The driver should now pass the 'busy' state of the device to the cx2341x
module whenever controls are set or tried. -EBUSY will be returned if
the device is busy and the user attempts to modify certain 'dangerous'
controls. It concerns controls that change the audio or video
compression mode and bitrates.
The cx88-blackbird and pvrusb2 drivers currently always pass '0' (not busy)
to the cx2341x, effectively keeping the old behavior for now.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Changing the VBI format requires a CX2341X_ENC_INITIALIZE_INPUT firmware
call. This can only be done if no capture is in progress. So return
-EBUSY if the encoder is busy.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
- Always clear when stopping the decoder
- Clear if the filehandle that is being close was used for UDMA_YUV output.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
The driver has long claimed to support mmap, but it didn't work at all. Some
of the dma buffer parameters weren't set, and since video_buf uses vmalloc to
allocate the buffer, a page callback is needed too.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
The driver kepts a static global array of snd_card pointers for each card
probed, which was never used.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Rework the way the DMA buffer is handled and IRQs are generated.
ALSA uses a ring-buffer of multiple periods. Each period is supposed to
corrispond to one IRQ.
The existing driver was generating one interrupt per ring-buffer, as opposed
to per period. This meant that as soon as the IRQ was generated, the hardware
was already starting to re-write the beginning of the buffer. Since the DMA
happens on a per-line basis, there was only a narrow window to copy the data
out before the buffer was overwritten.
The cx88 core RISC program generator is modified so that it can set the IRQ
and counter flags to count every X lines of DMA transfer. This way we can
generate an interrupt every period instead of every full ring-buffer. Right
now only period of one line are supported, but it should be possible to
support longer periods. Note that a WRITE instruction generates an IRQ when
it starts, not when the transfer is finished. Thus to generate an IRQ when
line X is done, one must set the IRQ flag on the instruction that starts line
X+1, not the one that ends line X.
Change the line size so that there are four lines in the SRAM FIFO. If there
are not four lines, the analog output from the cx88's internal DACs is full of
clicks and pops.
Try to handle FIFO sync errors. Sometimes the chip generates many of these
errors before audio data starts. Up to 50 sync errors will be ignored and the
counter reset.
Have the IRQ handler save the RISC counter to the chip struct, and then have
the pointer callback use this to calculate the pointer position. We could
read the counter from the pointer callback, but sometimes the sync errors on
start up cause the counter to go crazy. ALSA sees this and thinks there has
been an overrun. The IRQ hander can avoid saving the counter position on
sync errors.
The chip "opened" flag wasn't necessary. ALSA won't try to open the same
substream multiple times. Probably this code was cut&pasted from the bt87x
driver, which has multiple sub-streams for one chip.
Do error checking for the videobuf mapping functions.
snd_card_cx88_runtime_free() is useless and can be deleted.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Convert the initialization of ivtv_i2c_algo_template to C99-style.
Otherwise a future change to struct i2c_algo_bit_data would break
this code.
Also declared that structure const - it's a template after all.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Since at least kernel 2.6.12-rc2, module.h includes moduleparm.h. This
patch removes all occurences of moduleparm.h from drivers/media files.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
The file, tuner-driver.h was originally named tuner-hw.h, but we decided to
rename it. At the time, I had forgotten to change the #ifdef __TUNER_HW_H__
to #ifdef __TUNER_DRIVER_H__ . This patch corrects that.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Add DVB-T support for Avermedia Super 007
Analog television is untested. The device lacks input adapters for radio,
svideo & composite -- seems to be a DVB-T ONLY device.
Signed-off-by: Edgar Simo <bobbens@gmail.com>
Acked-by: Hermann Pitton <hermann-pitton@arcor.de>
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
This is a simple whitespace cleanup.
Signed-off-by: Edgar Simo <bobbens@gmail.com>
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
The revision is part of the pci_dev struct, so there is no need to read it in.
Stop storing the revision and latency in the chip struct, since they're never
used after being printed out when the driver loads.
linux/pci.h wasn't included. It was getting picked up something else,
probably cx88.h, but this file uses struct pci_dev so it should include pci.h.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
channels_min should be 2, not 1. The hardware only supports stereo.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Currently the driver turns on audio interrupts, then sets the audio interrupt
mask to select which interrupts to get. One could received unwanted
interrupts since the mask is set _after_ interrupts have already been turned
on. Change the order of the operations, and clear any audio interrupt status
bits that are already set for good measure.
Before changing the SRAM FIFO parameters, make sure the FIFO isn't being used.
This shouldn't happen with just the ALSA driver, as it should never try to
turn on FIFO/RISC/DMA while they are already on. However, the V4L driver
needs to turn the audio FIFO on for analog audio output to work (undocumented
cx88 bug). The FIFO parameters are in an inconsistent state while they are
updated, and this results in many FIFO sync error IRQs if the FIFO is in use
while it's in this inconsistent state.
Also create and use a bunch of symbolic constants for audio interrupt mask
bits.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
When an irq handled by the cx88 core driver (currently IR and errors) occurs
and the cx88-alsa irq handler is the first called, it will claim to have
handled the irq but it doesn't call cx88_core_irq() to handle it.
The means loading cx88-alsa can disable the IR remote.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Used for the PCI_INTMSK and PCI_INT_STAT registers.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Add VIDIOC_OVERLAY to enable/disable the OSD. Also add the OSD state to the
log status report.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
The return value of ivtv_udma_fill_sg_list() was changed by the
highmem bugfix, but that return value was still used in ivtv-yuv.c.
Revert to the old return value, but in addition return -1 in case of
a memory allocation error.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
When trying to DMA userspace buffers to the cx23415 you need to check
whether the page is in highmem. If this is the case, then bounce buffers
have to be used to allow DMA.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Serialization is now done on the open/close/ioctl level and also when the
read/write/poll start an encoder/decoder stream.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
- add FILE debug flag for open/close/read/write/poll.
- show cmd for encoder/decoder command ioctl.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Some streams (PCM, VBI decoding) do not need that much memory, so specify
the allocated memory in kB instead of MB to limit memory usage. E.g. 1 MB
is overkill for the VBI decoding stream, 64 kB is enough.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
The TV output standard was set only on first use, which meant that the saa7127
was set to NTSC until then, leading to flickering on PAL systems.
Since the saa7127 has no firmware it is OK to initialize it immediately.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Whenever the 0x80b register is used the microcontroller should be reset.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
- Give up frame after three retries.
- When the last capture/decode ends, make sure to delete the dma_timer.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
- Invalid VBI packets should result in an empty VBI frame, not
in an zero-sized frame that causes the reader to incorrectly
return a 0 (EOF) value.
- PIO completion should not reset the sg_pending_size field.
- The DMA offset detection code should be ignored for PIO transfers:
it somehow messes up the data on the card and is not needed anyway
for PIO.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
It turns out that the cx23415/6 DMA engine cannot do scatter/gather DMA
reliably. Every so often depending on the phase of the moon and your
hardware configuration the cx2341x DMA engine simply chokes on it and
you have to reboot to get it working again.
This change replaced the scatter/gather DMA by single transfers at a time,
where the driver is now responsible for DMA-ing each buffer.
UDMA is still done using scatter/gather DMA, that will be fixed soon.
Many thanks to Mark Bryars <mark.bryars@etvinteractive.com> for discovering
the link between scatter/gather and the DMA timeouts.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
When there are no more free buffers, then buffers are stolen from the
predma queue. Buffers should be stolen from the head of that queue (which
is where the most recently added buffers are) and all buffers belonging
to a frame should be stolen. Otherwise 'half-frames' would remain in the
queue, which leads to ugly playback and complete sync failure for YUV
buffers.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
The CX2341X_ENC_INITIALIZE_INPUT firmware call requires careful handling,
otherwise the computer can freeze or the top-third of the screen can start
flickering. This patch ensures that CX2341X_ENC_INITIALIZE_INPUT is called
at the right time and in the right way.
In addition the stop capture handling was improved so that the last pending
DMA transfer is also processed. Otherwise this would be the first data that
arrived when a new capture was started which is not what you want.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
FM radio transmission use a preemphasis/deemphasis scheme to reduce
high-frequency noise. The cx88 audio decoder is supposedly set to no
deemphasis by the current driver.
However, the "no deemphasis" setting doesn't work. On my chip, cx23883, it
produces the same result as the 75 us time constant. Maybe the default
settings on the cx23881 are for 50 us?
Since the deemphasis time constant varies by country, allow setting it via a
module parameter.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Get rid of the "CORE" prefix from cx88 printks. It was only used a few times,
and it makes it look like they're coming from the kernel core or something.
Fix the message, "TV tuner 60 at 0x1fe, Radio tuner -1 at 0x1fe", by adding a
"cx88[0]" prefix to be consistent, and to keep people who grep their dmesg
output for cx88 from missing it. Get rid of the addresses, which are always
wrong. The addresses are always set to -1, but because it's an unsigned 8-bit
value, the left shift converts it to the nonsense address 0x1fe.
In the cx8802 driver, some cut and pasted code prefixed lines with "CORE
cx88[0]:", which has been changed to "cx88[0]/2:" like the other printks from
the cx8802 driver.
Also fix some ugly printks in the cx8802 driver that used __FUNCTION__ for
KERN_INFO and KERN_ERR messages.
The changed printks in cx88-mpeg.c also needed lots of whitespace and
80-column fixes.
A bunch of misc changes in cx88-dvb.c and cx88-video.c to add message levels
or a consistent "cx88[?]/2" or "cx88[?]/0" prefix.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
A lot of code in cx88-cards.c was only used by cx88-core.c when the core state
is first allocated and initialized. Moving that task to cx88-cards makes the
driver simpler and the files more self contained.
- Module parameters tuner, radio, card, and latency move to cx88-cards.c
- cx88_boards is made static
- cx88_subids is made static and const
- cx88_bcount is eliminated
- cx88_idcount is eliminated
- cx88_card_list() is made static
- cx88_card_setup_pre_i2c() is made static
- cx88_card_setup() is made static
- cx88_pci_quirks() is moved from cx88-core to cx88-cards
The function argument "char *name" is made const too
- get_ressources() is moved from cx88-core to cx88-cards, and renamed to
cx88_get_resources()
- The code to allocate and initialize the core state struct and the chip is
moved out of cx88-core.c:cx88_get_core() and into a new function in
cx88-cards.c, cx88_core_create(). This makes both functions simpler.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
The cx88 driver state stored the ID of the board type in core->board. Every
time the driver need to get some information about the board configuration, it
uses the board number as an index into board configuration array.
This patch changes it so that the board number is in core->boardnr, and
core->board is a copy of the board configuration information. This allows
access to board information without the extra indirection. e.g.
cx88_boards[core->board].mpeg becomes core->board.mpeg.
This has a number of advantages:
- The code is simpler to write.
- It compiles to be smaller and faster, without needing the extra array lookup
to get at the board information.
- The cx88_boards array no longer needs to be exported to all cx88 modules.
- The boards array can be made const
- It should be possible to avoid keeping the (large) cx88_boards array around
after the module is loaded.
- If module parameters or eeprom info override some board configuration
setting, it's not necessary to modify the boards array, which would
affect all boards of the same type.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
An error message for PCI resource allocation failure used the board type
before it was set. Just get rid of the error message, as get_ressources()
[sic] already prints one. Format that error message better, and add the pci
function and subsystem information to better associate the error with what
caused it.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
The vmux setting is only two bits, but was taking up a whole 32
in the input description struct. By changing it to a two-bit
bitfield, it can fit in what was padding space before and drop
the input size by 4 bytes, from 28 to 24. This drops the board
description struct, which has 9 inputs, from 280 to 244 bytes.
Total driver size decreases by 2108 bytes.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
tuner-simple is the only sub-driver that uses last_div, so we
can free up two bytes of memory for all other tuners, by moving
this into tuner-simple's private data area.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Some lines later filp->private_data is initialized to dev again.
Since there are some checks that might fail in the mean time
keep the later version.
Signed-off-by: Rolf Eike Beer <eike-kernel@sf-tec.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Correctly attribute the origin of the driver to Kazuhiko
Kawakami.
It took some time to get the S-O-B line from the original
tvaudio patch author, but here it is.
Signed-off-by: Kazuhiko Kawakami <kazz-0@mail.goo.ne.jp>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
