To avoid sleeps in providing buffers to user space and in handling incoming
buffers from the capture unit, converted the per stream mutex for locking
queues to 3 spin locks.  There is now a spin lock per queue
to increase concurrency when moving buffers around.

Also simplified queue manipulations and buffer handling of incoming buffers
of data from the capture unit.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

When sending an outgoing firmware command, prepare to wait before we raise the
interrupt, so we don't miss the wake_up() on the acknowledgment.  When waiting
for the acknowledgement, there is no need to loop around schedule(), as there
will only be one interrupt, and hence one wake_up(), issued.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This change defers sending all CX18_CPU_DE_SET_MDL commands, for a stream with
an ongoing capture, by adding a work queue to handle sending such commands when
needed.  This prevents any sleeps, caused by notifying the firmware of new
usable buffers, when a V4L2 application read() is being satisfied or when
an incoming buffer is processed by the cx18-NN-in work queue thread.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Rename the work queue to "in_work_queue" to indicate it is handling
incoming mailbox commands.  This is preparation for adding a work queue
for handling deferrable outgoing mailbox commands.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

* Return actual error values as returned by the i2c subsystem, rather
  than 0 or 1.
* If the registration of the second bus fails, unregister the first one
  before exiting, otherwise we are leaking resources.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Acked-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

A late v4l2_subdev framework change accidentally sent the audio input
routing value to the external multiplexer, instead of the muxer input routing
value to the external multiplexer.  This change corrects that error.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Remove support for the debug call VIDIOC_INT_S_AUDIO_ROUTING from cx18
and ivtv. These internal ioctls shouldn't be exposed. These were only
used through the cx18-ctl and ivtv-ctl utilities, and only when testing
a new card variant.

This cleanup allows the removal of this ioctl from v4l2-common.h.

Cc: Andy Walls <awalls@radix.net>
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

It is no longer needed to use a struct pointer as argument, since v4l2_subdev
doesn't require that ioctl-like approach anymore. Instead just pass the input,
output and config (new!) arguments directly.
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The functions v4l2_i2c_new_subdev and v4l2_i2c_new_probed_subdev relied on
i2c_get_adapdata to return the v4l2_device. However, this is not always
possible on embedded platforms. So modify the API to pass the v4l2_device
pointer explicitly.
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

s_std didn't belong in the tuner ops. Stricly speaking it should be part of
the video ops, but it is used by audio and tuner devices as well, so it is
more efficient to make it part of the core ops.
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The init callback was used in several places to load firmware. Make a separate
load_fw callback for that. This makes the code a lot more understandable.
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The audio and vbi parts still used an 'ioctl'-like interface. Replace this
with normal functions.

Cc: Andy Walls <awalls@radix.net>
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

There are several Kconfig items using CUSTOMIZE. Yet, most use the English
writing CUSTOMISE. This generates lots of trouble, because people sometimes type
the Kbuild item different.

Let's standardise every occurrence using the same syntax.

The changes were generated by this small shell script:

for i in `find linux -type f`; do sed s,CUSTOMIZE,CUSTOMISE,g $i >/tmp/a && mv /tmp/a $i; done
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Removed some unnecessary memcpy()'s by reworking the compress_*_vbi_buf()
functions.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The sliced VBI payloads that cx18 and ivtv would insert in the MPEG stream
did not have consistent endianess for the linemasks in the payload (a big
endian platform would write them out big endian).  This change ensures the
linemasks are always stored as little-endian in the MPEG stream to ensure
cross platform consistency in parsing the generated MPEG stream.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The cx18 private packet insertion code only expects to operate on an MPEG PS
when inserting private packets of sliced VBI data.  This patch keeps the cx18
driver from corrupting the MPEG TS or other non-PS stream types, in case the
user enables sliced VBI insertion for these MPEG stream types.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Based on a documentation clarification from Conexant, fix the register code
used for sliced VBI VPS service.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This change only affects comments.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

My s-parse builds never griped about be32_to_cpu() casting to __be32, but Hans'
builds did.  This change explictly declares the pointer into the VBI buffer
header as __be32, which is the correct thing to do as the data is always
big endian when we go to fetch it.  Hopefully this will quiet s-parse warnings.
Also corrected a misplaced parenthesis logic error in checking for the VBI
header magic number.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Fix a scaling check that was failing, due to a magic number I missed fixing
during previous slice VBI changes.  Now

  $ v4l2-ctl -v width=480,height=480,pixelformat=MPEG

yields proper visual results again.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Driver is now at version 1.1.0.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Remove variables that were holdovers from ivtv for supporting the CX23415
MPEG decoder output.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Give messages originating from internal subdevs a header using the subdev's
name.  Fixed an uninitialized variable problem with reading the EEPROM, noticed
from log output.  Got rid of the unused cx18_av_exit() function.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

We leaked buffers every time a device was removed, if the user had enabled
sliced VBI insertion into the MPEG stream.  MythTV uses that.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Convert GPIO line functions, such a audio routing and device resets, to
v4l2_subdevices.  This essentially completes the conversion of cx18 to the
v4l2_device/v4l2_subdevice framework.  No regression testing has taken place as
of yet.  Also an ivtv legacy bug with GPIO mux routing and going to/from radio
mode was commented, but not fixed.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This is a major perturbation to cx18 I2C device handling to convert it to the
v4l2_device/subdeivce framework.  This change breaks GPIO audio multiplexer
control for the time being.  It will be fixed in a coming change.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Added a new chip identifer to v4l2-chip-ident for the integrated A/V broadcast
decoder core internal to the CX23418.  Completed separation and encapsulation
of the A/V decoder core interface as a v4l2_subdevice.  The cx18 driver now
compiles and links again.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Convert card instance handling to a lighter weight mechanism like ivtv.
Also convert name strings and debug messages to use v4l2_device.name.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This is the next step in converting the cx18 driver to use the v4l2_device/
v4l2_subdevice framework.  This is a straightforward conversion of the
cx18_av_*[ch] files.  It compiles, but leaves the driver in an unlinkable
state at the moment.

Note, the cx18 integrated A/V digitizer will now make a host match at address 1,
as far as v4l2-dbg is concerned.  That means it identifies itself as a separate
"chip", and acts as an alias to the integrated A/V decoder registers that are
also available with the host match at address 0.
Signed-off-by: NAndy Walls <awalls@radix.net>
[mchehab@redhat.com: fix merge conflicts due to the removal of v4l2_ctrl_query_fill_std()]
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The v4l2_ctrl_query_fill_std() function wasn't one the best idea I ever had.
It doesn't add anything valuable that cannot be expressed equally well with
v4l2_ctrl_query_fill and only adds overhead.

Replace it with v4l2_ctrl_query_fill() everywhere it is used and remove it
from v4l2_common.c.
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Control arrays as are used with v4l2_ctrl_next must be sorted from
low to high. Add a comment at the top of all such arrays to warn
about this.
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The raw video device didn't report the image size correctly.

When setting a new image the image height has to be a multiple of 32 lines.
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Sliced VBI, in the manner that ivtv implements it as a separate data stream,
now works for 525 line 60 Hz systems like NTSC-M.  It may work for 625 line
50 Hz systems, but I have more engineering work to do, to verify it is operating
properly.  Sliced data insertion into the MPEG PS should be working, but is
untested.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The cx23418 appears to send Raw VBI buffers with a PTS on a per frame
basis, not per field, so process Raw VBI on a whole frame basis and reduce
some complexity.  Fix VBI buffer size computation to handle a whole
frame of Raw VBI for a 625 line system, which is the worst case and will
work for 525 lines systems as well.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Fixed an endianess problem with the collection of the PTS from
the VBI buffer given to us by the encoder.  Also extrapolated the
last 12 bytes of the last line of each field, to remove artifacts
created by removing the first 12 bytes of each field for raw VBI.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The driver was incorrectly setting 0 line counts in a firmware
API call to set the maximum amount of lines per field.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The Encoder will only allow the Raw VBI parameters, along with a number of
other API parameters, to take effect when no analog captures are in
progress.  These parameters must be set before the first analog capture starts,
be it MPEG, VBI, YUV, etc., and cannot be changed until the last one stops.  It
is not obvious to me what capture channel API parameters are shared and which
ones must be set per capture channel, so set them all for every analog
capture channel start up.  This fixes the driver so that VBI capture can be
started up after the MPEG capture is going.
Signed-off-by: NAndy Walls <awalls@radix.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>