Yann E. MORIN [Mon, 3 Jun 2013 10:15:48 +0000 (10:15 +0000)]
package/linux-firmware: bump version for new firmware
Bump version to get a few new firmware files for some WiFi adapters.
In the rtl81xx familly, it's hard to differentiate each members: I have
a USB stick that identifies itself as TRL8188CUS, but is handled by the
rtl8192ce driver. So I moved all of rtl81xx firmwares under the same
option to avoid confusion.
It will be the responsibility of the user to eventually get rid of
unwanted firmware files in a post-build script, if needed.
Gregory Hermant [Fri, 31 May 2013 06:31:08 +0000 (06:31 +0000)]
Add support for the Calao-systems TNY-A9G20-LPW
[Peter: lock kernel/headers version to 3.9.4, use devtmpfs, add comments] Signed-off-by: Gregory Hermant <[email protected]> Signed-off-by: Peter Korsgaard <[email protected]>
Patrick Ziegler [Tue, 28 May 2013 23:41:19 +0000 (23:41 +0000)]
ext-toolchain-wrapper: fix paths if executable was resolved by PATH
If ext-toolchain-wrapper or any symbolic link to it was resolved by PATH,
the wrapper takes the working directory to calculate the relative paths.
Now '/proc/self/exe' is used to resolve the absolute path to the toolchain
directory if the wrapper was called neither with a relative nor an absolute
path.
[Peter: fix off-by-one, swap value == var checks around] Signed-off-by: Patrick Ziegler <[email protected]> Tested-by: Thomas Petazzoni <[email protected]> Signed-off-by: Peter Korsgaard <[email protected]>
Thomas Petazzoni [Mon, 27 May 2013 13:01:02 +0000 (13:01 +0000)]
gutenprint: ensure target and host variant have the same patch hooks
In preparation for the out-of-tree support, it is important that all
packages have the same patch hooks for the host variant and the target
variant (because the source tree will be shared between host and
target variants when doing out-of-tree build).
In the case of gutenprint, the patch hook executed for the target just
creates one directory which is needed for autoreconf to work, and the
host variant is not autoreconf'ed. So there is no harm in also
applying this patch hook to the host variant.
[Peter: autoreconf for host as pointed out by Arnout] Signed-off-by: Thomas Petazzoni <[email protected]> Signed-off-by: Peter Korsgaard <[email protected]>
Thomas Petazzoni [Mon, 27 May 2013 13:01:04 +0000 (13:01 +0000)]
xlib_libX11: ensure target and host variant have the same patch hooks
In preparation for the out-of-tree support, it is important that all
packages have the same patch hooks for the host variant and the target
variant (because the source tree will be shared between host and
target variants when doing out-of-tree build).
In the case of the xlib_libX11 package, the post-patch hook removes
some flags from the build of the makekeys utility, so that when
xlib_libX11 is built for the target, the build of makekeys for the
host works properly. This modification can be applied to the host
variant of xlib_libX11 without any problem.
Thomas Petazzoni [Mon, 27 May 2013 13:01:03 +0000 (13:01 +0000)]
libgtk2: remove useless post-patch hook
There used to be a post-patch hook to apply a patch only on the host
variant of libgtk2. However, after the reorganization of the patch
naming policy, this was no longer working properly, so dd1848bc73ee9dfe211ca3ed3a233707e3bc5955 changed this by a patch that
works on both the target and host variants.
We can therefore get rid of the useless post-patch hook.
Thomas Petazzoni [Mon, 27 May 2013 13:01:01 +0000 (13:01 +0000)]
binutils: apply post-patch hook on both the host and target builds
When a Xtensa specific patch needs to be applied, it should be applied
both when doing the target and the host build.
This change is part of the effort to remove the differences between
host and target post-extract and post-patch hooks, that are
problematic for the out-of-tree support.
Thomas Petazzoni [Mon, 27 May 2013 13:01:00 +0000 (13:01 +0000)]
gettext: don't patch different the host and target build
Until now, the gettext source code was being patched differently for
the host and the target build, which is going to be incompatible with
out-of-tree build. The difference is that the gettext tools could be
disabled in the target build, but are always enabled on the host
build.
Therefore, this commit switches the gettext package to using proper
patches against the relevant configure.ac and Makefile.am
files. gettext is now being autoreconfigured. Configuration options
are also passed for the host variant, in order to disable the build of
a large number of things we don't care about.
Finally, the two existing patches are renamed to include a number in
their filename.
Will Wagner [Thu, 30 May 2013 13:08:05 +0000 (14:08 +0100)]
opengl: correct detection of virtual packages having no dependencies
I believe the original intent was to make it that the configure step
for the opengl virtual packages fails if there is not at least one
dependency. This patch fixes the logic so that it actually fails if
dependency list is empty
Thomas Petazzoni [Tue, 28 May 2013 10:27:06 +0000 (10:27 +0000)]
qt5jsbackend: only available on ARM, x86, x86-64 and mipsel
The v8 Javascript virtual machine that is the core of qt5jsbackend is
only available on certain architectures. At the moment, only ARM, x86,
x86-64 and mipsel are supported.
[Peter: also add dependency for qt5graphicaleffects] Signed-off-by: Thomas Petazzoni <[email protected]> Signed-off-by: Peter Korsgaard <[email protected]>
Thomas Petazzoni [Tue, 28 May 2013 09:45:35 +0000 (09:45 +0000)]
libtirpc: prevent selection when ADI Blackfin external toolchains are used
libtirpc fails to build with the Blackfin external toolchains provided
by Analog Devices, because their uClibc configuration doesn't include
AI_ADDRCONFIG. In order to prevent such build failures to happen, we
simply prevent libtirpc from being selected when such toolchains are
used. This is not a big problem, since they provide native RPC
support.
Also, since they provide native RPC, we don't have to propagate this
new dependency to the reverse dependencies of libtirpc, because they
all use native RPC when available. The exception to this rule is the
rpcbind package, which can only use the libtirpc implementation of
RPCs, and not the native one of C libraries. Therefore, the dependency
is propagated to the rpcbind package.
Thomas Petazzoni [Tue, 28 May 2013 09:45:34 +0000 (09:45 +0000)]
czmq: prevent selection when ADI Blackfin external toolchains are used
czmq fails to build with the Blackfin external toolchains provided by
Analog Devices, because their uClibc configuration doesn't include
AI_ADDRCONFIG. In order to prevent such build failures to happen, we
simply prevent the selection of czmq or one of its reverse
dependencies when such toolchains are used.
Thomas Petazzoni [Sun, 26 May 2013 12:08:30 +0000 (12:08 +0000)]
network-manager: bump to 0.9.8.0
This upstream version (the latest available at the time of this
writing) contains
http://cgit.freedesktop.org/NetworkManager/NetworkManager/commit/?id=6b64e4db2f3c9cfc0e0e240cf0bc58f3b3e90c1f
which fixes a build issue of network-manager against recent kernel
headers.
This patch adds a new package for the OpenGL, OpenVG and EGL
proprietary implementations specific to the Vivante GPU of i.MX6
processors.
[Peter: show comment on uClibc toolchains, tweak help text] Signed-off-by: Henk Fijnvandraat <[email protected]> Signed-off-by: Thomas Petazzoni <[email protected]> Signed-off-by: Peter Korsgaard <[email protected]>
This commit moves the firmware-imx package to the newly introduced
freescale-imx/ directory, and uses the new variables provided by
freescale-imx/freescale-imx.mk.
This commit moves the imx-lib package to the newly introduced
freescale-imx/ directory, and uses the new variables provided by
freescale-imx/freescale-imx.mk.
Yann E. MORIN [Tue, 28 May 2013 12:11:37 +0000 (12:11 +0000)]
fs/ext2: further fix to the UUID
Turned out that setting a nil-UUID is no better than clearing it.
What currently happens is as follows:
- first, genext2fs does not generate a UUID
- then we tune2fs to upgrade the filesystem
- then we run fsck, which generates a random UUID
- then we re-run tune2fs to set a nil-UUID
So, on the target, if the file system is improperly unmounted (eg.
with a power failure), on next boot, fsck may be run, and a new
random UUID will be generated.
*However*, fsck improperly updates the filesystem when it adds the
UUID, and there are a few group descriptor checksum errors.
Those errors will go undetected until the next fsck, which will then
block for user input (bad on embedded systems, bad).
Fix that by systematically generating a random UUID _before_ we call
to fsck.
A random UUID is not so bad, after all, since there are already so
many sources of unpredictability in the filesystem: files date and
ordering, files content (date, paths...) which renders a fixed UUID
unneeded.
And it is still possible to set the UUID in a post-image script if
needed, anyway.
[Peter: Note where wchar dep comes from, show comment if not available] Signed-off-by: Thomas Petazzoni <[email protected]> Signed-off-by: Peter Korsgaard <[email protected]>
Carlo Caione [Sun, 26 May 2013 12:08:22 +0000 (12:08 +0000)]
sunxi-boards: new package
[Peter: depend on arm, wrap help, change to _FEX_FILE option, error on empty] Signed-off-by: Carlo Caione <[email protected]> Signed-off-by: Thomas Petazzoni <[email protected]> Signed-off-by: Peter Korsgaard <[email protected]>
Thomas Petazzoni [Mon, 27 May 2013 13:00:59 +0000 (13:00 +0000)]
bzip2: improve the packaging
Instead of doing some nasty SED tricks, use a patch that modifies the
bzip2 build system to make it a little bit nicer. Those SEDs were made
differently for the host and target packages, which is contrary to the
logic of not having any difference in the source tree between the host
build and the target build.
This commit also makes sure that bzip2.mk behaves properly in a
BR2_PREFER_STATIC_LIB=y context.
Thomas Petazzoni [Mon, 27 May 2013 13:03:45 +0000 (13:03 +0000)]
gdb: apply post-patch hook on both the host and target builds
When a Xtensa specific patch needs to be applied, it should be applied
both when doing the target and the host build.
This change is part of the effort to remove the differences between
host and target post-extract and post-patch hooks, that are
problematic for the out-of-tree support.
This commit adds a patch to fix this build failure, related to the
access of the ULLONG_MAX define on compilers that did not default to
the C99 variant of the C language.
Thomas Petazzoni [Tue, 28 May 2013 01:05:47 +0000 (01:05 +0000)]
flex: fix autobuilder failures
Since fe6a9e5e9d (flex: needs M4 at runtime), the autobuilders have
been producing a number of flex related build failures. They have been
hard to track down, because even on the same machine, with the same
Git commit ID and the same configuration, the failure could not be
reproduced.
However, a close inspection of flex's config.log file allowed to find
out what the problem was. In its configure script, flex uses the
host-flex to generate a minimal example, and find out the name of the
output file of flex.
When the M4 environment is passed when building the target flex, it
also affects the *execution* of the host-flex, which tries to use
/usr/bin/m4 (which doesn't exist in the autobuilder machines) instead
of the one built in $(HOST_DIR)/usr/bin/m4. So generating the minimal
example fails. And this is where what I could reproduce and what the
autobuilders script produce differ: in my case, even though host-flex
fails to run, it creates an empty lex.yy.c, which is enough to make
the configure script happy. In the context of the autobuild scripts,
this file is apparently not created at all, for an unknown reason, and
this leads to the configure script to abort.
The fix is to set ac_cv_path_M4. This will affect the default m4 used
by the target flex, but it will not affect the m4 used by the
host-flex. It allows the test made during the configure script to work
properly, and therefore should fix the issue seen in the autobuilders.
Yann E. MORIN [Tue, 28 May 2013 00:33:35 +0000 (00:33 +0000)]
fs/ext2: explicitly set a NULL UUID
"tune2fs -U clear" creates an invalid filesystem, that fsck.ext2
whines about later:
$ make rootfs-ext2
[--SNIP--]
$ ./host/usr/sbin/fsck.ext4 images/rootfs.ext2
e2fsck 1.42.7 (21-Jan-2013)
Filesystem did not have a UUID; generating one.
images/rootfs.ext2: clean, 4616/5120 files, 53171/131072 blocks
$ ./host/usr/sbin/fsck.ext4 -f images/rootfs.ext2
e2fsck 1.42.7 (21-Jan-2013)
One or more block group descriptor checksums are invalid. Fix<y>? yes
Group descriptor 0 checksum is 0x4131, should be 0x8bdb. FIXED.
[--SNIP--]
Thomas Petazzoni [Sun, 26 May 2013 12:08:17 +0000 (12:08 +0000)]
flex: needs M4 at runtime
For proper runtime execution, flex requires m4 to be
installed. Passing a M4 variable at configure time is needed,
otherwise flex on the target will try to use a 'm4' binary with a
build machine path.
ADI officially supports the buildroot and related GNU toolchain for
Blackfin since ADI's 2012R1 release only. In order to avoid confusion,
it is better to remove the 2011R1 GNU toolchain for Blackfin. In
addition, the 2011R1 GNU toolchain for Blackfin doesn't support the
BF60x processors.
Thomas Petazzoni [Sun, 26 May 2013 12:08:13 +0000 (12:08 +0000)]
elf2flt: fix build when zlib is not installed on the host
For a reason that's fairly unclear to me, Peter added a '-lz' link
flag to the elf2flt.mk build in d5664ee99 ("elf2flt: fix link").
However, the zlib library may not necessarily be installed on the host
machine, so we should depend on host-zlib, and pass the appropriate
LDFLAGS. This is what this patch does.
Thomas Petazzoni [Sun, 26 May 2013 12:08:12 +0000 (12:08 +0000)]
dropbear: factorize the 'depends on BR2_PACKAGE_DROPBEAR'
Enclose all dropbear sub-options into a 'if BR2_PACKAGE_DROPBEAR'
... 'endif' block rather than having 'depends on BR2_PACKAGE_DROPBEAR'
for each option.
Will Newton [Sun, 26 May 2013 12:08:11 +0000 (12:08 +0000)]
numactl: new package
[Thomas: updated to 2.0.8, restricted to supported architectures,
added licensing informations, updated to latest Buildroot package
infrastructure]
[Peter: Fix LICENSE_FILES typo, needs largefile support in toolchain] Signed-off-by: Will Newton <[email protected]> Signed-off-by: Thomas Petazzoni <[email protected]> Signed-off-by: Peter Korsgaard <[email protected]>
projects / buildroot-mgba.git / log