| Age | Commit message (Collapse) | Author |
|---|
|
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>
This Kconfig option was added in 2.6.23. On older kernels, use !PPC64 to set
it.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Recent patches to try to handle kernel object file directories haven't been
doing it correctly. This patch reworks the way the kernel location(s) are
handled.
KDIR et al are removed and replaced with two variables, SRCDIR and OUTDIR.
SRCDIR is the location of:
The main kernel Makefile
The kernel headers
The kernel source (if present)
OUTDIR is the location of:
The kernel .config file
The kernel [qxm]conf binaries (if present)
The following kernel situations should be handled correctly:
1. A kernel installed from source. One can specify the version via "make
release VER=version". If VER is not specified, uname -r is used. This the
default if make is run on a clean tree. OUTDIR is /lib/modules/$(VER)/build
and SRCDIR is /lib/modules/$(VER)/source, which are both symlinks pointing to
the same location.
2. A kernel installed from source which was was built with the
"O=/output/directory" option to the kernel Makefile. The version is specified
the same was as above. OUTDIR and SRCDIR are the same as above, except in
this case they are symlinks to different directories. OUTDIR will be a link
the the directory used in the "O=" option and SRCDIR will be the source
location.
3. An installed binary kernel with only a kernel-headers package. The
version can be specified the same way as above. OUTDIR will be
/lib/modules/$(VER)/build and SRCDIR will be the same as OUTDIR, as no
'source' directory exists. The kernel source and Kconfig program are most
likely not present.
4. An un-installed kernel source tree, for a kernel that was built without
using the "O=/outdir/directory" option. The kernel does not need to be fully
built, only the "modules_prepare" target needs to be built. The location of
the tree is specified with "make release DIR=directory". Both SRCDIR and
OUTDIR will be $(DIR).
5. An un-installed kernel source tree, for a kernel that was built using the
"O=/output/directory" option. The location of the tree is specified using
"make release DIR=/output/directory". Note that DIR is not the directory with
the kernel source! It is set to the directory that was used with the O option
when the kernel was built. The Makefile in this directory has a link back to
the kernel source directory and the v4l-dvb build system will find this.
OUTDIR will be the '/output/directory' and SRCDIR will be the location of the
source that built it.
All all these situation the v4l-dvb tree should build correctly and all the
various scripts will work. In addition, if the binaries for the [qxm]conf
programs are present, "make menuconfig", etc. will work as well. If the
binaries are not present, but the source is, the Makefile will automatically
build the binaries when "make menuconfig", etc. is used. The source should
be present in all situation except (3), a kernel-headers only install.
Note that if the kernel output directory (usually the same as the kernel
source directory, except for (2) and (5)) is not writable, then building the
Kconfig programs automatically will not be possible.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.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: 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>
|
|
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: 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: Trent Piepho <xyzzy@speakeasy.org>
Files changed in an mq patch weren't being checked.
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: 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: 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: 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>
|
|
From: Ludwig Nussel <ludwig.nussel@suse.de>
If defined, use $(DESTDIR) as a prefix for installing and removing
modules. depmod will also use $(DESTDIR) as a prefix when it is run as
part of module installation.
Signed-off-by: Ludwig Nussel <ludwig.nussel@suse.de>
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: Trent Piepho <xyzzy@speakeasy.org>
When parsing hgrc files, don't get confused by 'username' appearing in
other sections, like [smtp].
Check the repo's hgrc file first, then ~/.hgrc
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
tea5761 is a driver wrote back on 2005, to support Philips tea5671
radio, meanting to be used on some devices that would be lauched with
those chips.
Recently, some people seems to be interested on having support for those
chips at Linux Kernel.
Let's enable compilation of this driver and ask they to test if this
works fine.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
gentree.pl were part of Gerd Knorr scripts, used to generate patches
from his tree against kernel. Several newer capabilities were added to
the original script, including the ability to handle with multiple
levels of #if/#else/#endif and support for #elif.
This patch corrects the credits for this valuable piece of code.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Mostly re-write gentree.pl to make a few improvements:
About three times faster
Correctly processes BTTV_VERSION_CODE checks
Properly handles #elif
Doesn't need absolute path names specified in command line arguments
Improved debug output, which now shows input line number
Can handle almost any general expression for a conditional compilation
directive. Can understand all these:
#if 0
#ifdef BROKEN_XAWTV
#ifndef MM_KERNEL
#if defined(STV0297_CS2)
#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,20)
#elif defined(MM_KERNEL) || !BROKEN_XAWTV
#if BTTV_VERSION_CODE > KERNEL_VERSION(0,8,0) || defined(BROKEN_XAWTV)
#if (1 > 2) || (I2C_PEC && defined(I2C_CLASS_TV_ANALOG))
#ifndef CONFIG_XC3028 /* Keep */ <- knows to keep this one in the code
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Relaxed some regular expressions to better handle whitespaces at the
middle of cpp macros;
Added #elif processing capability to the script.
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>
The code to get the file list from hg manifest wasn't escaped right and
the error value check was the opposite of what it should have been.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
aci.c needs some OSS include files from the kernel to compile. Typical
kernel-headers packages don't have these include files, and so aci.c
fails to compile. Have make_kconfig.pl check for the headers, and
disable aci if they aren't there.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Skip the parsing code that only make sense inside a Kconfig menu block
when not in a menu block. Mostly to get rid of Perl warnings.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Since Kconfig.sound is processed by make_kconfig.pl like all the other
Kconfig files, it was getting put into v4l/Kconfig twice. Once from a
source directive v4l/Kconfig and again when it was processed by the
script.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
A subtle bug in the code that changed Kconfig depends expressions into
Perl expressions would keep changes to Kconfig variables from
propagating. The subroutine was supposed to change a _copy_ of the
expression from Kconfig to Perl syntax, but was changing the expression
itself.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
Restructure make_kconfig.pl and clean the code up some more.
The dependency tracking code is rewritten. This version converts the
"depends on" and "select X if" lines into Perl expressions, which are
then evaluated. This way all complex expressions are handled correctly.
Dependencies are tracked recursively until all are found. "select" lines
with an "if" clause are handled correctly; the selected variable is only
required when the "if" clause is true.
Another improvement is that hex/string/int variables that don't have
their dependencies met will not appear in the .config file. They aren't
supposed to.
make_kconfig will fail with an error message if MODULES is turned off in
the Kernel's config, as building out of tree makes no sense in that
situation.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
|
From: Trent Piepho <xyzzy@speakeasy.org>
The script which combines the individual Makefils into Makefile.media
would include a Makefile each time it was referenced. This caused some
Makefiles to get included multiple times.
Also print comments in Makefile.media with the names of the Makefiles that
are being included.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
|
