V4L/DVB (6654): mt352: support oversampled IF input

Rework the input frequency calculation so that it produces the right values
when the ADC oversamples the IF input.

This means MT352 devices can now process a near-zero IF (according to the,
specs 4.57MHz is supported with the default crystal).

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6653): Add support for the DViCO FusionHDTV NANO2 w/ZL10353 and firmware

Add support for the DViCO FusionHDTV DVB-T NANO with zl10353 demodulator and
firmware in ROM on the device.

Again, this is based on the great work of Mike Krufky with my modifications
to use the in-tree XC2028 driver.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6652): xc2028: try non-8MHZ init1 firmware

When loading init1 firmware, there may not be an 8MHz specific version.
Load the non-8MHz version if it exists.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6651): xc2028: mask off type correctly when searching for standard-specific types

When searching for standard-specific analog firmware, only certain
type bits are valid, much like for DTV.  Mask them off when finding
the firmware to load.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6650): xc2028: base firmwares should have std0

When loading BASE firmware, we must use std = 0.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6649): Add support for the DViCO FusionHDTV Dual Digital 4

Add support for DViCO's Dual Digital 4 with xc3028 tuner, zl10353 DVB-T
demodulator and a new-style I2C IR remote control receiver.

This would not have been possible without the work of and advice from
Mike Krufky, who originally got the Dual Digital 4 and second-gen DVB-T
NANO devices working with the out-of-tree XC3028 driver.

I converted it to use the in-tree XC3028 driver (after making it suitable
for our use), and added the IR remote control support based on his advice.

NB: a firmware package is required to use this device.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6648): xc2028: add sleep hook

Add sleep method to enable putting the tuner into standby mode.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6647): xc2028: retry firmware load if tuner does not respond

In practice, the tuner occasionally fails to respond correctly after a
firmware load.  Retry the firmware load if the firmware/hardware version
we read back from the tuner after programming does not match what we
expect.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6646): xc2028: rework firmware (re)loading process

Define a list of valid "firmware types" for each combination of BASE,
DTV and SCODEs.  By masking the appropriate firmware bits off we can
just use one "type" for the firmware searching and also flag when we
are looking for a BASE, DTV or SCODE type firmware.  This makes it
much easier to track if we need to change device modes or flash an
individual firmware part.

Add a structure to remember what firmware properties we have.  This
contains the currently loaded/wanted  base firmware (type), video std
(id), video std requested (std_req), scode file and number in use.

Incorporate said structure into the tuner private data.

When checking whether the current firmware needs to be reloaded, first
figure out exactly what "type" of firmware we want (base, std and
scode), and then proceed to load the appropriate matching base,
std-specific and scode records iff there are any changes required.
This removes guesswork from the process because we no longer need to
individually code a check for every tuning parameter's interactions.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6645): xc2028: allow selection of D2633 firmware

Add a bit to select D2633 DTV firmware to struct xc2028_ctrl, so that it can
be enabled via .set_config.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6644): xc2028: use correct offset into scode firmware

When validating and loading SCODE firmware we need to take into account the
two-byte size header before each entry.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6643): xc2028: use best match instead of first partial match during firmware selection

Rather than picking the first video standard firmware that supports any of
the standards that the user has requested, try to select one that supports
as many of them as possible.  This improves the likelihood that the firmware
we select will support the user's desired TV standard.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6642): xc2028: don't duplicate max_len in priv

There is no need to duplicate the max_len field from the ctrl structure
in the private data.  If we use it directly from priv->ctrl, we can memcpy
the structure (apart from strings) to reduce maintenance as it grows.

Enforce a minimum max_len length of 8 data bytes (+ 1 address byte) as seems
to be required by the tuner.

Also, use kstrdup instead of open coding the string duplication.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6641): xc2028: correct tuner offset for 7MHz DTV

7MHz bandwidth DVB-T needs an adjusted offset at the PLL to ensure the IF
output is correctly centered.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6640): xc2028: correctly select 8MHz firmware

We were using priv->bandwidth to select the base firmware to load, not the
requested bandwidth value, oops.  Also, 7MHz Digital TV needs 8MHz base
firmware loaded.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6639): xc2028: correct divisor length

The frequency divisor should only be four bytes long.  Also, display the
frequency and divisor correctly in the debug output.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6638): xc2028: firmware loading cleanup

Hold the private lock over set_config and set priv->firm_size to 0 after a
failed firmware load to prevent firmware accidentally being freed on us.

Clean up the firmware load/error messages somewhat and rename priv->version
to priv->firm_version to make it clear which "version" it is.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6637): xc2028: add missing break

Add break to stop us from following the default failure path even upon success.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6636): xc2028: protect device list

Protect refcount changes and modifications to xc2028_list with a mutex.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6635): xc2028: v4l2_std_id needs to be long long to display completely

Cast v4l2_std_id variables to unsigned long long so they will printk properly.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6634): xc2028: error messages missing whitespace

Fix some missing spaces in errors that may be emitted during attach failure.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6633): xc2028: make register reads atomic

Issuing register reads as a separate address write and data read transactions
means that other I2C activity could occur in between and state could get out
of sync.  Issue both the write and read in a single transaction so that the
i2c layer can prevent other users accessing the bus until we are complete.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6632): xc2028: fix inverted logic in audio standard check

strcasecmp returns 0 on match, not true.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6631): xc2028: eliminate i2c macro side-effects

The I2C macros have side effects and send_seq could cause a return from
a function with a mutex held.  Change them to behave like real functions.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6630): zl10353: calculate input frequency register instead of using hardcoded value

Now we know the zl10353's correct ADC clock, we can calculate the input
frequency registers correctly instead of just blindly setting them.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6628): zl10353: Improve support for boards without a tuner on secondary i2c

Issue FSM_GO instead of TUNER_GO if there is no tuner attached to the
secondary i2c bus.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6627): CXUSB: handle write then read from different address

The path to perform a read immediately after a write was not checking that
the address being read from was the same as the one that was written.

Handling this case correctly should mean that we now can handle more than
two i2c messages at a time.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6626): CXUSB: support only-read i2c requests

Any i2c read request that was not immediately preceded by a write request was
incorrectly taking the write path.  Add the capability to handle individual
read requests.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6625): CXUSB: i2c transfer failure notification

The i2c master_xfer routine should return a negative result if not all
transfers completed successfully.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6624): CXUSB: return control message transfer result to caller

Callers to cxusb_ctrl_msg currently do not receive any indication that their
transfer failed.  Return the true return code from dvb_usb_generic_{rw,write}.

Signed-off-by: Chris Pascoe <c.pascoe@itee.uq.edu.au>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6623): remove saa7134-oss

The saa7134-oss is deprecated for quite some time, it's the only remaining OSS
user outside of sound/oss/, and considering how few and what kind of
soundcards are left supported by OSS I hardly see any use cases left.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6622): tda827x: prevent possible NULL pointer dereference in tda827xa_lna_gain

If tda827x_config hasn't been defined, exit the function.

Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6621): tda827x: fix NULL pointer dereference during tda827x_probe_version

Fix the following oops:

[  750.807586] CPU:    1
[  750.807587] EIP:    0060:[<f9dde2a3>]    Tainted: P       VLI
[  750.807589] EFLAGS: 00010296   (2.6.22-14-generic #1)
[  750.807599] EIP is at tda827x_probe_version+0xc3/0x130 [tda827x]
[  750.807603] eax: 00000000   ebx: f5a45a00   ecx: 00000000   edx: 00000001
[  750.807607] esi: f9de05ac   edi: e9b897e0   ebp: ed8015ac   esp: ed735f58
[  750.807611] ds: 007b   es: 007b   fs: 00d8  gs: 0000  ss: 0068
[  750.807615] Process kdvb-fe-1 (pid: 10662, ti=ed734000 task=d3b76530 task.ti=ed734000)
[  750.807618] Stack: f9e23972 00000038 ed735fd0 00010060 41ae0001 ed735f73 88b89608 e9b89608
[  750.807629]        e9b89608 ed801400 f9dde338 e9b89608 f9b2a46b 000000ae 00000a18 d3b76640
[  750.807639]        fffffffc f9b2bad6 00000003 f4433e68 e962b8c0 f4433e64 00000292 ed735fd0
[  750.807649] Call Trace:
[  750.807653]  [<f9e23972>] tda10046_init+0x2d2/0x6c0 [tda1004x]
[  750.807700]  [<f9dde338>] tda827x_initial_init+0x8/0x20 [tda827x]
[  750.807713]  [<f9b2a46b>] dvb_frontend_init+0x2b/0x60 [dvb_core]
[  750.807745]  [<f9b2bad6>] dvb_frontend_thread+0x66/0x2f0 [dvb_core]
[  750.807780]  [complete+64/96] complete+0x40/0x60
[  750.807805]  [<f9b2ba70>] dvb_frontend_thread+0x0/0x2f0 [dvb_core]
[  750.807822]  [kthread+66/112] kthread+0x42/0x70
[  750.807828]  [kthread+0/112] kthread+0x0/0x70
[  750.807841]  [kernel_thread_helper+7/16] kernel_thread_helper+0x7/0x10
[  750.807869]  =======================
[  750.807872] Code: 8b 74 24 20 8b 7c 24 24 83 c4 28 c3 8b 0d 00 0c de f9 85 c9 75 42 be e0 04 de f9 81 c7 0c 01 00 00 b9 33 00 00 00 f3 a5 8b 43 08 <c7> 40 14 50 e3 dd f9 8b 5c 24 1c 31 c0 8b 74 24 20 8b 7c 24 24
[  750.807925] EIP: [<f9dde2a3>] tda827x_probe_version+0xc3/0x130 [tda827x] SS:ESP 0068:ed735f58

Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6620): kconfig: VIDEO_SAA7134 must select VIDEO_TVEEPROM

The ability to read Hauppauge eeprom's was recently added to saa7134,
so we must build the tveeprom module.

Thanks to Matthias Schwarzott for pointing this out.

Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6619): Use MTS firmware for the HVR-900

The HVR-900 requires the MTS version of the xc3028 firmware in order
to get any sound. The below patch selects this firmware variant on
HVR-900 cards, as well as splitting the HVR-950 into its own entry
(since I don't know if it uses the MTS variant and it will have to be
split off eventually anyway).

Signed-off-by: Aidan Thornton <makosoft@googlemail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6618): drivers/media/dvb: Add missing "space"

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6617): drivers/media/video: Add missing "space"

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6616): drivers/media/radio: Add missing "space"

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6614): Fix driver for i386 architectures

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6613): Fix: add a missing continue statement

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6612): Allow RESET_CLK callback and avoids unneeded loading

TM5600/TM6000 needs clock reset during firmware load. This patch adds the
capability of caling a callback method for this.

Also, avoids uneeded firmware loads.

Signed-off-by: Michel Ludwig <michel.ludwig@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6611): Change xc2028_attach method to make easier for DVB

Removes uneeded parameters and adds an structure for passing the parameters

This patch is co-authored by Mauro Carvalho Chehab.

Signed-off-by: Michel Ludwig <michel.ludwig@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6610): Fix a wrong typecast

Signed-off-by: Michel Ludwig <michel.ludwig@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6607): saa7134: add support for reading Hauppauge eeprom

Increased size of dev->eedata from 128 to 256, since the Hauppauge data begins
at byte 128.  This has been tested on boards with smaller eeproms, and caused
no problems.

Added comments to distinguish between the various versions of the HVR1110.

Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6606): saa7134: add autodetection support for alternate subids of Hauppauge HVR1110

Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6605): Add a modprobe option to manually select audio standard

While there's no public API to define audio standard, adds a hack option for
select them. This is needed only for NICAM and A2 firmwares, since AM, BTSC and
EAIJ are already properly handled, on firmware version 2.7.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6603): V4L: videobuf: convert streaming and reading to bitfields

Signed-off-by: Brandon Philips <bphilips@suse.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6600): V4L: videobuf: don't chew up namespace STATE_.*, convert to VIDEOBUF_

s/STATE_NEEDS_INIT/VIDEOBUF_NEEDS_INIT/g
s/STATE_PREPARED/VIDEOBUF_PREPARED/g
s/STATE_QUEUED/VIDEOBUF_QUEUED/g
s/STATE_ACTIVE/VIDEOBUF_ACTIVE/g
s/STATE_DONE/VIDEOBUF_DONE/g
s/STATE_ERROR/VIDEOBUF_ERROR/g
s/STATE_IDLE/VIDEOBUF_IDLE/g

Signed-off-by: Brandon Philips <bphilips@suse.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6598): Fix standard name

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6595): Corrects printk lines

Make the driver less verbose by default. It adds a debug parameter to make the
driver more verbose.
Also, error messages were using KERN_ERR level, instead of KERN_INFO.
A few printk messages were reviewed to make them more clear.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6594): Add tuner_err macro

Some tuners, like xc3028, need to print error messages. Instead of declaring
local macros, create a tuner global macro for printing tuner errors.

To preserve CodingStyle on all tuner_macros, a few CodingStyle violations were
fixed at the other macros:
- lines with more than 80 columns
- two statements at the same line

The patch also removes the CodingStyle violation of having emacs declarations
inside de source code (CodingStyle chapter 18).

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6593): Fix scode table loading

Xceive 2028/3028 has a concept of scode/dcode.
Scode is a table of 16 values (each with 12 bytes i2c sequence).
Dcode is the entry of Scode table that should be used, given a certain
frequency.

The idea is that, depending on what frequency is selected, and according with a
country-based (or standard-based?) table, the Xceive should be "hacked" to
fine-tune that specific frequency.

By default, Scode=0 is used, for undefined frequencies. Also, Scode=0 seems to
be the most used value.

This patch adds the capability of selecting a scode. However, extra work will
be needed to allow auto-selecting the proper scode, for a given set of
frequencies.

I'm not sure what would be the proper way for implementing the dcode selection.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6592): Add the capability to work with more complete firmwares

Firmware version 2.7 has other firmware types. This patch adds the capability
for the driver to work with those newer types.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6591): Adds support for MT9V111 on sn9c102

Adds a new image sensor to the sn9c102 driver.

Signed-off-by: Luca Risolia <luca.risolia@studio.unibo.it>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6590): Adding support for VHF with MT2266-devices

MT2266 :
- support for VHF
- Minor enhancements

Signed-off-by: Olivier DANET <odanet@caramail.com>
Signed-off-by: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6589): Gigabyte u7000 usb dvb-t support

Signed-off-by: dominik <dominik95@free.fr>
Signed-off-by: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6588):  Leadtek Winfast DTV Dongle remote control

Add remote controller support for Leadtek Winfast DTV Dongle based on
DiB7700P, MT2060 and using last drivers (6040:6a79c243aecc), firmware
dvb-usb-dib0700-03-pre1.fw

Signed-off-by: Jaroslav Barton <djaara@djaara.net>
Signed-off-by: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6587): Cleanup at tv norm selection

With the conversion to the vidio_ioctl2, tvnorms array is not required anymore.

Also, removed some code from V4L1 time (VIDEO_MODE_foo), specied at the
non-used video_decoder.h.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6586): Remove some dead code and make drive fully V4L2 compatible

There were some vestiges of an old V4L1 I2C driver that were called by em28xx.
This patch removes this dead code, and replaces videodev.h to videodev2.h

Now, this driver doesn't require V4L1 anymore.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6585): Convert em28xx to video_ioctl2

Uses the newer ioctl handler at videodev. This patch also cleans up some
bad logic at the driver and do CodingStyle and other cleanups at the
resulting driver.

Also, since VIDIOCMBUF were not working, the V4L1 compat code were removed.
The compat code will eventually be re-inserted, if we find a clean way for
implementing compatibility with the old API.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6584): Fix read() method

Backport read() fixes from Markus Rechberger.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6583): Fix em28xx read stream locking

On some situations, closing an streaming application and re-opening were
returning -EBUSY.

Uses the same locking schema also present on cx88.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6582): Fix em28xx to allow multiple open

Allows shared access support for em28xx. Just one userspace application is
allowed to get stream. The other(s) application(s) can change V4L2 controls,
set video standards, etc.

This patch were splited from Markus Rechberger's tree and backported to 2.6.17
by Pádraig Brady.

The original patch were ported to the latest em28xx version and had CodingStyle
corrected to solve the issues pointed by scripts/checkpatch.pl.

Thanks to Pádraig Brady <P@draigBrady.com> for pointing this.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6580): Set slave's master before master's attach call.

V4L: Int if: Set slave's master before attach, remove master argument

The master also now gets its own pointer from slave's structure.

Signed-off-by: Sakari Ailus <sakari.ailus@nokia.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6578): dvb-usb: make some debug vars static

This patch makes some needlessly global debug variables static.

opera1.h became so small that I removed it.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Reviewed-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6577): et61x251/: make 5 functions static

This patch makes the following needlessly global functions in
et61x251_core.c static:
- et61x251_read_reg()
- et61x251_i2c_try_read()
- et61x251_i2c_try_write()
- et61x251_i2c_read()
- et61x251_i2c_write()

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6576): cx88-mpeg.c: make 4 functions static

This patch makes the following needlessly global functions static:
- cx8802_init_common()
- cx8802_fini_common()
- cx8802_suspend_common()
- cx8802_resume_common()

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6575): cx23885/: cleanups

This patch contains the following cleanups:
- make the following needlessly global code static:
  - cx23885-core.c: struct cx23885_sram_channels[]
  - cx23885-core.c: struct cx23887_sram_channels[]
  - cx23885-core.c: cx23885_wakeup()
  - cx23885-core.c: cx23885_sram_channel_setup()
  - cx23885-core.c: cx23885_sram_channel_dump()
  - cx23885-core.c: cx23885_risc_disasm()
  - cx23885-core.c: cx23885_shutdown()
  - cx23885-core.c: cx23885_reset()
  - cx23885-core.c: cx23885_dev_unregister()
  - cx23885-core.c: cx23885_risc_databuffer()
  - cx23885-core.c: cx23885_risc_stopper()
- #if 0 the following unused functions:
  - cx23885-core.c: cx23885_risc_buffer()
  - cx23885-core.c: cx23885_cancel_buffers()
- remove the following unused EXPORT_SYMBOL's:
  - cx23885-cards.c: cx23885_boards
  - cx23885-i2c.c: cx23885_call_i2c_clients

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Reviewed-by: Steven Toth <stoth@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6574): common/ir-functions.c: make a function static

ir_rc5_decode() can become static.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6573): unexport flexcop_reset_block_300

This patch removes the unused EXPORT_SYMBOL(flexcop_reset_block_300).

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6572): dvb-usb/vp702x.c: cleanups

This patch contains the following cleanups:
- make the needlessly global vp702x_usb_out_op() static
- #if 0 the unused vp702x_power_ctrl()

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6571): dvb-usb/gp8psk.c: #if 0 gp8psk_bcm4500_reload()

This patch #if 0's the unused gp8psk_bcm4500_reload()

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6570): core/dvb_ringbuffer.c: remove unused exports

This patch removes the following unused EXPORT_SYMBOL's:
- dvb_ringbuffer_flush
- dvb_ringbuffer_pkt_write
- dvb_ringbuffer_pkt_read
- dvb_ringbuffer_pkt_dispose
- dvb_ringbuffer_pkt_next

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6569): bt8xx/dst.c: make code static

This patch makes the following needlessly global code static:
- dst_gpio_outb()
- dst_gpio_inb()
- rdc_8820_reset()
- dst_pio_enable()
- dst_command()
- struct tuner_list[]

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6564): Move check before lock

Signed-off-by: Roel Kluin <12o3l@tiscali.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
CC: Antti Palosaari <crope@iki.fi>
CC: Carl Lundqvist <comabug@gmail.com>
CC: Aapo Tahkola <aet@rasterburn.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6563): tda8290: optimize for loop in tda829x_probe function

Thanks to Trent Piepho for pointing this out.

Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6562): Make HVR900 to use also tveeprom

Hauppauge firmwares can be decoded using tveeprom. This patch adds HVR-900 as
a tveeprom client. It also adds xc3028 tuner entry to tveeprom.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6561): Fix xc2028 get register functions and calls

The status registers require a dword for setting register. Fix it on all
occurrences, and at xc3028_get_reg. Also, improves the hardware/firmware
detection printk.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6560): Fix a bug when setting tuner type

Tuner-type were correctly filled only by the hint function.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6559): Fix a buffer overflow at xc2028_get_reg

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6557): tea5767: remove unnecessary warning

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6556): tuner: convert to bus-based I2C API

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6555): tuner: reorder functions to prepare for i2c conversion

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6553): tuner: replace default_mode_mask

The default_mode_mask global is replaced by a list of tuner structs. The
tuner driver now walks that list to see which radio and/or tv tuner
devices are registered to a given i2c adapter.

The default_mode_mask global had to go since this is no longer supported
with the new bus-based I2C API.

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6551): tda8290: rule out tda988x before detecting tda8290/tda8295

To ensure prevention of detecting a tda9885/6/7 as a tda8290 or tda8295, we
will rule out the tda988x before testing the tda8290 / tda8295 id registers.

We read 8 bytes from the chip.  If they are all equal, then it is not a
tda829x, or some other error has occurred.

Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6550): tda8290: return -ENODEV on probe failures

Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB(6548b) Fix compilation for em28xx

From: Mauro Carvalho Chehab <mchehab@infradead.org>

In file included from drivers/media/video/em28xx/em28xx-i2c.c:31:
drivers/media/video/tuner-xc2028.h:10:26: error: dvb_frontend.h: No such file or directory

In file included from drivers/media/video/em28xx/em28xx-i2c.c:31:
drivers/media/video/tuner-xc2028.h:32: warning: 'struct dvb_frontend' declared inside parameter list
drivers/media/video/tuner-xc2028.h:32: warning: its scope is only this definition or declaration, which is probably not what you want

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB(6548a) Fix compilation for TDA8290

From: Mauro Carvalho Chehab <mchehab@infradead.org>

drivers/media/video/tda8290.c:27:21: error: tda827x.h: No such file or directory
drivers/media/video/tda8290.c:28:22: error: tda18271.h: No such file or directory
drivers/media/video/tda8290.c:52: error: field 'cfg' has incomplete type
drivers/media/video/tda8290.c: In function 'tda829x_find_tuner':
drivers/media/video/tda8290.c:590: error: implicit declaration of function 'tda18271_attach'
drivers/media/video/tda8290.c:598: error: implicit declaration of function 'tda827x_attach'

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6546): Add comments for the hint methods

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6545): em28xx: autodetect Cinergy 200 USB and VGear PocketTV

Adds autodetection support for the Cinergy200 USB and the VGear PocketTV.

Whenever a usb device with generic empia em2800 usb ids is detected the device
gets scanned for connected i2c devices. If the device list matches an em2800
device in the device list the model id gets changed accordingly.

Signed-off-by: Sascha Sommer <saschasommer@freenet.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6543): tda8290: enable probing of tda8295

Prevent the tda8295 from falsely being detected as a tda9887

Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6541): Add V-Gear PocketTV to Cardlist.em28xx

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6539): em28xx: add support for vgear pockettv

attached patch adds support for the vgear pockettv.
It seems to require a write to another register for audio to work.
I checked my old cinergydrv and we did the same register write there. I
therefore enabled it for all em2800 based devices.

Signed-off-by: Sascha Sommer <saschasommer@freenet.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6538): em28xx: fix locking to allow accesses from 2 different threads at the same time

The attached patch modifies the em28xx driver so that there can be ioctls from
multiple different threads.

This is necessary for capture apps like MPlayer that use different threads for
capturing and channel tuning.

Now the locking is only done for the ioctls that change properties of the
device or access the i2c bus.

It also removes some locks that look unnecessary:

In em28xx_init_dev:
  the videodevice is not registered yet so nothing can access the hardware
 meanwhile, the device struct is not assigned to the interface yet so no race
 with disconnect is possible

In em28xx_release_resources:
  it gets only called when dev->lock is already held

Signed-off-by: Sascha Sommer <saschasommer@freenet.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6537): Add entry for Pixelview Prolink PlayTV USB 2.0

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6536): Add a hint for boards without unique USB ID

This patch adds a function to allow trying to detect boards that shares
the generic IDs.

The current detection method is based at eeprom checksum.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6535): Fix: Adds the generic PCI IDs for em28xx

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6519): Fix HVR900/HVR950 entry

- Television is now default;
- Add HVR950 name at the entry.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6517): CodingStyle fixup

Used scripts/Lindent + manual check + scripts/checkpatch.pl

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

V4L/DVB (6516): Allow faster loading by using 64 bytes block by em28xx i2c write

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>