Michael Krufky [Tue, 22 Apr 2008 17:46:12 +0000 (14:46 -0300)]
V4L/DVB (7407): tuner-simple: add module options to specify rf input
Add module options to tuner-simple, called "atv_input" and "dtv_input"
to specify which rf input to use on devices with multiple rf inputs.
If the module option is not specified, then the driver will autoselect
the rf input, as per previous behavior.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
This patch fixes several kernel oops, when unplugging device while it is in
use:
Basically the patch delays freeing of the internal variables in
si470x_usb_driver_disconnect, until the the last user closed the device in
si470x_fops_release. This was implemented a while ago with the help of Oliver
Neukum.
I tested the patch five times (unplugging while in use) without oops coming
from the radio-si470x driver anymore. A remaining oops was coming from the
usbaudio driver, but this is someone else task. Hopefully this fixed all
unplugging issues.
V4L/DVB (7399): Removes video_dev from tuner-xc2028 config struct
The video_dev parameter, on tuner-xc2028 were used to pass i2c private data to
tuner_callback. Since the driver already have a pointer to i2c_adap->algo_data,
uses this instead.
This parameter were used also as a magic number to idenfity if two drivers are
trying to register the same xc3028 tuner. This occurs with boards with DVB
support, where both DVB and V4L drivers will share the same tuner.
Instead of using the algo_data as a private number, after this patch, the
driver will use i2c_adap->dev, with seems more consistent.
V4L/DVB (7393): tda827x: fixed support of tuners with LNA
Tuner refactoring broke support of tuners with LNA configurations 1 and 2
for both, analog TV and DVB-T.
Additionally, this patch initializes the saa713x gpios defined by the gpiomask
at driver init to avoid undefined stated at dvb.
V4L/DVB (7392): saa7134: support 2nd DVB-S section of the MD8800
There are some restrictions:
- The 2nd DVB-S section will only work if the 1st is configured for DVB-S too.
so "options saa7134-dvb use_frontend=0,1" won't work.
- Currently it is not possible to set the higher LNB supply voltages, so
14V instead of 13V in the 2nd section.
- It is not possibe to turn off the 2nd LNB supply independently.
This comes from the problem that the 2nd section can't access the i2c interface
of the LNB supply chip.
V4L/DVB (7390): saa7134: clear audio DSP interface after access error
In the case of an access error to the high latency registers of
the audio DSP, the interface needs to be cleared, otherwise a cascade
of errors occurs.
This patch is closely modeled after a proposal by Mirek Slugen
drivers/media/video/bt8xx/bttv-cards.c:3030:38: warning: "/*" within comment
drivers/media/video/bt8xx/bttv-cards.c:3032:20: warning: "/*" within comment
Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Michael Krufky [Tue, 22 Apr 2008 17:46:04 +0000 (14:46 -0300)]
V4L/DVB (7379): tuner: prevent instance sharing if i2c adapter is NULL
We currently do not have a method to enable instance staring if i2c adapter
is NULL, in the cases of dvb demods that write to the tuner directly using
calc_regs. Prevent possible wrong instance sharing for these cases until
a better solution can be found.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
flags used for spinlocks don't need to be initialized, except where the
compiler has no way to see, that the spin_unlock_irqrestore is only called
if the spin_lock_irqsave has been called before. Local variable
initialization doesn't have to be protected.
V4L/DVB (7376): Improve compile-time type-checking in videobuf
Make the dev member of the struct videobuf_queue of type "struct device *"
to avoid future problems. Also change the prototype of the
videobuf_queue_core_init() function.
V4L/DVB (7375): cx88/saa7134: fix magic number for xc3028 reusage detection
tuner-xc2028 needs to know when a DVB module is sharing the same analog tuner.
This is done by comparing a magic number that needs to be the same on analog
and on digital. To make easier, this magic number is a pointer to some data
struct.
With the previous code, two different pointers were using, causing a
miss-detection.
It has an audio connector coming out from a chip labeled
ATMEL
0242
AT89C2051-24PI
It is identified as follows on my Debian GNU/Linux Etch (kernel 2.6.18)
...
01:0a.0 Multimedia video controller: Brooktree Corporation Bt878 Video Capture (rev 11)
01:0a.1 Multimedia controller: Brooktree Corporation Bt878 Audio Capture (rev 11)
...
01:0a.0 0400: 109e:036e (rev 11)
Subsystem: 008a:763c
Flags: bus master, medium devsel, latency 64, IRQ 58
Memory at dfffe000 (32-bit, prefetchable) [size=4K]
Capabilities: [44] Vital Product Data
Capabilities: [4c] Power Management version 2
01:0a.1 0480: 109e:0878 (rev 11)
Subsystem: 008a:763c
Flags: bus master, medium devsel, latency 64, IRQ 58
Memory at dffff000 (32-bit, prefetchable) [size=4K]
Capabilities: [44] Vital Product Data
Capabilities: [4c] Power Management version 2
It was being detected as a GENERIC UNKNOWN CARD both by the 2.6.18
kernel and the latest v4l-dvb drivers, but it did not work at all. The
card has sixteen (16) BNC video inputs, four of them on the board itself
and twelve on three daughter-cards. It has a single bt878 chip, no tuner
and what looks like and audio input. After doing some research I managed
to get only eight channels working by forcing card=125 and those DID NOT
match channels 0-7 on the card, and no audio.
Based on what was working for card=125, I added the card definition
block, added a specific muxsel routine and got the card working fully
with xawtv, where the sixteen channels show up as Composite0 to
Composite15, matching the channel labels in the card and daughter-cards.
I have made no efforts yet to get audio working, but would appreciate
any pointers.
Marcin Slusarz [Tue, 22 Apr 2008 17:45:57 +0000 (14:45 -0300)]
V4L/DVB (7365): reduce stack usage of v4l1_compat_sync
poll_one allocated on stack struct poll_wqueues which is pretty big
structure (>500 bytes on x86_64). v4l1_compat_sync invokes poll_one
in a loop, so allocate struct poll_wqueues in v4l1_compat_sync (with
kmalloc) and pass it to poll_one.
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Marcin Slusarz [Tue, 22 Apr 2008 17:45:57 +0000 (14:45 -0300)]
V4L/DVB (7364): reduce stack usage of v4l_compat_translate_ioctl
v4l_compat_translate_ioctl used 1376 bytes of stack (x86_64),
so split this 800 lines long function into ~20 small noinline functions;
the biggest function takes now 712 bytes (v4l1_compat_sync)
fix VIDIOCSWIN handler which printked wrong errors
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Michael Krufky [Tue, 22 Apr 2008 17:45:56 +0000 (14:45 -0300)]
V4L/DVB (7360): tuner-simple: fix return value of simple_dvb_configure
simple_dvb_configure returns the actual tuned frequency to its caller, so
it must be declared as a u32 rather than an int. As a result, we will
return 0 to indicate a failure.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Michael Krufky [Tue, 22 Apr 2008 17:45:52 +0000 (14:45 -0300)]
V4L/DVB (7348): tuner-simple: do not send i2c commands if there is no i2c adapter
if (priv->i2c_props.adap == NULL) then exit any function that would
send commands over the i2c bus. We allow drivers to attach without an
i2c adapter for cases where the dvb demod accesses the tuner directly
via calc_regs.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Michael Krufky [Tue, 22 Apr 2008 17:45:51 +0000 (14:45 -0300)]
V4L/DVB (7346): tda9887: allow multiple driver instances for the same hardware to share state
Convert tda9887 to use the new hybrid_tuner_request_state and
hybrid_tuner_release_state macros to manage state sharing between
hybrid tuner instances.
Some ATSC/DVB cards need to put the analog demodulator into standby
before tuning digital. This patch allows us to attach the tda9887
driver to the digital side of the bridge driver and be able to put
it into standby without jeopardizing the analog demod driver's state.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Hans Verkuil [Tue, 22 Apr 2008 17:45:50 +0000 (14:45 -0300)]
V4L/DVB (7341): ivtv: rename tunerhz to tunertype
There are two tuner types: those for M/N standards and those for all others.
However, M/N standards are not always 60 Hz (PAL-N/Nc are 50 Hz), so rename
the module option accordingly.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Improve hardware parameter negotiation between the camera host driver and
camera drivers. Parameters like horizontal and vertical synchronisation,
pixel clock polarity shall be set depending on capabilities of the
parties.
V4L/DVB (7329): add flag to allow software demux to recognize the output type
Previously, the macro DVR_FEED, which is used to recognize and filter
out duplicate packets going to the DVR device, used the TS_PAYLOAD_ONLY
flag to identify a packet's destination.
This kind of filtering was introduced by the following two changesets:
Now, that it is possible to record TS PIDs using the demux device by
setting the output type to DMX_OUT_TSDEMUX_TAP, checking TS_PAYLOAD_ONLY
is not sufficient anymore. Therefore another flag, TS_DEMUX, is added to
specify the output type of a feed.
This allows multiple clients to filter the same TS PID on a demux device
simultaneously.
Signed-off-by: Andreas Oberritter <obi@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Michael Krufky [Tue, 22 Apr 2008 17:45:45 +0000 (14:45 -0300)]
V4L/DVB (7323): pvrusb2: set default video standard to NTSC for OnAir HDTV / Creator
The pvrusb2 driver normally picks up the default video standard from the
eeprom on Hauppauge devices, but the OnAir HDTV and OnAir Creator are not
Hauppauge devices, and do not store this information in any eeprom.
These devices support NTSC/ATSC, so we should use NTSC by default when in
analog mode.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Mike Isely [Tue, 22 Apr 2008 17:45:45 +0000 (14:45 -0300)]
V4L/DVB (7321): pvrusb2: Rework context handling and initialization
This change significantly rearranges pvr2_context level initialization
and operation:
1. A new kernel thread is set up for management of the context.
2. Destruction of the pvr2_context instance is moved into the kernel
thread. No other context is able to remove the instance; doing
this simplifies lock handling.
3. The callback into pvrusb2-main, which is used to trigger
initialization of each interface, is now issued from this kernel
thread. Previously it had been indirectly issued out of the work
queue thread in pvr2_hdw, which led to deadlock issues if the
interface needed to change a control setting (which in turn
requires dispatch of another work queue entry).
4. Callbacks into the interfaces (via the pvr2_channel structure) are
now issued strictly from this thread. The net result of this is
that such callback functions can now also safely operate driver
controls without deadlocking the work queue. (At the moment this
is not actually a problem, but I'm anticipating issues with this in
the future).
5. There is no longer any need for anyone to enter / exit the
pvr2_context structure. Implementation of the kernel thread here
allows this all to be internal now, simplifying other logic.
6. A very very longstanding issue involving a mutex deadlock between
the pvrusb2 driver and v4l should now be solved. The deadlock
involved the pvr2_context mutex and a globals-protecting mutex in
v4l. During initialization the driver would take the pvr2_context
mutex first then the v4l2 interface would register with v4l and
implicitly take the v4l mutex. Later when v4l would call back into
the driver, the two mutexes could possibly be taken in the opposite
order, a situation that can lead to deadlock. In practice this
really wasn't an issue unless a v4l app tried to start VERY early
after the driver appeared. However it still needed to be solved,
and with the use of the kernel thread relieving need for
pvr2_context mutex, the problem should be finally solved.
Signed-off-by: Mike Isely <isely@pobox.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Mike Isely [Tue, 22 Apr 2008 17:45:45 +0000 (14:45 -0300)]
V4L/DVB (7320): pvrusb2: Eliminate timer race during tear-down
The pvrusb2 tear-down logic was clearing two timers before stopping
its internal work queue. That left a tiny window open where the work
queue might run after the timers are stopped, possibly starting them
again. This could lead to dangling pointers and an oops. Solution:
Kill the work queue first, then delete the timers.
Signed-off-by: Mike Isely <isely@pobox.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Mike Isely [Tue, 22 Apr 2008 17:45:44 +0000 (14:45 -0300)]
V4L/DVB (7319): pvrusb2: Close potential race condition during initialization
There is a callback that is issued to into pvr2_context from pvr2_hdw
after initialization is done. There was a probability that this
callback could get missed. Fixed.
Signed-off-by: Mike Isely <isely@pobox.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
projects / linux-2.6 / log