| Age | Commit message (Collapse) | Author |
|---|
|
From: Maxim Levitsky <maximlevitsky@gmail.com>
/lib/modules/$(uname -r)/source isn't available if only kernel-headers
were installed. use /lib/modules/$(uname -r)/build in that case
Signed-off-by: Maxim Levitsky <maximlevitsky@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
checkpatch.pl is the official tool for checking CodingStyle violations on kernel
patches. This changeset makes it run every time a "make commit" is called. This
allows detecting CodingStyle violations during development phase.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Kay Sievers <kay.sievers@vrfy.org>
The currently used "struct class_device" will be removed from the
kernel. Here is a patch that converts all users in drivers/media/video/
to struct device.
Reviewed-by: Thierry Merle <thierry.merle@free.fr>
Reviewed-by: Mike Isely <isely@pobox.com>
Reviewed-by: Luca Risolia <luca.risolia@studio.unibo.it>
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
If a kconfig option goes away in the v4l-dvb Kconfig files, but is still
present in the kernel Kconfig files, then v4l-dvb code using that option
would appear to be ok.
This is fixed by not reading the drivers/media Kconfig files from the kernel
source and merging the options defined there with those from the v4l-dvb
version. If a Kconfig file is read from v4l-dvb, the kernel version of that
file is not read.
Fix a regex bug that made #ifdef CONFIG_WHATEVER_MODULE appear to be missing
when WHATEVER did exist.
Check #ifndef too.
Get Kconfig options from menuconfig lines too.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Much better now. Also handles some more error conditions and prints a bit
more information.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
Currently it is impossible to build v4l drivers if kernel output
directoty is not the same as kernel source directory
This patch attempts to fix that.
Signed-off-by: Maxim Levitsky <maximlevitsky@gmail.com>
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>
A few distro kernels, like Mandriva 2.6.22.9-desktop-1mdv are starting to be
shipped with the newer alsa drivers. This patch changes the compat code for
older kernels in a way that it will be compatible also with those distro
kernels.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
- Preserve original email's date;
- Handles reviewed-by tag;
- avoids upperscase troubles when processing tags.
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>
based on cx88.pl -- minimal changes required for cx23885
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Michael Krufky <mkrufky@linuxtv.org>
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: Michael Krufky <mkrufky@linuxtv.org>
Break tuner.ko into separate modules. This was a quick change -
Tuner sub-drivers are still static-linked to tuner.ko, this will
change after using dvb_attach and removing the probing functions.
After this change, one can deselect undesired tuner sub-drivers via Kconfig.
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
Acked-by: Hans Verkuil <hverkuil@xs4all.nl>
Acked-by: Mike Isely <isely@pobox.com>
Acked-by: Steven Toth <stoth@hauppauge.com>
Acked-by: Patrick Boettcher <pb@linuxtv.org>
Acked-by: Jarod Wilson <jwilson@redhat.com>
Acked-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Those scripts were used with the legacy patch submission.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: 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>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
The current compat code for device_create/destroy doesn't work properly for
kernels 2.6.14 or below, and 2.6.18 to 2.6.23.
For kernels 2.6.14 and below, the device_create/destroy compat code wasn't
enabled. Those kernels need it too.
For kernels 2.6.18 to 2.6.23, compat code _was_ getting including, which
caused errors since 2.6.18 and above already have device_create and
device_destroy.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Removed some ALSA compat code from bt87x and put it into compat.h, then use
this code in cx88-alsa and saa7134-alsa to avoid lots and lots of #ifs.
All the kernel version checks in cx88-alsa and saa7134 were off by one too,
which is now fixed.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
class_device_create/class_device_destroy are being replaced by
device_create/device_destroy. This patch make it backward compatible.
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: Trent Piepho <xyzzy@speakeasy.org>
Files changed in an mq patch weren't being checked.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
If you have tools like "dkms" installed, you may lose your kernel
config at kernel source tree. The new "init" target allows
recreating kernel basic init files from V4L tree.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
---
v4l/Makefile | 5 +++++
1 file changed, 5 insertions(+)
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
This device is internal to the Panasonic VP27S tuner and is used to
set the mono/stereo/bilingual setting of the tuner.
It is used by two Japanese cx23416-based cards.
Signed-off-by: Takahiro Adachi <tadachi@tadachi-net.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
The function wasn't added until after 2.6.22.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
In 2.6.23, there is a new struct pci_dev->revision field to get the pci
revision. A couple drivers now use this to print the revision and on old
kernels don't print it anymore. This patch adds a simple function to get
the revision to compat.h to make it easy for the backward compay code to
keep printing the revision like is used too.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Thanks to Michael Krufky for pointing this issue.
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>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Add scripts/hg-pull-req.pl, which generates a nicely formatted pull
request. Mention it in README.patches.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
The first argument changed in 2.6.13, but it can be make backward compatible
with a macro.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
CONFIG_BIGPHYS_AREA is an out-of-tree patch. While it may be interesting
to have this configuration flag when compiling v4l-dvb out-of-tree, for
development purposes, there's no sense on keeping those unused code
inside kernel.
This patch do some cosmetic changes on zoran_driver.c, preserving both
supports there, and adds the defaults to be used by gentree.pl. This
way, gentree.pl will handle the removal of the dead code when submitting
code to mainstream, while keeping the complete code at v4l-dvb tree.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
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: 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: 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: 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>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
A number of drivers depend on "!VIDEO_HELPER_CHIPS_AUTO", so when VHCA is on,
they are disabled. Make the allyesconfig default be VHCA off so these drivers
will be enabled with allyesconfig.
Even though these drivers are disabled when VHCA is on, it is still possible
to enable them via a "select", at which point a driver will be on even though
its dependencies are not met. Which seems like it doesn't follow the
specification.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
A backported kernel patch added a "depends on" clause to a menu, which wasn't
yet handled.
Everything in a menu gains implicitly gains a copy of the menu's dependencies
We handle this the same way as if/endif is handled. It may not work properly
if someone makes a nonsense file like:
menu "foo"
if X
endmenu
endif
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
kernel-sync:
Backport the original patch from Mattia Dongili <malattia@linux.it>
meye: make meye use sony-laptop instead of sonypi
Change sonypi_camera_command() calls to sony_pic_camera_command() and use
the renamed macros.
Also, added compat stuff to allow compiling meye driver for kernels < 2.6.22
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Deal with string options and the default values.
There is only one string option, and it's disabled, but someone could
always add more.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Update make_myconfig.pl and make_kconfig.pl to handle menuconfig options the
same way as config options.
Update make_kconfig.pl to process if/endif directives. All config options
inside an if/endif block gain a dependency on the 'if' expression.
Make the script a little more efficient by not trying to match a line after it
has already found a match for it. This helped to enhance debug mode, so
that it will print lines which were not understood rather than silently
skipping them.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Changed usbvision cards table to allow:
1) Not repeat USB ID on two structs;
2) Not need to specify both usb and card description tables at
the same order, removing some magic;
Some cards had duplicated names. Fixed.
A test for an specific board were doing by using a string comparation.
The comparation were wrong. Also, it is not a good practice to recognize
a board based on his string name.
Acked-by: Thierry MERLE <thierry.merle@free.fr>
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: Trent Piepho <xyzzy@speakeasy.org>
Annoyingly, the Hg mq extension will strip lines starting with "From: "
from the patch header. Since the v4l-dvb format commit message contains
a line starting with "From: ", it will mess them up.
This changes prep_commit_msg to create a decoy "from" line to keep mq from
messing up commit messages.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
Adds a qrefresh target to the top-level Makefile. It's like commit, but
for use with the Hg mq extension. It will execute hg qrefresh, so you
should have a current mq patch created with hg qnew already!
It will run the whitespace and cardlist scripts. If you do not have a
commit message from the current mq patch, it will create a default one,
as make commit does. If one already exists, then it will put a comment
at the beginning of the existing message with the current diffstat and a
note of any changes the whitespace script made. The diffstat is done
with qdiff, which will show the total diffstat of the patch instead of
just the diffstat of from the last version of the patch to the current
version.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
In one case a Kbuild file would do something like:
foo-objs := a.o b.o
foo-objs += c.o
analyze_build didn't understand that the second lines _adds_ c.o to the
objects that make up foo.ko. It would think that foo.ko was made up of
only c.o. Should be fixed now.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
