.

USB2 software LPM support is optional for xHCI 0.96 hosts. xHCI 1.0 hosts
should support software LPM, and may support hardware LPM.

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxHCI: Clear PLC for USB2 root hub ports
Andiry Xu [Fri, 23 Sep 2011 21:19:50 +0000 (14:19 -0700)]
xHCI: Clear PLC for USB2 root hub ports

When the link state changes, xHC will report a port status change event
and set the PORT_PLC bit, for both USB3 and USB2 root hub ports.

The PLC will be cleared by usbcore for USB3 root hub ports, but not for
USB2 ports, because they do not report USB_PORT_STAT_C_LINK_STATE in
wPortChange.

Clear it for USB2 root hub ports in handle_port_status().

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxHCI: test and clear RWC bit
Andiry Xu [Fri, 23 Sep 2011 21:19:49 +0000 (14:19 -0700)]
xHCI: test and clear RWC bit

Introduce xhci_test_and_clear_bit() to clear RWC bit in PORTSC register.

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxHCI: set link state
Andiry Xu [Fri, 23 Sep 2011 21:19:48 +0000 (14:19 -0700)]
xHCI: set link state

Introduce xhci_set_link_state() to remove redundant codes.

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousbcore: check device's LPM capability
Andiry Xu [Fri, 23 Sep 2011 21:19:48 +0000 (14:19 -0700)]
usbcore: check device's LPM capability

Check device's LPM capability by examining the bmAttibutes field of the
USB2.0 Extension Descriptor.

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousbcore: get BOS descriptor set
Andiry Xu [Fri, 23 Sep 2011 21:19:47 +0000 (14:19 -0700)]
usbcore: get BOS descriptor set

This commit gets BOS(Binary Device Object Store) descriptor set for Super
Speed devices and High Speed devices which support BOS descriptor.

BOS descriptor is used to report additional USB device-level capabilities
that are not reported via the Device descriptor. By getting BOS descriptor
set, driver can check device's device-level capability such as LPM
capability.

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: gadget: u_serial.c: fixed a brace coding style issue
Shaun Silk [Mon, 26 Sep 2011 01:26:43 +0000 (11:26 +1000)]
USB: gadget: u_serial.c: fixed a brace coding style issue

Signed-off-by: Shaun Silk <g0del@bigpond.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoEHCI : introduce a common ehci_setup
Matthieu CASTET [Sat, 2 Jul 2011 17:47:33 +0000 (19:47 +0200)]
EHCI : introduce a common ehci_setup

This allow to clean duplicated code in most of SOC driver.

Signed-off-by: Matthieu CASTET <castet.matthieu@free.fr>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: stable <stable@kernel.org> # fixes 3.1 build error
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousbmon vs. tcpdump: fix dropped packet count
Johannes Stezenbach [Thu, 8 Sep 2011 13:39:15 +0000 (15:39 +0200)]
usbmon vs. tcpdump: fix dropped packet count

Report the number of dropped packets instead of zero
when using the binary usbmon interface with tcpdump.

# tcpdump -i usbmon1 -w dump
tcpdump: listening on usbmon1, link-type USB_LINUX_MMAPPED (USB with padded Linux header), capture size 65535 bytes
^C2155 packets captured
2155 packets received by filter
1019 packets dropped by kernel

Signed-off-by: Johannes Stezenbach <js@sig21.net>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: gadget: udc-core: fix bug on soft_connect interface
Felipe Balbi [Thu, 8 Sep 2011 11:12:18 +0000 (14:12 +0300)]
usb: gadget: udc-core: fix bug on soft_connect interface

We should not be using dev_get_drvdata() because we
never call dev_set_drvdata(). Let's use container_of()
as all other sysfs attributes.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: pl2303: add id for SMART device
Eric Benoit [Sat, 24 Sep 2011 06:04:50 +0000 (02:04 -0400)]
USB: pl2303: add id for SMART device

Add vendor and product ID for the SMART USB to serial adapter. These
were meant to be used with their SMART Board whiteboards, but can be
re-purposed for other tasks. Tested and working (at at least 9600 bps).

Signed-off-by: Eric Benoit <eric@ecks.ca>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: qcserial: Add support for Sierra Wireless MC8355/Gobi 3000
Richard Hartmann [Tue, 20 Sep 2011 18:50:51 +0000 (20:50 +0200)]
USB: qcserial: Add support for Sierra Wireless MC8355/Gobi 3000

Simple patch to make qcserial recognize the USB id of the Sierra
Wireless MC8355 which is based on the Gobi 3000 chip.

Both UMTS and GPS work fine.

Signed-off-by: Richard Hartmann <richih.mailinglist@gmail.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxhci-mem.c: xhci_segment_free: No need for checking seg argument
Kautuk Consul [Mon, 19 Sep 2011 23:53:16 +0000 (16:53 -0700)]
xhci-mem.c: xhci_segment_free: No need for checking seg argument

The seg argument to xhci_segment_free is never passed as NULL, so
no need to check for this in xhci_segment_free.

Signed-off-by: Kautuk Consul <consul.kautuk@gmail.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxhci-mem.c: Check for ring->first_seg != NULL
Kautuk Consul [Mon, 19 Sep 2011 23:53:12 +0000 (16:53 -0700)]
xhci-mem.c: Check for ring->first_seg != NULL

There are 2 situations wherein the xhci_ring* might not get freed:
- When xhci_ring_alloc() -> xhci_segment_alloc() returns NULL and
  we goto the fail: label in xhci_ring_alloc. In this case, the ring
  will not get kfreed.
- When the num_segs argument to xhci_ring_alloc is passed as 0 and
  we try to free the rung after that.
  ( This doesn't really happen as of now in the code but we seem to
    be entertaining num_segs=0 in xhci_ring_alloc )

This should be backported to kernels as old as 2.6.31.

Signed-off-by: Kautuk Consul <consul.kautuk@gmail.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Cc: stable@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: When hot reset for USB3 fails, try warm reset.
Sarah Sharp [Wed, 14 Sep 2011 21:24:52 +0000 (14:24 -0700)]
USB: When hot reset for USB3 fails, try warm reset.

When a hot reset (standard USB port reset) fails on a USB 3.0 port, the
host controller transitions to the "Error" state.  It reports the port
link state as "Inactive", sets the link state change flag, and (if the
device disconnects) also reports the disconnect and connect change status.
It's also supposed to transition the link state to "RxDetect", but the NEC
µPD720200 xHCI host does not.

Unfortunately, Harald found that the combination of the NEC µPD720200 and
a LogiLink USB 3.0 to SATA adapter triggered this issue.  The USB core
would reset the device, the port would go into this error state, and the
device would never be enumerated.  This combination works under Windows,
but not under Linux.

When a hot reset fails on a USB 3.0 port, and the link state is reported
as Inactive, fall back to a warm port reset instead.  Harald confirms that
with a warm port reset (along with all the change bits being correctly
cleared), the USB 3.0 device will successfully enumerate.

Harald also had to add two other patches ("xhci: Set change bit when warm
reset change is set." and "usbcore: refine warm reset logic") to make this
setup work.  Since the warm reset refinement patch is not destined for the
stable kernels (it's too big), this patch should not be backported either.

This fixes https://bugzilla.kernel.org/show_bug.cgi?id=41752

Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Tested-by: Harald Brennich <harald.brennich@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxhci: USB 3.0 BW checking.
Sarah Sharp [Tue, 13 Sep 2011 23:41:13 +0000 (16:41 -0700)]
xhci: USB 3.0 BW checking.

The Intel Panther Point xHCI host tracks SuperSpeed endpoints in a
different way than USB 2.0/1.1 endpoints.  The bandwidth interval tables
are not used, and instead the bandwidth is calculated in a very simple
way.  Bandwidth for SuperSpeed endpoints is tracked individually in each
direction, since each direction has the full USB 3.0 bandwidth available.
10% of the bus bandwidth is reserved for non-periodic transfers.

This checking would be more complex if we had USB 3.0 LPM enabled, because
an additional latency for isochronous ping times need to be taken into
account.  However, we don't have USB 3.0 LPM support in Linux yet.

Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxhci: Fix mult base in endpoint bandwidth info.
Sarah Sharp [Tue, 13 Sep 2011 23:41:12 +0000 (16:41 -0700)]
xhci: Fix mult base in endpoint bandwidth info.

The "Mult" bits in the SuperSpeed Endpoint Companion Descriptor are
zero-based, and the xHCI host controller wants them to be zero-based in
the input context.  However, for the bandwidth math, we want them to be
one-based.  Fix this.

Fix the documentation about the endpoint bandwidth mult variable in the
xhci.h file, which says it is zero-based.  Also fix the documentation
about num_packets, which is also one-based, not zero-based.

Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousbcore: refine warm reset logic
Andiry Xu [Tue, 13 Sep 2011 23:41:11 +0000 (16:41 -0700)]
usbcore: refine warm reset logic

Current waiting time for warm(BH) reset in hub_port_warm_reset() is too short
for xHC host to complete the warm reset and report a BH reset change.

This patch increases the waiting time for warm reset and merges the function
into hub_port_reset(), so it can handle both cold reset and warm reset, and
factor out hub_port_finish_reset() to make the code looks cleaner.

This fixes the issue that driver fails to clear BH reset change and port is
"dead".

Signed-off-by: Andiry Xu <andiry.xu@amd.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb/xhci: ignore xhci version while checking for the link quirk
Sebastian Andrzej Siewior [Tue, 13 Sep 2011 23:41:10 +0000 (16:41 -0700)]
usb/xhci: ignore xhci version while checking for the link quirk

instead of reading the xhci interface version each time _even_ if the
quirk is not required, simply check if the quirk flag is set. This flag
is only set of the module parameter is set and here is where I moved the
version check to.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: Realtek cr: Fix driver freeze issue
edwin_rong [Fri, 16 Sep 2011 08:53:37 +0000 (16:53 +0800)]
USB: Realtek cr: Fix driver freeze issue

After auto-delink command is triggered, the CSW won't be sent back
to host side, in which scenario, the USB Mass Storage driver will
wait for the completion of the URB for MAX_SCHEDULE_TIMEOUT.

Signed-off-by: edwin_rong <edwin_rong@realsil.com.cn>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: add RESET_RESUME for webcams shown to be quirky
Oliver Neukum [Tue, 13 Sep 2011 06:42:21 +0000 (08:42 +0200)]
USB: add RESET_RESUME for webcams shown to be quirky

The new runtime PM code has shown that many webcams suffer
from a race condition that may crash them upon resume.
Runtime PM is especially prone to show the problem because
it retains power to the cameras at all times. However
system suspension may also crash the devices and retain
power to the devices.
The only way to solve this problem without races is in
usbcore with the RESET_RESUME quirk.

Signed-off-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: document ehci-hcd's "companion" sysfs attribute
Alan Stern [Wed, 14 Sep 2011 16:33:16 +0000 (12:33 -0400)]
USB: document ehci-hcd's "companion" sysfs attribute

This patch (as1484) adds documentation for ehci-hcd's "companion"
sysfs attribute, which was added to the kernel over four years ago but
never documented.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoMerge branch 'for-next' of git://gitorious.org/usb/usb into usb-next
Greg Kroah-Hartman [Sun, 18 Sep 2011 08:42:59 +0000 (01:42 -0700)]
Merge branch 'for-next' of git://gitorious.org/usb/usb into usb-next

* 'for-next' of git://gitorious.org/usb/usb: (47 commits)
  usb: musb: Enable DMA mode1 RX for transfers without short packets
  usb: musb: fix build breakage
  usb: gadget: audio: queue wLength-sized requests
  usb: gadget: audio: actually support both speeds
  usb: gadget: storage: make FSG_NUM_BUFFERS variable size
  USB: gadget: storage: remove alignment assumption
  usb: gadget: storage: adapt logic block size to bound block devices
  usb: dwc3: gadget: improve debug on link state change
  usb: dwc3: omap: set idle and standby modes
  usb: dwc3: ep0: introduce ep0_expect_in flag
  usb: dwc3: ep0: giveback requests on stall_and_restart
  usb: dwc3: gadget: drop the useless dma_sync_single* calls
  usb: dwc3: gadget: fix GCTL programming
  usb: dwc3: define ScaleDown macro helper
  usb: dwc3: Fix definition of DWC3_GCTL_U2RSTECN
  usb: dwc3: gadget: do not map/unmap ZLP transfers
  usb: dwc3: omap: fix IRQ handling
  usb: dwc3: omap: change IRQ name to dwc3-omap
  usb: dwc3: add module.h to dwc3-omap.c and core.c
  usb: dwc3: omap: distinguish between SW and HW modes
  ...

12 years agoUSB: irq: Remove IRQF_DISABLED
Yong Zhang [Wed, 7 Sep 2011 08:10:52 +0000 (16:10 +0800)]
USB: irq: Remove IRQF_DISABLED

This flag is a NOOP and can be removed now.

Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: ehci: remove the 1st wmb in qh_append_tds
Ming Lei [Mon, 5 Sep 2011 13:05:58 +0000 (21:05 +0800)]
usb: ehci: remove the 1st wmb in qh_append_tds

According to ehci spec 4.10.2, Advance Queue

If the fetched qTD has its Active bit set to a zero, the
host controller aborts the queue advance and follows the
queue head's horizontal pointer to the next schedule data
structure.

the 'qtd' will be linked into qh hardware queue after the line
below

*dummy = *qtd;

is executed and observed by EHCI HC, but EHCI HC won't have chance to
fetch the qtd descriptor pointed by 'qtd' in qh_append_tds until the
line below

dummy->hw_token = token; #set Active bit here

is executed by CPU and observed by EHCI HC.

There is already one 'wmb' to order writing to 'dummy'/'qtd' descriptors
and writing 'token' to 'dummy' descriptor(set Active bit), so the 1st
wmb is not needed and can be removed.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: ehci: fix comment for EHCI_SHRINK_JIFFIES
Ming Lei [Mon, 5 Sep 2011 13:05:57 +0000 (21:05 +0800)]
usb: ehci: fix comment for EHCI_SHRINK_JIFFIES

EHCI_SHRINK_JIFFIES should be 5ms, which was just used originally,
and not 200ms, so fix it.

Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: ehci: only prepare zero packet for out transfer if required
Ming Lei [Mon, 5 Sep 2011 13:05:56 +0000 (21:05 +0800)]
usb: ehci: only prepare zero packet for out transfer if required

Obviously, ZLP is only required for transfer of OUT direction,
so just take same policy with UHCI for ZLP packet.

Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: ehci: remove wmb in qh_update
Ming Lei [Mon, 5 Sep 2011 13:05:55 +0000 (21:05 +0800)]
usb: ehci: remove wmb in qh_update

qh_refresh is always called when the qh is idle and has not been
linked into hardware queue, so EHCI will not access overlay of
the qh at this time. Just before linking qh into hardware queue, there
has already one wmb to order writing qh descriptor and writing dma
address of the qh into hardware queue, so HC can always see
up-to-date qh descriptor once the qh is fetched with its dma address
by EHCI.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: cdc-acm: Owen SI-30 support
Denis Pershin [Sun, 4 Sep 2011 10:37:21 +0000 (17:37 +0700)]
usb: cdc-acm: Owen SI-30 support

here is the patch to support Owen SI-30 device.
This is a pulse counter controller.
http://www.owen.ru/en/catalog/93788515

usb-drivers output:
T:  Bus=01 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#=  4 Spd=12  MxCh= 0
D:  Ver= 2.00 Cls=02(commc) Sub=00 Prot=00 MxPS= 8 #Cfgs=  1
P:  Vendor=03eb ProdID=0030 Rev=01.01
C:  #Ifs= 2 Cfg#= 1 Atr=c0 MxPwr=0mA
I:  If#= 0 Alt= 0 #EPs= 1 Cls=02(commc) Sub=02 Prot=00 Driver=cdc_acm
I:  If#= 1 Alt= 0 #EPs= 2 Cls=0a(data ) Sub=00 Prot=00 Driver=cdc_acm

This patch is installed on my home system which receives data from this
controller connected to cold water counter.

Signed-off-by: Denis Pershin <dyp@perchine.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb-storage: fix realtek cr configuration
Vincent Palatin [Thu, 1 Sep 2011 21:05:15 +0000 (14:05 -0700)]
usb-storage: fix realtek cr configuration

A typo in the configuration variable name prevents from activating the
USB autosuspend on the device.

Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: g_printer: fix bug in unregistration
Fabian Godehardt [Thu, 1 Sep 2011 12:15:46 +0000 (14:15 +0200)]
USB: g_printer: fix bug in unregistration

The allocated chardevice region range is only 1 device but on
unregister it currently tries to deregister 2.

Found this while doing a insmod/rmmod/insmod/rm... of the module
which seemed to eat major numbers.

Signed-off-by: Fabian Godehardt <fg@emlix.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agos3c-hsudc: implement vbus_draw hook
Heiko Stübner [Sun, 4 Sep 2011 19:56:02 +0000 (21:56 +0200)]
s3c-hsudc: implement vbus_draw hook

When a transceiver is available use otg_set_power to submit
the target current to it.

Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb: Provide usb_speed_string() function
Michal Nazarewicz [Tue, 30 Aug 2011 15:11:19 +0000 (17:11 +0200)]
usb: Provide usb_speed_string() function

In a few places in the kernel, the code prints
a human-readable USB device speed (eg. "high speed").
This involves a switch statement sometimes wrapped
around in ({ ... }) block leading to code repetition.

To mitigate this issue, this commit introduces
usb_speed_string() function, which returns
a human-readable name of provided speed.

It also changes a few places switch was used to use
this new function.  This changes a bit the way the
speed is printed in few instances at the same time
standardising it.

Signed-off-by: Michal Nazarewicz <mina86@mina86.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: option: add various ZTE device network interfaces to the blacklist
Dan Williams [Tue, 13 Sep 2011 18:52:52 +0000 (13:52 -0500)]
USB: option: add various ZTE device network interfaces to the blacklist

IDs found in the Windows driver's ZTEusbnet.inf file from the
ZTE MF100 drivers (O2 UK).  Also fixes the ZTE MF626 device
since it really is distinct from the 4G Systems stick and
apparently needs the net interface blacklisted too, while
there's no indication (yet) that the 4G Systems stick does.

Signed-off-by: Dan Williams <dcbw@redhat.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: option: add ZTE product 0x0037 to sendsetup blacklist
Dan Williams [Tue, 13 Sep 2011 18:51:45 +0000 (13:51 -0500)]
USB: option: add ZTE product 0x0037 to sendsetup blacklist

Signed-off-by: Dan Williams <dcbw@redhat.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: option: convert Huawei K3765, K4505, K4605 reservered interface to blacklist
Dan Williams [Tue, 13 Sep 2011 18:51:13 +0000 (13:51 -0500)]
USB: option: convert Huawei K3765, K4505, K4605 reservered interface to blacklist

That's what the blacklist is for...

Signed-off-by: Dan Williams <dcbw@redhat.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: option: convert interface blacklisting to bitfields
Dan Williams [Tue, 13 Sep 2011 18:49:41 +0000 (13:49 -0500)]
USB: option: convert interface blacklisting to bitfields

It's cleaner than the array stuff, and we're about to add a bunch
more blacklist entries.  Second, there are devices that need both
the sendsetup and the reserved interface blacklists, which the
current code can't accommodate.

Signed-off-by: Dan Williams <dcbw@redhat.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agousb gadget: clean up FSF boilerplate text
Klaus Schwarzkopf [Fri, 9 Sep 2011 14:10:44 +0000 (16:10 +0200)]
usb gadget: clean up FSF boilerplate text

remove the following two paragraphs as they are not needed:

This program is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public
License for more details.

You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc.,59
Temple Place - Suite 330, Boston, MA  02111-1307, USA.

Signed-off-by: Klaus Schwarzkopf <schwarzkopf@sensortherm.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agomusb_gadget: Fix for spurious interrupts on endpoint zero.
Hans Petter Selasky [Fri, 2 Sep 2011 06:17:17 +0000 (08:17 +0200)]
musb_gadget: Fix for spurious interrupts on endpoint zero.

There is a multi-year old bug in the MUSB hardware which is not documented.
It causes spurious interrupts and have various symptoms, like endless
"SetupEnd came in a wrong ep0stage" messages. The fix is taken from the
FreeBSD's musb driver.

How to reproduce:
For example issue clear-stall on a couple of endpoints very fast,
like one request per 125us. After a while the bug triggers and the
musb-chip becomes unusable until next re-enumeration.

Signed-off-by: Hans Petter Selasky <hps@bitfrost.no>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: for usb_autopm_get_interface_async -EINPROGRESS is not an error
Jim Wylder [Wed, 7 Sep 2011 02:07:20 +0000 (21:07 -0500)]
USB: for usb_autopm_get_interface_async -EINPROGRESS is not an error

A return value of -EINPROGRESS from pm_runtime_get indicates that
the device is already resuming due to a previous call.  Internally,
usb_autopm_get_interface_async doesn't treat this as an error and
increments the usage count, but passes the error status along
to the caller.  The logical assumption of the caller is that
any negative return value reflects the device not resuming
and the pm_usage_cnt not being incremented.  Since the usage count
is being incremented and the device is resuming, return success (0)
instead.

Signed-off-by: James Wylder <james.wylder@motorola.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoUSB: storage: Use normalized sense when emulating autosense
Luben Tuikov [Thu, 11 Nov 2010 23:43:11 +0000 (15:43 -0800)]
USB: storage: Use normalized sense when emulating autosense

This patch solves two things:
1) Enables autosense emulation code to correctly
interpret descriptor format sense data, and
2) Fixes a bug whereby the autosense emulation
code would overwrite descriptor format sense data
with SENSE KEY HARDWARE ERROR in fixed format, to
incorrectly look like this:

Oct 21 14:11:07 localhost kernel: sd 7:0:0:0: [sdc]  Sense Key : Recovered Error [current] [descriptor]
Oct 21 14:11:07 localhost kernel: Descriptor sense data with sense descriptors (in hex):
Oct 21 14:11:07 localhost kernel:        72 01 04 1d 00 00 00 0e 09 0c 00 00 00 00 00 00
Oct 21 14:11:07 localhost kernel:        00 4f 00 c2 00 50
Oct 21 14:11:07 localhost kernel: sd 7:0:0:0: [sdc]  ASC=0x4 ASCQ=0x1d

Signed-off-by: Luben Tuikov <ltuikov@yahoo.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Matthew Dharm <mdharm-usb@one-eyed-alien.net>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoxhci: Redundant check in xhci_check_args for xhci->devs
sifram.rajas@gmail.com [Fri, 2 Sep 2011 18:06:00 +0000 (11:06 -0700)]
xhci: Redundant check in xhci_check_args for xhci->devs

The xhci_hcd->devs is an array of pointers rather than pointer to pointer.
Hence this check is not required.

Signed-off-by: Sifram Rajas <Sifram Rajas sifram.rajas@gmail.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>