pxa_dma_update_sg_tail() is called only once, runs exactly the same loop as the
caller and has to recalculate the last element in an sg-list, that the caller
has already calculated. Eliminate redundancy by merging the two loops and
re-using the calculated pointer. This also saves a bit of performance which is
always good during video-capture.
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The last buffer queued will often overrun, as the DMA chain
is finished, and the time the dma irq handler is activated,
the QCI fifos are filled by the sensor.
The fix is to ignore the overrun condition on the last
queued buffer, and restart the capture only on intermediate
buffers of the chain.
Moreover, a fix was added to the very unlikely condition
where in YUV422P mode, one channel overruns while another
completes at the very same time. The capture is restarted
after the overrun as before, but the other channel
completion is now ignored.
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The DMA transfers in pxa_camera showed some weaknesses in
multiple queued buffers context :
- poll/select problem
The bug shows up with capture_example tool from v4l2 hg
tree. The process just "stalls" on a "select timeout".
- multiple buffers DMA starting
When multiple buffers were queued, the DMA channels were
always started right away. This is not optimal, as a
special case appears when the first EOF was not yet
reached, and the DMA channels were prematurely started.
- Maintainability
DMA code was a bit obfuscated. Rationalize the code to be
easily maintainable by anyone.
- DMA hot chaining
DMA is not stopped anymore to queue a buffer, the buffer
is queued with DMA running. As a tribute, a corner case
exists where chaining happens while DMA finishes the
chain, and the capture is restarted to deal with the
missed link buffer.
This patch attemps to address these issues / improvements.
create mode 100644 Documentation/video4linux/pxa_camera.txt
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
All planes were PAGE aligned (ie. 4096 bytes aligned). This
is not consistent with YUV422 format, which requires Y, U
and V planes glued together. The new implementation forces
the alignement on 8 bytes (DMA requirement), which is almost
always the case (granted by width x height being a multiple
of 8).
The test cases include tests in both YUV422 and RGB565 :
- a picture of size 111 x 111 (cross RAM pages example)
- a picture of size 1023 x 4 in (under 1 RAM page)
- a picture of size 1024 x 4 in (exactly 1 RAM page)
- a picture of size 1025 x 4 in (over 1 RAM page)
- a picture of size 1280 x 1024 (many RAM pages)
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Due to DMA constraints, the DMA chain always transfers bytes
from the QCI fifos to memory in 8 bytes units. In planar
formats, that could mean 0 padding between Y and U plane
(and between U and V plane), which is against YUV422P
standard.
Therefore, a frame size is required to be a multiple of 16
(so U plane size is a multiple of 8). It is enforced in
try_fmt() and set_fmt() primitives, be aligning height then
width on 4 multiples as need be, to reach a 16 multiple.
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
CUSTOMISE appears to be the one actually defined...
Signed-off-by: Kyle McMartin <kyle@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial: (28 commits)
trivial: Update my email address
trivial: NULL noise: drivers/mtd/tests/mtd_*test.c
trivial: NULL noise: drivers/media/dvb/frontends/drx397xD_fw.h
trivial: Fix misspelling of "Celsius".
trivial: remove unused variable 'path' in alloc_file()
trivial: fix a pdlfush -> pdflush typo in comment
trivial: jbd header comment typo fix for JBD_PARANOID_IOFAIL
trivial: wusb: Storage class should be before const qualifier
trivial: drivers/char/bsr.c: Storage class should be before const qualifier
trivial: h8300: Storage class should be before const qualifier
trivial: fix where cgroup documentation is not correctly referred to
trivial: Give the right path in Documentation example
trivial: MTD: remove EOL from MODULE_DESCRIPTION
trivial: Fix typo in bio_split()'s documentation
trivial: PWM: fix of #endif comment
trivial: fix typos/grammar errors in Kconfig texts
trivial: Fix misspelling of firmware
trivial: cgroups: documentation typo and spelling corrections
trivial: Update contact info for Jochen Hein
trivial: fix typo "resgister" -> "register"
...
Setting ->owner as done currently (pde->owner = THIS_MODULE) is racy
as correctly noted at bug #12454. Someone can lookup entry with NULL
->owner, thus not pinning enything, and release it later resulting
in module refcount underflow.
We can keep ->owner and supply it at registration time like ->proc_fops
and ->data.
But this leaves ->owner as easy-manipulative field (just one C assignment)
and somebody will forget to unpin previous/pin current module when
switching ->owner. ->proc_fops is declared as "const" which should give
some thoughts.
->read_proc/->write_proc were just fixed to not require ->owner for
protection.
rmmod'ed directories will be empty and return "." and ".." -- no harm.
And directories with tricky enough readdir and lookup shouldn't be modular.
We definitely don't want such modular code.
Removing ->owner will also make PDE smaller.
So, let's nuke it.
Kudos to Jeff Layton for reminding about this, let's say, oversight.
http://bugzilla.kernel.org/show_bug.cgi?id=12454
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
The tuner ops pointer wasn't set, so s_std never worked here.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
I2C_DRIVERIDs are no longer needed with v4l2_subdev and will be removed
from the kernel soon.
Cc: Devin Heitmueller <dheitmueller@linuxtv.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
No i2c modules are called with the V4L1 API anymore, so this can be removed.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Replace 'VIDIOC_' references in v4l i2c drivers by their new v4l2_subdev
callback names.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro 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: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The audio and vbi functions where still called through an ioctl-like
interface, even though this is no longer needed with v4l2-subdev. Just
change each 'case' into a proper function and call that directly.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
tuner-core.c contains a hack for cx88 board to prevent probing of certain
addresses:
/* HACK: Ignore 0x6b and 0x6f on cx88 boards.
* FusionHDTV5 RT Gold has an ir receiver at 0x6b
* and an RTC at 0x6f which can get corrupted if probed.
*/
With the new i2c API this hack no longer works. So instead change the
list of tuner probe addresses in the cx88 driver itself, which is much
more clean.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Several i2c drivers still used v4l2_subdev_command, even though they were
converted to v4l2_subdev. Remove those unused .command callbacks.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
tda7432 is no longer used with the old-style i2c API, so I can remove
the support for that.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
tda9875 is no longer used with the old-style i2c API, so I can remove
the support for that.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
All drivers that use wm8775 are now converted to v4l2_subdev, so I can remove
the support for the old-style i2c API.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Convert cx88 to use v4l2_subdev since the old i2c autoprobing mechanism
will be removed.
Added code to explicitly load tvaudio where needed. Also fix the rtc-isl1208
support: since that driver no longer supports autoprobing it has to be
loaded using the new i2c API.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
All drivers that use cx25840 are now converted to v4l2_subdev, so I can remove
the support for the old-style i2c API.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Convert this driver to v4l2_subdev. Note that currently the only card with
analog support in this driver is the HVR-1800. The analog tuner support in
this driver is limited to what is needed for this board. When analog support
is added for other cards, then the tuner load code will probably have to be
expanded to take care of those boards. For example, there is currently no
support for either radio tuners or tda9887 demods.
I'd like to thank Steven Toth for testing this on his HVR-1800.
Tested-by: Steven Toth <stoth@linuxtv.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
If the firmware failed to be found the error message indicated
the incorrect filename.
Signed-off-by: Steven Toth <stoth@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
When enumerating formats with VIDIOC_ENUM_FMT the uvcvideo driver does not
fill the reserved fields of the struct v4l2_fmtdesc with zeros as required by
V4L2 API revision 0.24 [1]. Add the missing initializations.
The patch was tested with v4l-test 0.10 [2] with CNF7129 webcam found on EeePC
901.
References:
[1] V4L2 API specification, revision 0.24
http://v4l2spec.bytesex.org/spec/r8367.htm
[2] v4l-test: Test environment for Video For Linux Two API
http://v4l-test.sourceforge.net/
Signed-off-by: Márton Németh <nm127@freemail.hu>
Signed-off-by: Laurent Pinchart <laurent.pinchart@skynet.be>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This patch introduces support for DVB-T for the following dibcom based card:
Elgato EyeTV DTT (USB-ID: 0fd9:0021)
Signed-off-by: Klaus Flittner <klaus@flittner.org>
Signed-off-by: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This patch adds support for the Yuan MC770 DVB-T (1164:0871).
Thanks to Xoan Loureiro.
Signed-off-by: Xoan Loureiro <x04n2.0@gmail.com>
Signed-off-by: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This is the second part of the code cleanup changing the usb and pci-driver cores.
Signed-off-by: Uwe Bugla <uwe.bugla@gmx.de>
Signed-off-by: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This patch cleans up the source code of the b2c2 flexcop-driver. It is the first of a total of two.
The code is now passing the checkpatch-script.
Signed-off-by: Uwe Bugla <uwe.bugla@gmx.de>
Signed-off-by: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
When resetting the PID-filter block a safety msleep(1) was injected. Now that the function is called from interrupt context, it was replaced with a udelay. It seems that this delay is not necessary at all, let's remove it for now and test a while.
Signed-off-by: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This patch adds a workaround in the i2c-code of the flexcop-driver to
fix support for SkyStar2 rev 2.7. There are not many devices out
there, that's why this bug was not revealed earlier.
Signed-off-by: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
saa6588 is now only used through v4l2_subdev, so we can remove the old legacy
code.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The Terratec Cinergy 600 TV MK3 supports the RDS decoder saa6588.
Add support to saa7134 for such devices.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
saa6588 can also be used by other drivers than just bttv. Move it to a
new RDS decoders category and add it as helper chip to bttv.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Since tda9875 support was merged into tvaudio the bttv driver no
longer needs tda9875 as helper driver.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Since i2c autoprobing will disappear bttv needs to be converted to use
v4l2_subdev instead.
Without autoprobing the autoload module option has become obsolete. A warning
is generated if it is set, but it is otherwise ignored.
Since the bttv card definitions are of questionable value a new option was
introduced to allow the user to control which audio module is selected:
msp3400, tda7432 or tvaudio (or none at all).
By default bttv will use the card definitions and fallback on tvaudio as the
last resort.
If no audio device was found a warning is printed.
The saa6588 RDS device is now also explicitly probed since it is no longer
possible to autoprobe it. A new saa6588 module option was added to override
the card definition since I suspect more cards have this device than one
would guess from the card definitions.
Note that the probe addresses of the i2c modules are hardcoded in this
driver. Once all v4l drivers are converted to v4l2_subdev this will be
cleaned up. Such data belongs in an i2c driver header.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
In the tvaudio_remove function del_timer_sync(&chip->wt) is called.
However, chip->wt isn't always initialized depending on the type of
audio chip. Since del_timer_sync hangs when given an uninitialized timer
we should always initialize it.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This change allows bttv to use tvaudio for this device. Since this device
has the same i2c address as the tda9874 we need to support both in the same
tvaudio driver. This makes it possible for tvaudio to detect which chip is
used. Originally the tda9875 was only available in the dedicated tda9875
driver, but that makes life very hard for bttv since loading tvaudio might
misdetect a tda9875 as a tda9874.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The mute control depends on CHIP_HAS_INPUTSEL, so test for that first.
The s/g_tuner code should check whether getmode/setmode is set at the
beginning instead of filling in the struct and discovering at
the end that this chip doesn't implement audiomodes after all (i.e. is
a simple muxer chip).
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The ->vidioc_(s|g|try|enum)_fmt_vid_cap() methods are only called on
VIDEO_CAPTURE buffers. Thus, there is no need to check or set the buffer's
'type' field since it must already be set to VIDEO_CAPTURE. Checking the
buffer type can be removed from zr364xx_vidioc_(s|g|try|enum)_fmt_vid_cap().
The v4l2 core code in v4l2_ioctl will zero out the structure the driver is
supposed to fill in for read-only ioctls. For read/write ioctls, all the
fields which aren't supplied from userspace will be zeroed out.
Zeroing code can be removed from zr364xx_vidioc_querycap(),
zr364xx_vidioc_enum_input(), zr364xx_vidioc_enum_fmt_vid_cap(), and
zr364xx_vidioc_g_fmt_vid_cap().
Cc: Antoine Jacquet <royale@zerezo.com>
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The v4l2-ioctl core only allows buffer types for which the corresponding
->vidioc_try_fmt_xxx() methods are defined to be used with
vidioc_(q|dq|query)buf() and vidioc_reqbufs().
Since this driver only defines ->vidioc_try_fmt_vid_cap() the checks can be
removed from vidioc_reqbufs(), vidioc_qbuf(), and vidioc_dqbuf().
The ->vidioc_(s|g|try|enum)_fmt_vid_cap() methods are only called on
VIDEO_CAPTURE buffers. Thus, there is no need to check or set the buffer's
'type' field since it must already be set to VIDEO_CAPTURE. So setting the
buffer type in vidioc_enum_fmt_vid_cap() can be removed.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The ->vidioc_(s|g|try|enum)_fmt_vid_cap() methods are only called on
VIDEO_CAPTURE buffers. Thus, there is no need to check or set the buffer's
'type' field since it must already be set to VIDEO_CAPTURE.
The v4l2-ioctl core only allows buffer types for which the corresponding
->vidioc_try_fmt_xxx() methods are defined to be used with
vidioc_(q|dq|query)bufs() and vidioc_reqbufs().
Since this driver only defines ->vidioc_try_fmt_vid_cap() the checks can be
removed from vidioc_reqbufs(), vidioc_querybuf(), vidioc_qbuf(), and
vidioc_dqbuf(). Also, the buffer index is unsigned so it's not necessary
to check if it is less than zero.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The ->vidioc_(s|g|try|enum)_fmt_vid_cap() methods are only called on
VIDEO_CAPTURE buffers. Thus, there is no need to check or set the 'type'
field since it must already be set to VIDEO_CAPTURE.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The v4l2 core code in v4l2_ioctl will zero out the structure the driver is
supposed to fill in for read-only ioctls. For read/write ioctls, all the
fields which aren't supplied from userspace will be zeroed out.
Zeroing code is removed from enum_input, g_tuner, g_frequency, querycap,
enum_fmt_vid_cap, g_fmt_vid_cap, and try_fmt_vid_cap.
The ->vidioc_(s|g|try|enum)_fmt_vid_cap() methods are only called on
VIDEO_CAPTURE buffers. Thus, there is no need to check or set the buffer's
'type' field since it must already be set to VIDEO_CAPTURE.
There also appeared to be a copy and paste error in
vidioc_try_fmt_vid_cap() that would set f->fmt.pix.sizeimage to zero.
Note that the s_fmt_vid_cap method doesn't appear to actually do anything.
Whatever parameters were requested are just silently ignored. Was this
intentional? Who knows, as the commit log entry for the driver just says,
"Add generic cx23417 hardware encoder support." There are no docs. A
comment like "this driver totally ignores the v4l2 spec w.r.t.
VIDIOC_S_FMT because ..." would have gone a long way.
Cc: Steven Toth <stoth@hauppauge.com>
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The v4l2-ioctl core only allows buffer types for which the corresponding
->vidioc_try_fmt_xxx() methods are defined to be used with
vidioc_(q|dq|query)bufs() and vidioc_reqbufs().
This driver only defines ->vidioc_try_fmt_vid_cap() so only VIDEO_CAPTURE
buffers are allowed to be used with cafe_vidioc_reqbufs(),
cafe_vidioc_querybuf(), cafe_vidioc_qbuf(), and cafe_vidioc_dqbuf().
The ->vidioc_(s|g|try|enum)_fmt_vid_cap() methods are only called on
VIDEO_CAPTURE buffers. Thus, there is no need to check or set the buffer's
'type' field since it must already be set to VIDEO_CAPTURE. So the
check in cafe_vidioc_enum_fmt_vid_cap() can be removed.
The 'index' field of v4l2_buffer is unsigned so the checks for it being
less than zero can be removed too.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The v4l2-ioctl core only allows buffer types for which the corresponding
->vidioc_try_fmt_xxx() methods are defined to be used in vidioc_(g|s)_parm,
vidioc_(q|dq|query)buf, and vidioc_reqbufs.
Remove buffer type checking from vino_g_parm(), vino_s_parm(),
vino_reqbufs(), vino_querybuf(), vino_qbuf(), and vino_dqbuf(). This
reduced the indent level of the code so a few lines can be wrapped better.
Also fixed the C++ type comments.
The v4l2-ioctl core also provides structs that have been pre-zeroed for all
fields that driver is supposed to fill in, so remove zeroing code from
vino_enum_fmt_vid_cap(). Also, the format index is unsigned so it's not
necessary to check if it's less than zero.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
