[media] cx18: Fix warnings introduced during cleanup

I misused the ktime API, and failed to remove some traces of the
in-kernel format conversion. Fix these, so the the driver builds
without warnings.

Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Make the API public

Move the public API definitions to include/linux/uvcvideo.h and bump the
version number to 1.1.0. Compatibility with the old API is kept,
application can still be compiled against the private header and will
not break.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Return -ERANGE when userspace sets an unsupported menu entry

Instead of passing the value down to the device and getting an error
back (or worse, crashing the firmware), return -ERANGE when the
requested menu entry is not supported.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Don't report unsupported menu entries

Supported menu entries are reported by the device in response to the
GET_RES query. Use the information to return -EINVAL to userspace for
unsupported values when enumerating menu entries.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Override wrong bandwidth value for Hercules Classic Silver

Set the FIX_BANDWIDTH quirk for the device.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Add support for JMicron USB2.0 XGA WebCam

The camera requires the PROBE_MINMAX quirk. Add a corresponding entry
in the device IDs list

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Add support for V4L2_PIX_FMT_RGB565

The TomTom navigation system used in the Sony XNV 660BT and 770BT
reports an RGB565 (RGBP) format, support it in the driver.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Add driver documentation

This adds detailed documentation about the driver's extension unit control
features. Part of it has been adopted from the dynctrl.txt that used to be
available in the driver's previous Subversion repository.

The documentation has been rewritten to match the current implementation, in
particular to capture the new XU control support including UVCIOC_CTRL_QUERY.

Signed-off-by: Martin Rubli <martin_rubli@logitech.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Rename UVC_CONTROL_* flags to UVC_CTRL_FLAG_*

This makes the public driver API more uniform, in preparation of moving
uvcvideo.h to include/linux. Keep the old names for backward
compatibility with existing applications.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Deprecate UVCIOC_CTRL_{ADD,MAP_OLD,GET,SET}

Those ioctls are deprecated, list them in the features removal schedule
for 2.6.42.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] uvcvideo: Add UVCIOC_CTRL_QUERY ioctl

This ioctl extends UVCIOC_CTRL_GET/SET by not only allowing to get/set
XU controls but to also send arbitrary UVC commands to XU controls,
namely GET_CUR, SET_CUR, GET_MIN, GET_MAX, GET_RES, GET_LEN, GET_INFO
and GET_DEF. This is required for applications to work with XU controls,
so that they can properly query the size and allocate the necessary
buffers.

Signed-off-by: Martin Rubli <martin_rubli@logitech.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx18: Bump driver version to 1.5.0

To simplify maintainer support of this driver, bump the version to
1.5.0 - this will be the first version that is expected to support
mmap() for raw video frames.

Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk>
Acked-by: Andy Walls <awalls@md.metrocast.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx18: Clean up mmap() support for raw YUV

The initial version of this patch (commit
d5976931639176bb6777755d96b9f8d959f79e9e) had some issues:

 * It didn't correctly calculate the size of the YUV buffer for 4:2:2,
   resulting in capture sometimes being offset by 1/3rd of a picture.

 * There were a lot of variables duplicating information the driver
   already knew, which have been removed.

 * There was an in-kernel format conversion - libv4l can do this one,
   and is the right place to do format conversions anyway.

 * Some magic numbers weren't properly explained.

Fix all these issues, leaving just the move from videobuf to videobuf2
to do.

Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk>
Acked-by: Andy Walls <awalls@md.metrocast.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cpia2: fix typo in variable initialisation

Currently 'fh' initialises to whatever happens to be on stack. This
looks like a typo and this patch fixes that.

Signed-off-by: Mariusz Kozlowski <mk@lab.zgora.pl>
Acked-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cxd2820r: switch automatically between DVB-T and DVB-T2

Remove old DVB-T2 freq module param and detect DVB-T/T2 automatically.

Implementation is trial and error, if DVB-T does not lock try DVB-T2
and vice versa. That's done by replacing normal DVBFE_ALGO_SW with
DVBFE_ALGO_CUSTOM which gives better control for tuning process.

DVB-C still uses normal software ZigZag, DVBFE_ALGO_SW.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cxd2820r: whitespace fix

Make checkpatch.pl happy by fixing whitespaces introduced by
commit 79e8b8e3b8cbf67130247a3f6d25732373fe2f34

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] em28xx: Disable audio for EM28174

Signed-off-by: Steve Kerrison <steve@stevekerrison.com>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cxd2820r: make C, T, T2 and core components as linked objects

Signed-off-by: Steve Kerrison <steve@stevekerrison.com>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] Add support for PCTV nanoStick T2 290e [2013:024f]

Supports DVB-T/T2/C, USB ID: 2013:024f.
Empia EM28174, Sony CXD2820R and NXP TDA18271HD/C2.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] Sony CXD2820R DVB-T/T2/C demodulator driver

It is very first DVB-T2 Linux driver!

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] em28xx: add support for EM28174 chip

EM28174 is very similar as already supported EM2874.
I am not sure what are differences, but it could be analog support.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] em28xx: Multi Frontend (MFE) support

Register multiple FEs for same adapter. After that it is
possible to register two FEs for same adapter. For example
one for DVB-T and one for DVB-C.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] tda18271: add DVB-C support

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx24116: make FW DL split more readable

Change firmware download split, which I introduced few patch earlier,
a little bit to make it more readable as requested [1].

Anyhow, for some reason this seems to increase compiled binary size
52 bytes, on my AMD64 box, which is rather much for so small change.

[1] http://www.spinics.net/lists/linux-media/msg31968.html

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] anysee: add support for Anysee E7 S2

It is DVB-S/S2 USB box.

* DNBU10512IST NIM
 * STV0903 demod
 * STV6110 tuner

* Intersil ISL6423 LNB controller

Signed-off-by: Antti Palosaari <crope@iki.fi>
Cc: info@anysee.com
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] anysee: add support for Anysee E30 S2 Plus

It is USB DVB-S/S2 box.

Conexant cx24116/cx24118 demod tuner combo
Intersil ISL6423 LNB controller

Signed-off-by: Antti Palosaari <crope@iki.fi>
Cc: info@anysee.com
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx24116: add config option to split firmware download

It is very rare I2C adapter hardware which can provide 32kB I2C write
as one write. Add .i2c_wr_max option to set desired max packet size.
Split transaction to smaller pieces according to that option.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Cc: Steven Toth <stoth@hauppauge.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] media: fsl_viu: fix bug in streamon routine

Currently video capturing using streaming I/O method
doesn't work if capturing to overlay buffer took place
before.

When enabling the stream we have to check the overlay
enable driver flag and reset it so that the interrupt
handler won't execute the overlay interrupt path after
enabling DMA in streamon routine. Otherwise the capture
interrupt won't be handled correctly causing non working
VIDIOC_DQBUF ioctl.

Signed-off-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] fsl_viu: add VIDIOC_OVERLAY ioctl

Currently the driver enables overlay when running
VIDIOC_S_FMT ioctl with fmt type V4L2_BUF_TYPE_VIDEO_OVERLAY.
Actually, this is wrong. Add proper VIDIOC_OVERLAY support
instead of using VIDIOC_S_FMT for overlay enable.

Signed-off-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] gspca/stv06xx_pb0100: Don't potentially deref NULL in pb0100_start()

usb_altnum_to_altsetting() may return NULL. If it does we'll dereference a
NULL pointer in
drivers/media/video/gspca/stv06xx/stv06xx_pb0100.c::pb0100_start().
As far as I can tell there's not really anything more sensible than
-ENODEV that we can return in that situation, but I'm not at all intimate
with this code so I'd like a bit of review/comments on this before it's
applied.
Anyway, here's a proposed patch.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Cc: Jean-Francois Moine <moinejf@free.fr>
Cc: Hans de Goede <hdegoede@redhat.com>
Cc: Erik Andren <erik.andren@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] rc-core: use ir_raw_event_store_with_filter in winbond-cir

Using ir_raw_event_store_with_filter() saves about 20 lines of code.

Signed-off-by: David Härdeman <david@hardeman.nu>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] rc-core: add trailing silence in rc-loopback tx

If an IR command is sent (using the LIRC userspace) to rc-loopback
which doesn't include a trailing space, the result is that the message
won't be completely decoded. In addition, "leftovers" from a previous
transmission can be left until the next one. Fix this by faking a long
silence after the end of TX data.

Signed-off-by: David Härdeman <david@hardeman.nu>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] rc-core: add TX support to the winbond-cir driver

This patch adds preliminary IR TX capabilities to the
winbond-cir driver.

Signed-off-by: David Härdeman <david@hardeman.nu>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] rc-core: int to bool conversion for winbond-cir

Using bool instead of an int helps readability a bit.

Signed-off-by: David Härdeman <david@hardeman.nu>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] v4l: Add mt9v032 sensor driver

The MT9V032 is a parallel wide VGA sensor from Aptina (formerly Micron)
controlled through I2C.

The driver creates a V4L2 subdevice. It currently supports binning and
cropping, and the gain, auto gain, exposure, auto exposure and test
pattern controls.

Signed-off-by: Detlev Casanova <detlev.casanova@gmail.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] STV0299 Register 02 on Opera1/Bsru6/z0194a/mantis_vp1033

Bits 4 and 5 on register 02 should always be set to 1.

Opera1/Bsru6/z0194a/mantis_vp1033

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Acked-by: Oliver Endriss <o.endriss@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] STV0299 incorrect standby setting issues register 02 (MCR)

Issues with Register 02 causing spurious channel locking from standby.
Should have always bits 4 & 5 written to 1.
Lower nibble not used in any current driver. Usage if necessary can be
applied through initab to mcr_reg. stv0299 not out of standby before
writing inittab.

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Acked-by: Oliver Endriss <o.endriss@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] fsl-viu: replace .ioctl by .unlocked_ioctl

Use the core-assisted locking in fsl-viu driver and switch
to .unlocked_ioctl.

Signed-off-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx88: don't use atomic_t for core->users

users is always read or written with core->lock held.  A plain int is
simpler and faster.

Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx88: don't use atomic_t for core->mpeg_users

mpeg_users is always read or written with core->lock held except
in mpeg_release (where it looks like a bug).  A plain int is simpler
and faster.

Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx88: gracefully reject attempts to use unregistered cx88-blackbird driver

It should not be possible to enter mpeg_open and acquire core->lock
without the blackbird driver being registered, so just error out if it
is not.  This makes the code more readable and should prevent the bug
fixed by the patch "hold device lock during sub-driver initialization"
from resurfacing again.

Similarly, if we enter mpeg_release and acquire core->lock then either
the blackbird driver is registered (since open files keep it loaded)
or the sysadmin forced the driver's removal.  In the latter case the
state will be inconsistent and this is worth a loud warning.

Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx88: protect cx8802_devlist with a mutex

Add and use a mutex to protect the cx88-mpeg device list.  Previously
the BKL prevented races.

Based on work by Ben Hutchings <ben@decadent.org.uk>.

Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx88: hold device lock during sub-driver initialization

cx8802_blackbird_probe makes a device node for the mpeg sub-device
before it has been added to dev->drvlist.  If the device is opened
during that time, the open succeeds but request_acquire cannot be
called, so the reference count remains zero.  Later, when the device
is closed, the reference count becomes negative --- uh oh.

Close the race by holding core->lock during probe and not releasing
until the device is in drvlist and initialization finished.
Previously the BKL prevented this race.

Reported-by: Andreas Huber <hobrom@gmx.at>
Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Cc: stable@kernel.org
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx88: fix locking of sub-driver operations

The BKL conversion of this driver seems to have gone wrong.
Loading the cx88-blackbird driver deadlocks.

The cause: mpeg_ops::open in the cx2388x blackbird driver acquires the
device lock and calls the sub-driver's request_acquire, which tries to
acquire the lock again.  Fix it by clarifying the semantics of
request_acquire, request_release, advise_acquire, and advise_release:
now all will rely on the caller to acquire the device lock.

Based on work by Ben Hutchings <ben@decadent.org.uk>.

Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=31962

Reported-by: Andi Huber <hobrom@gmx.at>
Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Cc: stable@kernel.org
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx88: protect per-device driver list with device lock

The BKL conversion of this driver seems to have gone wrong.  Various
uses of the sub-device and driver lists appear to be subject to race
conditions.

In particular, some functions access drvlist without a relevant lock
held, which will race against removal of drivers.  Let's start with
that --- clean up by consistently protecting dev->drvlist with
dev->core->lock, noting driver functions that require the device lock
to be held or not to be held.

After this patch, there are still some races --- e.g.,
cx8802_blackbird_remove can run between the time the blackbird driver
is acquired and the time it is used in mpeg_release, and there's a
similar race in cx88_dvb_bus_ctrl.  Later patches will address the
remaining known races and the deadlock noticed by Andi.  This patch
just makes the semantics clearer in preparation for those later
changes.

Based on work by Ben Hutchings <ben@decadent.org.uk>.

Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Cc: stable@kernel.org
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] Fix panic on loading earth-pt1

Signed-off-by: HIRANO Takahito <hiranotaka@zng.info>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] usbvision: remove broken testpattern

Enabling force_testpattern module parameter in usbvision causes kernel panic.
Things like that does not belong to the kernel anyway so the fix is easy.

Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] usbvision: add Nogatech USB MicroCam

Add Nogatech USB MicroCam PAL (NV3001P) and NTSC (NV3000N) support to
usbvision driver.
PAL version is tested, NTSC untested.
Data captured using usbsnoop, init_values are listed in the INF file along
with image dimensions, offsets and frame rates.

Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] lmedm04: PID filtering changes

Improve PID filtering and program register 20 correctly.
 Make sure stream_on message is sent if streaming is off, otherwise
 PIDs are not registered.
 Move mutex outside lme2510_enable_pid.

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] lmedm04: don't write to buffer without a mutex

usb_buffer not inside mutex lock, waiting caller can alter buffer.
 Static added to lme2510_exit and lme2510_exit_int.

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] IX2505V Keep I2C gate control alive

Gate could close after first I2C message. On stv0288 it does.
 Keep 2nd and 3rd message I2C gate control alive.
 Remove unnecessary gate closing in this module.

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] Media, DVB, Siano, smsusb: Avoid static analysis report about 'use after free'

In drivers/media/dvb/siano/smsusb.c we have this code:
 ...
               kfree(dev);

               sms_info("device %p destroyed", dev);
 ...

at least one static analysis tool (Coverity Prevent) complains about this
as a use-after-free bug.
While it's true that we do use the pointer variable after freeing it, the
only use is to print the value of the pointer, so there's not actually any
problem here. But still, silencing the complaint is trivial by just moving
the kfree() call below the sms_info(), so why not just do it?. It doesn't
change the workings of the code in any way, but it makes the tool shut up.
The patch below also removes a rather pointless blank line.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] tm6000: add detection based on eeprom name

On some situations, it is desired to use eeprom data to detect
the board type. This patch adds a logic for it and fixes 2 detection
issues:
1) 10Moons UT-821 uses a generic Trident ID. Other boards
also share the same ID. So, better to use an alternative way for
it;
2) Sometimes, HVR-900H is loaded with the default Trident
ID. This seems to be some hardware bug or race condition.

The new logic will only be enabled if the device is detected as
having a generic ID.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] dvb-usb: don't return error if stream stop

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] dvb-usb return device errors to demuxer

Return device errors to demuxer from on/off streamming and
 pid filtering.

Please test this patch with all dvb-usb devices.

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] lmedm04: Do not unlock mutex if mutex_lock_interruptible failed

There are a couple of places where mutex_unlock() is called even
if mutex_lock_interruptible() failed. The patch fixes the issue.

Found by Linux Driver Verification project (linuxtesting.org).

Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx231xx: add support for Kworld

[mchehab@redhat.com: avoided board renumberation, removed an unused #define
 and re-used the existing mb86a20s dvb attach code]
Signed-off-by: Márcio A Alves <froooozen@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] Speed up DVB TS stream delivery from DMA buffer into dvb-core's buffer

Avoid unnecessary DVB TS 188 sized packet copying from DMA buffer into stack.
Backtrack one 188 sized packet just after some garbage bytes when possible.
This obsoletes patch https://patchwork.kernel.org/patch/118147/

Signed-off-by: Marko Ristola <marko.ristola@kolumbus.fi>
Acked-by: Andreas Oberritter <obi@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx23885: Don't leak firmware in cx23885_card_setup()

We leak the memory allocated to 'fw' (the firmware) when the variable goes
out of scope.
Fix the leak by calling release_firmware(fw) before 'fw' goes out of
scope.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] DVB, DiB9000: Fix leak in dib9000_attach()

If the second memory allocation in dib9000_attach() fails, we'll leak the
memory allocated by the first.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Cc: Patrick Boettcher <patrick.boettcher@dibcom.fr>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx18: mmap() support for raw YUV video capture

Add support for mmap method streaming of raw YUV video on cx18-based
hardware, in addition to the existing support for read() streaming of
raw YUV and MPEG-2 encoded video.

[simon.farnsworth@onelan.co.uk: I forward-ported this from Steven's original work,
 done under contract to ONELAN. The original code is at
 http://www.kernellabs.com/hg/~stoth/cx18-videobuf]

Signed-off-by: Steven Toth <stoth@kernellabs.com>
Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] tm6000: add CARDLIST

add CARDLIST

Signed-off-by: Stefan Ringel <stefan.ringel@arcor.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] s2255drv: jpeg enable module parameter

Adding jpeg enable module parameter.

Signed-off-by: Dean Anderson <linux-dev@sensoray.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] s2255drv: adding MJPEG format

adding MJPEG format

Signed-off-by: Dean Anderson <linux-dev@sensoray.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] wl128x: Remove unused NO_OF_ENTRIES_IN_ARRAY macro

Signed-off-by: Thiago Farina <tfransosi@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] DM04/QQBOX v1.84 added PID filter

A greatly simplified version of the PID Filter now added
back to the Driver.

The driver allows for the PID filter to be turned off.

applied after patch 683781.

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] STV0288 Register 42 - Incorrect settings

Register 42 bits 2,3,6 and 7 should be set to 0.
 This gives difficult locking on some channels and may be compensated
 for by other methods.

This affects any driver using the stv0288 frontend on the default
or earda inittab.

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] DM04/QQBOX stv0288 register 42 - incorrect setting

 stv0288 Register 42 bits 6 & 7 should be set to 0.
 This is causing intermittent lock, the dvb-usb-lmedm04 driver uses
 register 50 (auto fine mode) to correct for this, this register is
 now returned to its default setting.

Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] s2255drv: atomic submit urb in completion handler

An usb_submit_urb should be atomic in a completion handler. This fixes
"BUG: scheduling while atomic" messages.

Signed-off-by: Pete Eberlein <pete@sensoray.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] budget-ci: Add support for TT S-1500 with BSBE1-D01A tuner

Add support for TT S-1500 with BSBE1-D01A tuner.

Signed-off-by: Oliver Endriss <o.endriss@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] cx24116.c - fix for wrong parameter description

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] pwc: Handle V4L2_CTRL_FLAG_NEXT_CTRL in queryctrl

Signed-off-by: Jean-François Moine <moinejf@free.fr>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] dib0700: fix possible NULL pointer dereference

Seems like 'adap->fe' test for NULL was meant to be before we dereference
that pointer.

Signed-off-by: Mariusz Kozlowski <mk@lab.zgora.pl>
Acked-by: Patrick Boettcher <patrick.boettcher@dibcom.fr>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] pvrusb2: delete generic_standards_cnt

The generic_standards_cnt define is only used in one place and it's
more readable to just call ARRAY_SIZE(generic_standards) directly.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] pvrusb2: check for allocation failures

This function returns NULL on failure so lets do that if kzalloc()
fails.  There is a separate problem that the caller for this function
doesn't check for errors...

Signed-off-by: Dan Carpenter <error27@gmail.com>
Acked-By: Mike Isely <isely@pobox.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] saa7134: support MagicPro ProHDTV Pro2 Hybrid DMB-TH PCI card

This card has a TD18271 silicon tuner, and uses TDA8290 and LGS8G75 to
demodulate analog and digital broadcast respectively.  GPIO configurations
were derived using DScaler regspy.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] zoran: Drop unused module parameters encoder and decoder

The ability to force the encoder or decoder chip was broken by commit
0ab6e1c38d80ab586e3a1ca9e71844131d9f51dc in February 2009. As nobody
complained for over 2 years, I take it that these parameters were no
longer used so we can simply drop them.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] lmedm04: get rid of on-stack dma buffers

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Tested-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Florian Mickler <florian@mickler.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] au6610: get rid of on-stack dma buffer

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Signed-off-by: Florian Mickler <florian@mickler.org>
Acked-by: Antti Palosaari <crope@iki.fi>
Reviewed-by: Antti Palosaari <crope@iki.fi>
Tested-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] ce6230: get rid of on-stack dma buffer

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Signed-off-by: Florian Mickler <florian@mickler.org>
Acked-by: Antti Palosaari <crope@iki.fi>
Reviewed-by: Antti Palosaari <crope@iki.fi>
Tested-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] ec168: get rid of on-stack dma buffers

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Signed-off-by: Florian Mickler <florian@mickler.org>
Acked-by: Antti Palosaari <crope@iki.fi>
Reviewed-by: Antti Palosaari <crope@iki.fi>
Tested-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp702x: use preallocated buffer in the frontend

Note: This change is tested to compile only as I don't have the
hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp702x: use preallocated buffer in vp702x_usb_inout_cmd

If we need a bigger buffer, we reallocte a new buffer and free the old
one.

Note: This change is tested to compile only as I don't have the
hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp702x: use preallocated buffer

Note: This change is tested to compile only as I don't have the
hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp702x: fix locking of usb operations

Otherwise it is not obvious that vp702x_usb_in_op or vp702x_usb_out_op
will not interfere with any vp702x_usb_inout_op.

Note: This change is tested to compile only, as I don't have the
hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp702x: get rid of on-stack dma buffers

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Note: This change is tested to compile only, as I don't have the hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp702x: remove unused variable

struct vp702x_device_state.power_state is nowhere referenced.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp702x: preallocate memory on device probe

This sets up a buffer and a mutex protecting that buffer in
the struct vp702x_device_state.

The definition of struct vp702x_device_state is moved into the header
in order to use the buffer also in the frontend.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp702x: rename struct vp702x_state -> vp702x_adapter_state

We need a state struct for the dvb_usb_device.
In order to reduce confusion we rename the vp702x_state struct.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp702x: cleanup: whitespace and indentation

Some whitespace, one linebreak and one unneded variable
initialization...

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] opera1: get rid of on-stack dma buffer

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Note: This change is tested to compile only, as I don't have the hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] m920x: get rid of on-stack dma buffers

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Note: This change is tested to compile only, as I don't have the hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] dw2102: get rid of on-stack dma buffer

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Note: This change is tested to compile only, as I don't have the hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Igor M. Liplianin <liplianin@tut.by>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] friio: get rid of on-stack dma buffers

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Note: This change is tested to compile only, as I don't have the hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Akihiro Tsukada <tskd2@yahoo.co.jp>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] vp7045: get rid of on-stack dma buffers

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] a800: get rid of on-stack dma buffers

usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.

In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack

Note: This change is tested to compile only, as I don't have the hardware.

Signed-off-by: Florian Mickler <florian@mickler.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] dib0700: remove unused variable

This variable is never used.

Signed-off-by: Florian Mickler <florian@mickler.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] DiB0700: get rid of on-stack dma buffers

This patch removes the on-stack buffers for USB DMA transfers.

This is an alternative version of the patch discussed by Florian here:
http://thread.gmane.org/gmane.linux.kernel/1115695/

Signed-off-by: Olivier Grenie <olivier.grenie@dibcom.fr>
Signed-off-by: Patrick Boettcher <patrick.boettcher@dibcom.fr>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] lmedm04: correct indentation

This should not change anything except whitespace.

Signed-off-by: Florian Mickler <florian@mickler.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

[media] ignore Documentation/DocBook/media/

It only contains generated files

Signed-off-by: Michael Jones <michael.jones@matrix-vision.de>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>