[media] uvcvideo: Remove unneeded device disconnected flag

The UVC_DEV_DISCONNECTED flag is set when the USB device is
disconnected, and checked in the open() handler to refuse opening a
device that has been disconnected.

This is inherently racy, but the race condition is taken care of in the
V4L2 core, which also fails open() calls when the V4L2 device node has
been unregistered. As the uvcvideo USB disconnect handler unregisters
all its video device nodes, the UVC_DEV_DISCONNECTED flag doesn't serve
any purpose anymore and can be removed.

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

[media] uvcvideo: Fix incorrect bandwidth with Chicony device 04f2:b50b

The "TOSHIBA Web Camera - 5M" Chicony device (04f2:b50b) seems to
compute the bandwidth on 16 bits and erroneously sign-extend it to
32 bits, resulting in a huge bandwidth value. Detect and fix that
condition by setting the 16 MSBs to 0 when they're all equal to 1.

Reported-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] uvcvideo: Implement DMABUF exporter role

Now that videobuf2-vmalloc supports exporting buffers, add support for
the DMABUF exporter role by plugging in the videobuf2 ioctl helper.

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

[media] af9035: add USB ID 07ca:0337 AVerMedia HD Volar (A867)

It is AF9035 + MxL5007T.
Driver reports:
prechip_version=00 chip_version=03 chip_type=3802

Not sure if that USB ID is reserved only for HP brand or if it is
common, but the stick I have is branded as HP part no. 580715-001
rmn A867.

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

[media] m88ds3103: add I2C client binding

Implement I2C client device binding.
Wrap media attach to driver I2C probe.
Add wrapper from m88ds3103_attach() to m88ds3103_probe() via driver
core in order to provide proper I2C client for legacy media attach
binding.

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

[media] m88ds3103: use jiffies when polling DiSEqC TX ready

Use jiffies to set timeout for DiSEqC TX ready polling. Using jiffies
is more elegant solution than looping N times with sleep.

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

[media] m88ds3103: implement DVBv5 BER

Implement DVBv5 BER statistics.
Wrap legacy DVBv3 BER to DVBv5 BER.

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

[media] m88ds3103: implement DVBv5 CNR statistics

Implement DVBv5 CNR statistics.
Wrap legacy DVBv3 SNR to DVBv5 CNR.

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

[media] m88ds3103: do not return error from get_frontend() when not ready

Do not return error from get_frontend() when status is queried, but
the frontend didn't lock yet. The proper behavior is to only update the
cache after having a lock.

[mchehab@osg.samsung.com: fix the patch description]
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] gspca: sn9c2028: Add gain and autogain controls Genius Videocam Live v2

Autogain algorithm is very simple, if average luminance is low - increase gain,
if it's high - decrease gain. Gain granularity is low enough for this algo to
stabilize quickly.

Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] gspca: sn9c2028: Add support for Genius Videocam Live v2

This cam seems to return different values on long commands, so make status check
in sn9c2028_long_command() more tolerant. Anyway, read value isn't used anywhere
later.

Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] wl128x: Allow compile test of GPIO consumers if !GPIOLIB

The GPIO subsystem provides dummy GPIO consumer functions if GPIOLIB is
not enabled. Hence drivers that depend on GPIOLIB, but use GPIO consumer
functionality only, can still be compiled if GPIOLIB is not enabled.

Relax the dependency on GPIOLIB if COMPILE_TEST is enabled, where
appropriate.

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] return BUF_STATE_ERROR if streaming stopped during acquisition

When stop_streaming is called while a frame is currently being retrieved, the
buffer being filled will still be returned with BUF_STATE_DONE. By resetting
the sequence number and checking before returning the buffer, it can now
correctly be returned with BUF_STATE_ERROR.

Signed-off-by: Martin Kaltenbrunner <modin@yuri.at>
Signed-off-by: Florian Echtler <floe@butterbrot.org>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] add extra debug output, remove noisy warning

Add dev_dbg statements for easier future debugging; also change the warning
about packet ID mismatches to debug output to avoid flooding the logs. This
warning is only important in a very specific/rare use case when trying to
correlate input events with video data.

Signed-off-by: Martin Kaltenbrunner <modin@yuri.at>
Signed-off-by: Florian Echtler <floe@butterbrot.org>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] add frame size/frame rate query functions

Add missing functions to query the single fixed frame size (960x540) and
supported frame rates. Technically, the SUR40 supports any arbitrary frame
rate up to 60 FPS, as it is polled and not interrupt-driven. For now, we
just report 30 and 60 FPS, which is sufficient to make most V4L2 tools work.

Signed-off-by: Martin Kaltenbrunner <modin@yuri.at>
Signed-off-by: Florian Echtler <floe@butterbrot.org>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] reduce poll interval to allow full 60 FPS framerate

The SUR40 hardware can deliver images at up to 60 FPS; at full USB2 bandwidth,
one raw frame will take about 11 ms to transmit. If the poll interval is above
5 ms, fully handling one frame will take longer than 16 ms and the overall
frame rate will drop below 60 FPS. To get the full frame rate without blocking
all the time and still allowing for a bit of timing jitter, we reduce the poll
interval to 4 ms.

Signed-off-by: Martin Kaltenbrunner <modin@yuri.at>
Signed-off-by: Florian Echtler <floe@butterbrot.org>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] v4l2: correct two SDR format names

U8 and U16LE format human readable names were translated with string
containing emulated word. These strings were taken from msi2500 driver,
where those formats were emulated (after module parameter is set). But
on API correct names should be used, without any special case notes.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] vivid: SDR cap add 'CU08' Complex U8 format

Add complex unsigned 8-bit sample format for SDR capture.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] vivid: use new V4L2_MAP_*_DEFAULT defines

Use these defines instead of hardcoding this in any driver that needs it.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] videodev2.h: add macros to map colorspace defaults

The mapping of COLORSPACE_DEFAULT, YCBCR_ENC_DEFAULT or QUANTIZATION_DEFAULT
to proper non-default values is fairly complex, and it is something that
needs to be done both in the kernel and in userspace.

So add macros that can do this conversion, making this available to both
kernel and userspace.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] DocBook/media: document COLORSPACE_RAW

Document this new colorspace define.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] videodev2.h: add COLORSPACE_RAW

V4L2_COLORSPACE_RAW is added for raw image formats where the picture
is minimally processed and is in the internal colorspace of the sensor.

This is typically used in digital cameras where the image processing is
done later.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] DocBook/media: document COLORSPACE_DEFAULT

Document this new colorspace define.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] videodev2.h: add COLORSPACE_DEFAULT

V4L2_COLORSPACE_DEFAULT is added so we have a specific define for
the default case where applications do not set it but leave it to 0.
In that case the driver will set the colorspace based on what it
captures.

This is already used, but we never had a define for the value 0.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] ivtv: fix incorrect audio mode report in log_status

The log_status ioctl should report if the audio input has mode Bilingual. However, the
check against the itv->dualwatch_stereo_mode is completely wrong and is a left-over from
the distant past. Not only is the bitmask obviously wrong, the test itself is broken too
since itv->dualwatch_stereo_mode is no longer a bitmask at all.

Fix this code properly.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] DocBook/media: add missing entry for V4L2_PIX_FMT_Y16_BE

This format was added but not documented. Do this now.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] media/vivid: Add support for Y16_BE format

Support for V4L2_PIX_FMT_Y16_BE, a 16 bit big endian greyscale format.

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] media/v4l2-core: Add support for V4L2_PIX_FMT_Y16_BE

16 bit greyscale format, structured in Big Endian. Such a format can be
converted into a PMN image just by adding a header.

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] media/videobuf2-dma-vmalloc: Save output from dma_map_sg

dma_map_sg returns the number of areas mapped by the hardware,
which could be different than the areas given as an input.
The output must be saved to nent.

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Reviewed-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] media/videobuf2-dma-contig: Save output from dma_map_sg

dma_map_sg returns the number of areas mapped by the hardware,
which could be different than the areas given as an input.
The output must be saved to nent.

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Reviewed-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] media/videobuf2-dma-sg: Fix handling of sg_table structure

When sg_alloc_table_from_pages() does not fail it returns a sg_table
structure with nents and nents_orig initialized to the same value.

dma_map_sg returns the number of areas mapped by the hardware,
which could be different than the areas given as an input.
The output must be saved to nent.

The output of dma_map, should be used to transverse the scatter list.

dma_unmap_sg needs the value passed to dma_map_sg (nents_orig).

sg_free_tables uses also orig_nent.

This patch fix the file to follow this paradigm.

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Reviewed-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] adv7604/cobalt: missing GPIOLIB dependency

The adv7604 driver depends on GPIOLIB, and therefore cobalt depends
on it as well.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] saa7164: fix sparse warning

drivers/media/pci/saa7164/saa7164-i2c.c:45:33: warning: Using plain integer as NULL pointer

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Steven Toth <stoth@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cx24120: fix sparse warning

drivers/media/dvb-frontends/cx24120.c:837:6: warning: symbol 'cx24120_calculate_ber_window' was not declared. Should it be static?

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cobalt: fix sparse warnings

drivers/media/pci/cobalt/cobalt-v4l2.c:189:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:191:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:192:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:193:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:194:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:195:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:196:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:197:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:198:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:199:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:201:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:202:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:203:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:234:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:240:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:246:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:266:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:267:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:271:28: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:275:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:276:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:312:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:313:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:314:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:315:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:317:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:320:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:321:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:321:36: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:324:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:327:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:328:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:328:41: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:329:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:331:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:332:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:334:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:335:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:336:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:362:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:367:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:368:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:420:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:421:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:422:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:423:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:516:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:516:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:518:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:518:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:518:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:525:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:525:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:531:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:531:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:531:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:531:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:531:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:531:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:531:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:546:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:546:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:548:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:549:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:550:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:551:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:552:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:553:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:554:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:555:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:556:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:556:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:556:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:563:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:564:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:564:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:569:9: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:595:16: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-v4l2.c:602:9: warning: dereference of noderef expression

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cobalt: fix sparse warnings

drivers/media/pci/cobalt/cobalt-i2c.c:130:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:147:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:151:26: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:156:34: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:206:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:210:26: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:215:34: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:225:27: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:335:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:336:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:337:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:348:34: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:352:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:353:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:356:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:357:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-i2c.c:359:17: warning: dereference of noderef expression

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cobalt: fix sparse warnings

drivers/media/pci/cobalt/cobalt-irq.c:62:33: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:64:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:65:23: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:72:21: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:73:25: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:74:25: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:82:33: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:83:33: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:91:25: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:94:23: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:103:25: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:107:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:109:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:116:13: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:119:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:120:17: warning: dereference of noderef expression
drivers/media/pci/cobalt/cobalt-irq.c:122:17: warning: dereference of noderef expression

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cobalt: fix sparse warnings

drivers/media/pci/cobalt/cobalt-flash.c:101:5: warning: symbol 'cobalt_flash_probe' was not declared. Should it be static?
drivers/media/pci/cobalt/cobalt-flash.c:126:6: warning: symbol 'cobalt_flash_remove' was not declared. Should it be static?
drivers/media/pci/cobalt/cobalt-cpld.c:101:6: warning: symbol 'cobalt_cpld_status' was not declared. Should it be static?
drivers/media/pci/cobalt/cobalt-cpld.c:240:6: warning: symbol 'cobalt_cpld_set_freq' was not declared. Should it be static?

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cobalt: fix sparse warnings

drivers/media/pci/cobalt/cobalt-flash.c:39:36: warning: incorrect type in initializer (different address spaces)
drivers/media/pci/cobalt/cobalt-flash.c:54:36: warning: incorrect type in initializer (different address spaces)
drivers/media/pci/cobalt/cobalt-flash.c:63:36: warning: incorrect type in initializer (different address spaces)
drivers/media/pci/cobalt/cobalt-flash.c:82:36: warning: incorrect type in initializer (different address spaces)
drivers/media/pci/cobalt/cobalt-flash.c:107:19: warning: incorrect type in assignment (different address spaces)

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] e4000: fix compiler warning

drivers/media/tuners/e4000.c:287:3: warning: this decimal constant is unsigned only in ISO C90
   .rangehigh  =  2208000000L,
   ^

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Cc: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cobalt: fix compiler warnings on 32 bit OSes

Fixes these warnings:

drivers/media/pci/cobalt/cobalt-omnitek.c: In function 'omni_sg_dma_start':
drivers/media/pci/cobalt/cobalt-omnitek.c:112:28: warning: right shift count >= width of type [-Wshift-count-overflow]
  iowrite32((u32)(desc->bus >> 32), DESCRIPTOR(s->dma_channel) + 4);
                            ^
drivers/media/pci/cobalt/cobalt-omnitek.c: In function 'descriptor_list_create':
drivers/media/pci/cobalt/cobalt-omnitek.c:222:28: warning: right shift count >= width of type [-Wshift-count-overflow]
     d->next_h = (u32)(next >> 32);
                            ^
drivers/media/pci/cobalt/cobalt-omnitek.c:268:32: warning: right shift count >= width of type [-Wshift-count-overflow]
    d->next_h = (u32)(desc->bus >> 32);
                                ^
drivers/media/pci/cobalt/cobalt-omnitek.c:275:27: warning: right shift count >= width of type [-Wshift-count-overflow]
    d->next_h = (u32)(next >> 32);
                           ^
drivers/media/pci/cobalt/cobalt-omnitek.c: In function 'descriptor_list_chain':
drivers/media/pci/cobalt/cobalt-omnitek.c:293:31: warning: right shift count >= width of type [-Wshift-count-overflow]
   d->next_h = (u32)(next->bus >> 32);
                               ^
drivers/media/pci/cobalt/cobalt-omnitek.c: In function 'descriptor_list_loopback':
drivers/media/pci/cobalt/cobalt-omnitek.c:332:30: warning: right shift count >= width of type [-Wshift-count-overflow]
  d->next_h = (u32)(desc->bus >> 32);
                              ^

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cobalt: fix 64-bit division link error

[linuxtv-media:master 1023/1029] ERROR: "__aeabi_uldivmod" [drivers/media/pci/cobalt/cobalt.ko] undefined!

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cobalt: fix irqs used for the adv7511 transmitter

The interrupt bit assignments use for the adv7511 were off by one.
This means that the current scheme (bit << (4 * stream_index)) can
no longer be used.

Fix this by precalculating and storing the correct masks in the
cobalt_stream struct.

This wasn't noticed before because the adv7511 interrupts are very
rare. But for CEC support these interrupts are essential, so this made
me realize that it wasn't working correctly.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] DocBook: Update DocBook version and fix a few legacy things

The DVB part of the media API documentation has several
legacy things on it:
- Examples that don't work;
- APIs unused and deprecated;
- places mentioning the wrong API version.

Fix them and bump the documentation version, in order to
reflect the cleanup efforts to make it more consistent with
the current status of the API.

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

[media] DocBook: Fix arguments on some ioctl documentation

Due to a cut-and-paste error, the argument is missing or wrong
on 3 ioctl documentation. Fix them.

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

[media] DocBook: cleaup the notes about DTV properties

The notes there are somewhat confusing and assumes that the
reader would have read the DVBv3 way. This is not true anymore,
as the DVBv3 is now on a separate section that is marked as
deprecated.

So, cleanup the notes.

While here, add a note about using libdvbv5, instead of using
the DVBv5 API directly.

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

[media] DocBook: Add an example for using FE_SET_PROPERTY

In order to make it clearer about how to use the DVBv5 calls,
add an example of its usage. That should make it clearer about
what's actually required for the DVBv5 calls to work.

While here, also mentions the libdvbv5 library.

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

[media] DocBook: improve documentation of the properties structs

Rename the tytle of the struct documentation to reflect
the name of the structures, and use links to do cross-ref.

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

[media] DocBook: improve documentation for hierarchy

Format it as a table and links it with the legacy API xml.

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

[media] DocBook: improve documentation for guard interval

Format it as a table and add more details, in special for
DTMB guard intervals.

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

[media] DocBook: improve documentation for FEC fields

Format it as a table and add more details. Also, remove the
duplicated occurrences.

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

[media] DocBook: move fe_bandwidth to the frontend legacy section

fe_bandwidth/fe_bandwidth_t is used only on DVBv3 API. So, move
it to the frontend legacy xml, and convert it into a table.

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

[media] DocBook: improve documentation for OFDM transmission mode

Format it as a table and add more details, in special, for
the DTMB modes.

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

[media] DocBook: improve documentation for DVB spectral inversion

Format it as a table and provide more details.

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

[media] DocBook: cross-reference enum fe_modulation where needed

At frontend legacy API description, there are three places
where fe_modulation_t is defined. Cross-reference it to point
to the right place at the documentation.

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

[media] DocBook/Makefile: improve typedef parser

The typedef parser is wrong and doesn't get some of the
types defined at the DVB API. Improve it, as we want to add
cross-references to those types.

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

[media] DocBook: some fixes at FE_GET_INFO

There are a few issues at FE_GET_INFO documentation:
- name is a string, not a pointer to a string;
- the return text should be after the paragraph.

While here, better to bold that two fields of the structure used
by FE_GET_INFO are actually deprecated.

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

[media] DocBook: Better document enum fe_modulation

Instead of using programlisting, use a table, as this provides
a better view of the structure.

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

[media] DocBook: add a proper description for dvb_frontend_info.fe_type

The fe_type is deprecated at the DVB API. However, it may still
be used by legacy DVBv3 applications. While this works with old
devices, modern devices may support more than one delivery
system.

Add an explanation about that and a point to what should be
used, instead, in order for legacy apps to support newer hardware.

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

[media] DocBook: Provide a high-level description for DVB frontend

Instead of just showing ioctls, let's add an introdutory text
briefly explaining the DVB frontend API.

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

[media] DocBook: fix FE_READ_STATUS argument description

What's written there about the arguments for this ioctl
is bogus: it doesn't return an enum (or a typedef)
for enum fe_status. Instead, it returns a bitmask with the
values defined by enum fe_status.

Also, the size of the integer returned is not 16 bits, but,
instead, sizeof(fe_status_t), e. g. sizeof(enum), with is
arch-dependent.

This should of course be fixed, but this should be done on
a separate patch.

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

[media] DocBook: better organize the function descriptions for frontend

Move the function ioctl definitions to the end of the chapter,
at their importance. That makes the document better organized,
as the DVB frontend system call index will look like:

open()
close()
ioctl FE_GET_INFO — Query DVB frontend capabilities and returns information about the front-end. This call only requires read-only access to the device
ioctl FE_READ_STATUS — Returns status information about the front-end. This call only requires read-only access to the device
ioctl FE_SET_PROPERTY, FE_GET_PROPERTY — FE_SET_PROPERTY sets one or more frontend properties. FE_GET_PROPERTY returns one or more frontend properties.
ioctl FE_DISEQC_RESET_OVERLOAD — Restores the power to the antenna subsystem, if it was powered off due to power overload.
ioctl FE_DISEQC_SEND_MASTER_CMD — Sends a DiSEqC command
ioctl FE_DISEQC_RECV_SLAVE_REPLY — Receives reply from a DiSEqC 2.0 command
ioctl FE_DISEQC_SEND_BURST — Sends a 22KHz tone burst for 2x1 mini DiSEqC satellite selection.
ioctl FE_SET_TONE — Sets/resets the generation of the continuous 22kHz tone.
ioctl FE_SET_VOLTAGE — Allow setting the DC level sent to the antenna subsystem.
ioctl FE_ENABLE_HIGH_LNB_VOLTAGE — Select output DC level between normal LNBf voltages or higher LNBf voltages.
ioctl FE_SET_FRONTEND_TUNE_MODE — Allow setting tuner mode flags to the frontend.

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

[media] DocBook: better document FE_DISEQC_RESET_OVERLOAD

Use the new format for the ioctl documentation and put the
struct dvb_diseqc_slave_reply together with the ioctl.

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

[media] DocBook: better document FE_DISEQC_SEND_MASTER_CMD

Use the new format for the ioctl documentation and put the
struct dvb_diseqc_slave_reply together with the ioctl.

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

[media] DocBook: better document FE_DISEQC_RECV_SLAVE_REPLY

Use the new format for the ioctl documentation and put the
struct dvb_diseqc_slave_reply together with the ioctl.

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

[media] DocBook: better document FE_DISEQC_SEND_BURST ioctl

Use the proper format for FE_DISEQC_SEND_BURST documentation
and  improve the documentation.

Keep the enum fe_sec_mini_cmd description together with
the ioctl, as both are used together.

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

[media] DocBook: better document FE_SET_TONE ioctl

Use the proper format for FE_SET_TONE documentation and
improve the documentation.

Keep the enum fe_sec_tone_mode description together with
the ioctl, as both are used together.

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

[media] DocBook: better document FE_SET_VOLTAGE ioctl

Use the proper format for FE_SET_VOLTAGE documentation and fix
the documentation. The description for the enum is not 100%,
and it is missing the voltage off value.

Also, it is better to keep the enum description together with
the ioctl, as both are used together.

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

[media] DocBook: reformat FE_ENABLE_HIGH_LNB_VOLTAGE ioctl

Use the proper format for FE_ENABLE_HIGH_LNB_VOLTAGE documentation.

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

[media] DocBook: reformat FE_SET_FRONTEND_TUNE_MODE ioctl

Use the proper format for FE_SET_FRONTEND_TUNE_MODE documentation.

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

[media] DocBook: move FE_GET_PROPERTY to its own xml file

That helps the xref logic at the Makefile to point to the
right place. Also, it becomes more organized and easier to
maintain if each ioctl have its own xml file.

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

[media] DocBook: Rename ioctl xml files

for the xml files describing ioctls, use the same nomenclature
as on V4L2: the ioctl name, in lower case, using - instead of _.

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

[media] DocBook: Fix false positive undefined ioctl references

The new code that detects undocumented ioctls hits some false
positives:

This one is not documented, nor it should, as this is
there just to reserve namespace:

Warning: can't find reference for VIDIOC_RESERVED ioctl

But those are already documented together with other ioctls:

Warning: can't find reference for VIDIOC_UNSUBSCRIBE_EVENT ioctl
Warning: can't find reference for FE_GET_PROPERTY ioctl
Warning: can't find reference for VIDIOC_SUBDEV_G_EDID ioctl
Warning: can't find reference for VIDIOC_SUBDEV_S_EDID ioctl
Warning: can't find reference for VIDIOC_SUBDEV_S_DV_TIMINGS ioctl
Warning: can't find reference for VIDIOC_SUBDEV_G_DV_TIMINGS ioctl
Warning: can't find reference for VIDIOC_SUBDEV_QUERY_DV_TIMINGS ioctl

So, we need to just be sure to point to the right documentation.

Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] DocBook: remove duplicated ioctl from v4l2-subdev

Those ioctls are already parsed. No need to explicitly add
them.

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

[media] DocBook: Improve xref check for undocumented ioctls

There are several badly documented undocumented ioctls.
Currently, it just generates an empty link. Instead of doing that,
only add references to the ones that exists, and add a warning
for all references that weren't found.

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

[media] DocBook: Add xref links for DTV propeties

Create xref links for all DTV properties and link the frontend.h
to each. Also use them at the DVB frontent API example.

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

[media] DocBook: Improve the description of the properties API

Make the text clearer about what the properties API does.

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

[media] DocBook: Merge FE_SET_PROPERTY/FE_GET_PROPERTY ioctl description

Instead of having two refentries, merge them into just one,
like what's done with other similar ioctls at V4L2 side.

That makes the entry cleaner and will allow to add the associated
structures together with the refentry.

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

[media] DocBook: fix xref to the FE open() function

Instead of going to the V4L2 open(), use the xref to the
proper place at the frontend ioctls that were already
reformatted.

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

[media] DocBook: rewrite FE_GET_PROPERTY/FE_SET_PROPERTY to use the std way

Use the standard way of documenting ioctls for FE_GET_PROPERTY
and FE_SET_PROPERTY.

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

[media] DocBook: move DVB properties to happen earlier at the document

The DVBv5 API uses DVB properties as the main way to set the frontend
and collect statistics. Move the definition to happen earlier, in
order to reflect its importance.

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

[media] DocBook: improve documentation for FE_READ_STATUS

Move the ioctl and enum fe_status to a separate xml file and
put it into a better format.

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

[media] DocBook: move FE_GET_INFO to a separate xml file

Keeping everything altogether makes harder to reorganize the
DocBook. So, move the FE_GET_INFO ioctl and the associated structures
into a separate file.

No functional changes.

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

[media] DocBook: reformat FE_GET_INFO ioctl documentation

The DVB part of the docbook has a completely different format
than the V4L2 part, as it was written as a separate document.

As the V4L2 documentation is on better shape, and its format
allows adding more information, let's use it for FE_GET_INFO
and gradually update the non-legacy DVB ioctls using the new
format.

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

[media] dvb: split enum from typedefs at frontend.h

Using typedefs is already bad enough, but doing it together
with enum declaration is even worse.

Also, it breaks the scripts at DocBook that would be generating
reference pointers for the enums.

Well, we can't get rid of typedef right now, but let's at least
declare it on a separate line, and let the scripts to generate
the cross-reference, as this is needed for the next DocBook
patches.

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

[media] DocBook: move DVBv3 frontend bits to a separate section

Although the recommended usage is the DVBv5 API calls, the
documentation doesn't make it clear about what's the recommended
calls and what's legacy.

So, move the legacy API bits to a separate xml, putting them into
a new section.

Please notice that more changes are needed, since some of the
bits there are cross-referenced elsewhere.

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

[media] DocBook: Improve DVB frontend description

The DVB frontend API got bitrotten. Update it.

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

[media] DocBook: fix emphasis at the DVB documentation

Currently, it is using 'role="tt"', but this is not defined at
the DocBook 4.5 spec. The net result is that no emphasis happens.

So, replace them to bold emphasis.

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

[media] DocBook: add drawing with a typical media device

An illustration of what's considered a typical media device
may help people to better understand the contents of the media
infrastructure API docbook. So, add it.

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

[media] DocBook: add a note about the ALSA API

The Docbook mistakenly makes to believe that all needed APIs for
media devices are there. Add a note there pointing that some
sub-devices are actually be controlled via ALSA API.

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

[media] DocBook: Update DVB supported standards at introduction

The list of standards at the media docbook is incomplete, and it
is mentioning that the DVB-S2 & friends is "currently being updated".
That's wrong, as such update occurred back in 2008.

So, provide a more complete list of supported standards and add
a reference to the actual list.

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

[media] cx24120: Assume ucb registers is a counter

The ucblocks register is probably a counter and not a rate; assume
it is so and change the calculations as required.

Signed-off-by: Jemma Denson <jdenson@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] dvb-core: fix 32-bit overflow during bandwidth calculation

Frontend bandwidth calculation overflows on very high DVB-S/S2
symbol rates. Use mult_frac() macro in order to keep calculation
correct.

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

[media] e4000: Fix rangehigh value

As reported by smatch:

drivers/media/tuners/e4000.c:287:32: warning: constant 2208000000 is so big it is long long
drivers/media/tuners/e4000.c:287:32: warning: decimal constant 2208000000 is between LONG_MAX and ULONG_MAX. For C99 that means long long, C90 compilers are very likely to produce unsigned long (and a warning) here
drivers/media/tuners/e4000.c:287:3: warning: this decimal constant is unsigned only in ISO C90
   .rangehigh  =  2208000000,
   ^

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

[media] e4000: implement V4L2 subdevice tuner and core ops

Implement V4L2 subdevice tuner and core ops. After that this driver
is hybrid driver implementing both V4L2 and DVB ops.

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

[media] e4000: various small changes

* Rename device state from 's' to 'dev'.
* Move single include to driver private header.
* Change error handling type of each function to one I tend use
nowadays.
* Remove dummy register write from init. Even Windows driver does this
multiple times remove it as I have never seen any I2C errors.
* Define I2C client pointer for each function and use it.
* Do not clean tuner ops during driver remove - not needed.
* Disable sysfs device bind / unbind. We are not allowed manually
bind / unbind device from the driver currently.
* Rename some other variables.

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

[media] e4000: revise synthesizer calculation

Update synthesizer calculation to model I prefer nowadays. It is
mostly just renaming some variables to ones I think are most standard.
Also add 'schematic' of synthesizer following my current understanding.

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

[media] rtl2832_sdr: cleanup some set_bit() calls

This code works fine but static checkers complain.  The test_bit()
function takes the bit number and not a mask.  Then the other issue is
that we were using USB_STATE_URB_BUF which is BIT(0) instead of URB_BUF.
Also we were open coding that instead of using the test/clear/set_bit()
functions.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] cobalt: add new driver

The cobalt device is a PCIe card with 4 HDMI inputs (adv7604) and a
connector that can be used to hook up an adv7511 transmitter or an
adv7842 receiver daughterboard.

This device is used within Cisco but is sadly not available outside
of Cisco. Nevertheless it is a very interesting driver that can serve
as an example of how to support HDMI hardware and how to use the popular
adv devices.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] adv7604/adv7842: replace FMT_CHANGED by V4L2_DEVICE_NOTIFY_EVENT

This makes it easier for the bridge driver to just passthrough such
events to the corresponding device node.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>