| Age | Commit message (Collapse) | Author |
|---|
|
From: Trent Piepho <xyzzy@speakeasy.org>
The out of tree v4l-dvb build system uses slightly different compiler
flags than the kernel does. This caused problems when trying to compile
the out of tree version of the source symlinked into the kernel tree.
The workaround is to have the script that creates the symlinks also
create a new file called drivers/media/Kbuild, which the kernel build
system will use in preference to divers/media/Makefile. The new Kbuild
file exports the necessary EXTRA_CFLAGS to make building work and then
includes the real drivers/media/Makefile.
The dvb Makefiles were changed in a previous patch from 'EXTRA_CFLAGS='
to 'EXTRA_CFLAGS+=' so they wouldn't replace the new EXTRA_CFLAGS.
The symlink creating script is also re-written to make the symlinks
faster with less duplication of commands and avoid some unnecessary
links.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
The Makefiles in the dvb directory tree used '=' when defining EXTRA_CFLAGS
rather than '+=', which is far more common in the rest of the kernel source.
= 14 times (9 of which this patch removes)
:= 25 times
+= 123 times
This change also has certain advantages for the out of kernel v4l-dvb build
system.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Michael Krufky <mkrufky@linuxtv.org>
Although it is safe to kfree(NULL), We only need to kfree(priv)
if the release callback is undefined. As it stands now, there
is some redundancy in the operation of releasing the priv data
structures. This patch will call kfree(priv) and set priv to NULL,
if the release callback isnt defined. Otherwise, let the release
callback handle this itself.
Thanks to Mauro Carvalho Chehab for suggesting this.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Michael Krufky <mkrufky@linuxtv.org>
Define tuner release callbacks for mt20xx, tda9887 and tda8290, so that
these drivers can release their own private structures themselves.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Michael Krufky <mkrufky@linuxtv.org>
Individual tuner drivers are now allocating memory themselves for
their own private data structures. This changeset adds a release
callback to the tuner operations, so that newer drivers that may
require more complex data structures may release this private data
themselves.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
Michael Schimek requested the addition of inverted alpha framebuffer
caps/flags to support such hardware.
'Normal' alpha uses this formula to mix the framebuffer and video:
output = fb pixel * fb alpha + video pixel * (1 - fb alpha)
and the 'inverted' alpha uses this formula:
output = fb pixel * (1 - fb alpha) + video pixel * fb alpha
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
The last user of dvb_pll_configure was the dvb-usb function
dvb_usb_tuner_calc_regs(), which was nothing more than a wrapper around
dvb_pll_configure(). It's just a copy of the functionality provided by
the tuner_ops calc_regs method, and can be deleted.
There were two users of dvb_usb_tuner_calc_regs().
One was dvb_usb_tuner_set_params_i2c(), which is converted to use
fe->ops.tuner_ops.calc_regs().
The other was the digitv driver. This driver can use one of two demods,
mt352 or nxt6000. For the mt352, the driver would set
tuner_ops.calc_regs to dvb_usb_tuner_calc_regs(). We can just attach
dvb_pll and use the tuner_ops.calc_regs() provided by that module.
For the nxt600, the driver would set tuner_ops.set_params to
digitv_nxt6000_tuner_set_params. That function would in turn use
dvb_usb_tuner_calc_regs(). We convert it to use tuner_ops.calc_regs()
instead, and use dvb_pll_attach.
The digitv_tuner_attach() needs to know which frontend was attached by
digitv_frontend_attach(), since the nxt6000 needs tuner_ops.set_params()
to be overridden with digitv_nxt6000_tuner_set_params(). So to do this a
digitv_state that says which frontend was used is added to the
dvb_usb_device private state field.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
V4L/DVB tree keeps backward compatibility with vanilla 2.6.x kernels.
However, if some API changes happens on 2.6.x.y, this wouldn't be
handled.
This patch improves the compatibility capabilities of the tree by
allowing customized scripts to be added on make_config_compat.pl.
So, when generating config-compat.h, some compat checks can be done
against the kernel tree.
The practical effect is allowing compilation on 2.6.17.x trees, used by
several distros, where some changes on netdevice.h affects dvb-net,
stopping its compilation.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Hermann Pitton <hermann-pitton@arcor.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Hermann Pitton <hermann-pitton@arcor.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Installation logs were very prolific. This patch makes the installation message
cleaner, while preserving all information.
The generated log is now much more clear, especially if compiling just a few
device drivers.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Michael Krufky <mkrufky@linuxtv.org>
Now that cx88-blackbird is only accepting the official firmware image,
we no longer have any need to store the size of the firmware inside the
cx88 data structure.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
Remove temporary support for older 256 kB firmwares.
ivtv, pvrusb2 and blackbird can now all handle the newer larger firmwares,
so support for the older (buggier) firmware can be removed.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Michael Krufky <mkrufky@linuxtv.org>
Create private data struct for device specific private data.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Michael Krufky <mkrufky@linuxtv.org>
replace leading spaces with tabs
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
The VIDIOC_CROPCAP ioctl was missing in ivtv.
The handling of output video cropping was wrong. This has now been fixed.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
- Remove last use of V4L2_CAP_VIDEO_OUTPUT_POS from v4l2-ctl and v4l2-dbg.
- Add VIDIOC_CROPCAP support to v4l2-ctl.
- Extend VIDIOC_G/S_CROP support for other buffers besides video capture
in v4l2-ctl.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
V4L2_CAP_VIDEO_OUTPUT_POS was initially introduced for 2.6.22 but never
actually used: remove it before the final 2.6.22 is made.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
For backwards compatibility firmware files of 256 Kb were allowed: all
drivers have now been updated to support the newer larger firmwares so
remove this compatibility code and only support the newer firmware.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Oliver Endriss <o.endriss@gmx.de>
Several people reported unreliable reception with the current driver.
Furthermore, STR and SNR values seem to be inverted.
This fix is based on a patch posted by Hartmut Hackman.
Thanks to Helmut Auer for testing and helping to optimize the patch.
tda826x:
- set baseband cut-off to 19 MHz
tda10086:
- change the parameters of the carrier recovery loop
- toggle register 0x02 between 0x35 (tuning) and 0x00 (locked)
- invert STR and SNR values
Signed-off-by: Oliver Endriss <o.endriss@gmx.de>
Thanks-to: Hartmut Hackmann <hartmut.hackmann@t-online.de>
Thanks-to: Helmut Auer <vdr@helmutauer.de>
|
|
From: "Robert P. J. Day" <rpjday@mindspring.com>
Remove the long-dead references to the obsolete MODULE_PARM macro.
Given that the first one of those is actually misspelled as
"MODULE_PARAM", it's clear that they can't have been doing any good.
Signed-off-by: Robert P. J. Day <rpjday@mindspring.com>
Acked-by: Thierry Merle <thierry.merle@free.fr>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Thierry Merle <thierry.merle@free.fr>
- fixed the urb allocation part that was not taking into account the current alternate setting
this fixes usb_submit_urb returning -90 errno in isocIrq.
- fixed usb_submit_urb returning -1 errno in isocIrq (need to ignore usb urb with status==-ENOENT)
Acked-by: Dwaine P. Garden <dwainegarden@rogers.com>
Signed-off-by: Thierry Merle <thierry.merle@free.fr>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
Removed unnecessary .hardware from video_device struct.
Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Most of the v4l-dvb drivers now depend on the kernel Kconfig option
HAS_IOMEM, which isn't defined by kernels older than 2.6.22.
Assume HAS_IOMEM is true for older kernels to make the v4l-dvb Kconfig
files work.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
The recent changes on Kconfig broke compilation when VIDEO_DEV is compiled
as module. On some cases, drivers like VIDEO_BUF are compiled with 'y' option
instead of 'm':
ERROR: "videobuf_read_stop" [drivers/media/video/video-buf-dvb.ko] undefined!
ERROR: "videobuf_waiton" [drivers/media/video/video-buf-dvb.ko] undefined!
ERROR: "videobuf_read_start" [drivers/media/video/video-buf-dvb.ko] undefined!
ERROR: "videobuf_mmap_free" [drivers/media/video/saa7134/saa7134.ko] undefined!
ERROR: "videobuf_mmap_mapper" [drivers/media/video/saa7134/saa7134.ko] undefined!
...
ERROR: "videobuf_read_stop" [drivers/media/common/saa7146_vv.ko] undefined!
ERROR: "videobuf_poll_stream" [drivers/media/common/saa7146_vv.ko] undefined!
ERROR: "videobuf_streamoff" [drivers/media/common/saa7146_vv.ko] undefined!
Thanks to Toralf Forster <toralf.foerster@gmx.de> for pointing this issue.
CC: Toralf Forster <toralf.foerster@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Uwe Bugla <uwe.bugla@gmx.de>
This patch synchronizes the Documentation for bt8xx-based cards to the
actual state of kernel 2.6.22-rc1.
Signed-off-by: Uwe Bugla <uwe.bugla@gmx.de>
Signed-off-by: Manu Abraham <manu@linuxtv.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Russell King <rmk@arm.linux.org.uk>
If the SAA7111 device is powered down, and requires re-initialisation when
the V4L device is opened (as on the NetWinder), the SAA7111 driver gets
completely confused about the current settings.
The problem is centred around the way the driver keeps _two_ cached copies
of the current settings - one of the struct video_picture settings, and one
of the registers.
When the decoder is re-initailised, the cached register settings are
overwritten, as are the values in the hardware registers. However, the
cached video_picture settings are not.
Resolve this by removing the useless and buggy second level of caching for
video_picture. Instead, provide a function which updates register values
if and only if the value we are going to write to the register has changed.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Andrew Morton <akpm@linux-foundation.org>
If this mutex_lock_interruptible() does fail due to signal_pending() then the
state of the driver will get trashed in interesting ways, because userspace
cannot and will not retry the close().
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Dmitry Torokhov <dtor at mail.ru>
In preparation for struct class_device -> struct device input
core conversion switch to using input_dev->dev.parent when
specifying device position in sysfs tree.
Signed-off-by: Dmitry Torokhov <dtor at mail.ru>
Acked-by: Thierry Merle <thierry.merle@free.fr>
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Jonathan Corbet <corbet@lwn.net>
A bug in the ov7670 sensor causes it to introduce noise unless the CLKRC
register is rewritten *after* setting the image mode. Naturally,
resetting CLKRC in this way will cause other modes to fail. So
carefully poke the register only when indicated.
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Jonathan Corbet <corbet@lwn.net>
The proper method for powering down the sensor on OLPC systems has
changed somewhat; in particular, the sensor must be powered down
completely (rather than simply told to power down) or the associated
"camera active" LED will stay on.
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
