Fix various messages displayed by these scripts:
- make start-stop-daemon quiet in order to avoid extra messages like
"stopped /usr/sbin/dhcpd (pid 174)" being output between the command
description and its result,
- fix the script names in the usage strings.
The dhcpd daemon does not require network interface names to be
specified on the command line.
>From dhcpd(8):
"The names of the network interfaces on which dhcpd should listen for
broadcasts may be specified on the command line. This should be done
on systems where dhcpd is unable to identify non-broadcast interfaces,
but should not be required on other systems. If no interface names
are specified on the command line dhcpd will identify all network
interfaces which are up, eliminating non-broadcast interfaces if
possible, and listen for DHCP broadcasts on each interface."
dhcpd exits with "Not configured to listen on any interfaces!" only if
no requested (those in INTERFACES, or all if empty) non-broadcast
interfaces matching the subnet declarations in dhcpd.conf are up.
Add an OPTIONS configuration variable in order to make it possible to
pass custom extra options to dhcpd. This keeps the systemd support
consistent with the SysV init script.
package/dhcp: make S80dhcp-relay read a file from /etc/default/
The S80dhcp-relay init script has configuration variables like
INTERFACES whose contents have to be passed to the daemon. These
variables are initialized as empty strings, but some of them are not
allowed to be empty and there was no means of filling them apart from
creating a root FS overlay to overwrite these scripts.
This commit adds support for reading dhcrelay under /etc/default/ to
set these configuration variables.
[Thomas: adapt to patch only S80dhcp-relay, since S80dhcp-server has
already been changed by previous commits.]
Thomas Petazzoni [Thu, 24 Dec 2015 13:48:28 +0000 (14:48 +0100)]
dhcp: read /etc/default/dhcpd, not /etc/default/dhcpd.conf
All daemons that read a file from /etc/default/ have it named just
after the name of daemon, without any extension. This commit fixes the
dhcp package to do the same.
Rodrigo Rebello [Thu, 24 Dec 2015 11:35:34 +0000 (09:35 -0200)]
doom-wad: bump to version 1.9
Also add a hash file.
The extraction method had to be changed, since the WAD file for this
version is not distributed as a stand-alone file, but comes inside a
2-part DOS self-extracting zip archive.
Thomas Petazzoni [Thu, 24 Dec 2015 10:18:27 +0000 (11:18 +0100)]
libnspr: fix build with the musl C library
Even though libnspr uses a configure script, it does look at __GLIBC__
and other macros to determine whether certain C library features are
available or not. Such things fails completely with the musl C
library, since it doesn't define __GLIBC__.
To fix this, we borrow some logic from Alpine Linux, which consists in
explicitly specifying which features are available in the musl C
library.
Thomas Petazzoni [Thu, 24 Dec 2015 10:07:53 +0000 (11:07 +0100)]
autofs: really fix build
The autofs configure script does a number of weird things.
First, instead of using the default search path for AC_CHECK_PROGS(),
it overrides it its own search path, ignoring the PATH environment
variable. Due to this, autofs fails to find flex even though flex is
available in the PATH. Fortunately, the authors provided a --with-path
configure option which allows to override this search path.
In addition, the configure script is checking on the build machine
that a number of programs are available, while those programs are in
fact only needed on the target. The one causing build issues is
modprobe, because it may not necessarily be installed on build
machines, so we explicitly tell autofs that modprobe is located in
/sbin/modprobe. It might be necessary to provide additional values for
other programs, as we discover other problems.
Noé Rubinstein [Wed, 4 Nov 2015 17:47:05 +0000 (18:47 +0100)]
tpm-tools: new package
[Thomas:
- move from Libraries/Crypto to System tools in menuconfig
- add patch to fix build with uClibc
- use 'select' for openssl rather than a 'depends on'
- propagate trousers dependencies
- add <pkg>_LICENSE and <pkg>_LICENSEFILES
- link against libiconv when needed, fixing !locale builds
- explicitly disable pkcs11 support, since it is not handled by
tpm-tools.mk currently.]
Noé Rubinstein [Wed, 4 Nov 2015 17:47:04 +0000 (18:47 +0100)]
trousers: new package
[Thomas:
- use a 'select' instead of a 'depends on' for openssl
- select libiconv when locale support is not available, otherwise the
build fails due to missing iconv.h
- disable on BR2_STATIC_LIBS configurations as it doesn't build. It
could be made to build by patching the trousers Makefile.
- add missing dependency on thread support in the toolchain
- add Config.in comment about the thread and dynamic library
dependencies.
- add missing <pkg>_LICENSE and <pkg>_LICENSE_FILES variables.]
Rodrigo Rebello [Fri, 6 Nov 2015 19:47:57 +0000 (17:47 -0200)]
chocolate-doom: new package
Chocolate Doom is a set of conservative source ports for Doom, Heretic,
Hexen and Strife, with a philosophy of preserving the look, feel, and
bugs of the vanilla versions of each.
Martin Bark [Wed, 23 Dec 2015 12:16:05 +0000 (12:16 +0000)]
package/nodejs: fixed comment for mips soft float
nodejs is only selected for mips hard float however the comment is
displayed for hard and soft float. Correct this by adding the missing
!BR2_MIPS_SOFT_FLOAT to the comment.
Martin Bark [Wed, 23 Dec 2015 12:16:03 +0000 (12:16 +0000)]
package/nodejs: removed version 0.12.9
Remove 0.12.9 to rationalise the number of nodejs releases supported by
buildroot. Going forward buildroot will only support the latest release
of nodejs and the 0.10.x branch for armv5 support.
Martin Bark [Wed, 23 Dec 2015 12:16:02 +0000 (12:16 +0000)]
package/nodejs: bump version to 5.3.0
Replace 4.2.3 with 5.3.0. 5.3.0 is the current Stable release. See
https://nodejs.org/en/blog/release/v5.3.0 for details on the release.
Copied 4.2.3 patched to 5.3.0 with the following exceptions:
- Removed 0004-fix-arm-vfpv2.patch, committed upstream and included in
5.3.0. See https://github.com/nodejs/node/commit/84dea1bd0c
- Added 0004-Fix-va_list-not-declared.patch, fix for a new bug. This
has already been fixed upstream but is not in 5.3.0
Rodrigo Rebello [Wed, 23 Dec 2015 15:38:56 +0000 (13:38 -0200)]
vnstat: bump to version 1.15
Also:
- Remove no longer needed patch (merged upstream).
- Adapt vnstat.mk to the new autotools-based build system.
- Include patch to improve gd library usability check + add
configure option to explicitly disable that check.
Sam bobroff [Tue, 22 Dec 2015 21:22:00 +0000 (22:22 +0100)]
core/pkg-kconfig: handle generated defconfigs
As of Linux kernel commit ea4d1a8 "powerpc/configs: Replace
pseries_le_defconfig with a Makefile target using merge_config" some
kconfig defconfigs (one so far: "pseries_le_defconfig") can be
generated using "make <defconfig>" and they do not exist on disk as a
single defconfig file.
This causes buildroot's build to fail for those configs with:
'arch/powerpc/configs/pseries_le_defconfig' for 'linux' does not exist
To handle this case and keep the makefile steps as simple as possible,
introduce a new package variable, *_KCONFIG_DEFCONFIG, that can be
used to indicate that a defconfig rule is being used, rather than a
file.
This allows the rule that generates the .config file to use either the
provided file (by copying) or a generated defconfig (by running "make
<defconfig>") as its starting point. merge_config.sh can then be run the
same way in either case, using .config as both input and output.
Note that merge_config.sh is now modifying .config in-place but this
is safe because it uses a temporary copy while making changes.
This patch introduces the new variable but does not make use of it.
Use of the new variable will be introduced in separate patches.
Signed-off-by: Sam Bobroff <[email protected]>
[[email protected]:
- include all the other kconfig-package hunks into this one patch
- do not transform the 'echo "..."; exit 1' into $(error ...) calls
- use a make $(if)-block instead of a shell if-block to copy the file
or run make (like is done to check the kconfig snippets)
- misc typoes and rephrasing in the commit log
- do not force the _defconfig suffix in the infra (Thomas)
] Signed-off-by: "Yann E. MORIN" <[email protected]> Cc: Thomas Petazzoni <[email protected]> Signed-off-by: Thomas Petazzoni <[email protected]>
Thomas Petazzoni [Tue, 22 Dec 2015 21:35:20 +0000 (22:35 +0100)]
wine: handle all *-config scripts
We already had a patch to allow us to explicitly pass the location of
the sane-config script, but wine's configure script uses several other
*-config scripts. This commit extends the wine patch and wine.mk to
explicitly pss the location of all *-config scripts that may be used
by wine.
Thomas Petazzoni [Tue, 22 Dec 2015 20:52:11 +0000 (21:52 +0100)]
ltrace: add patch to fix build on sparc
Due to a missing header inclusion, the build of ltrace of sparc fails
with the following error:
plt.c: In function 'sym2addr':
plt.c:33:12: error: dereferencing pointer to incomplete type
return sym->enter_addr;
^
plt.c:34:1: warning: control reaches end of non-void function [-Wreturn-type]
}
This commit fixes that by adding a patch to the ltrace code base.
Thomas Petazzoni [Tue, 22 Dec 2015 20:39:17 +0000 (21:39 +0100)]
lockdev: fix build with the musl C library
When the musl C library is used, the __GNU_LIBRARY__ symbol is not
defined, so the lockdev code doesn't know what to do. To work around
this, we explicitly pass __GNU_LIBRARY__ when building with musl, this
musl provides what's needed for lockdev to build, as if we were
building against glibc.
Doug Kehn [Wed, 2 Dec 2015 14:56:17 +0000 (08:56 -0600)]
package/unscd: new package
A daemon which handles passwd, group and host lookups for running
programs and caches the results for the next query. This package is
intended to replace glibc nscd. Presently, glibc nscd is not installed
by Buildroot.
unscd depends on glibc because it relies on glibc function
__nss_disable_nscd.
nscd.conf is taken from glibc with unused configurations removed.
Installs /etc/nfsmount.conf on the target, mainly to be able to configure the
the NFS version that mount.nfs uses by default. Without this ability you are
likely to encounter mount errors when automounting directories from NFS
version 3 servers.
Tested on i.MX6 using NFS mounts only. Not tested with DAS mounts as
of this commit.
[Thomas:
- remove the "# rpcbind" comments on the thread and mmu dependencies,
since we don't select rpcbind. The thread and mmu dependencies are
still needed though, since autofs uses pthread_*() functions and
fork().
- Remove "Requires a toolchain with native RPC." in the Config.in
help text and instead add a proper Config.in comment about the
thread and RPC dependency.
- Add patch to fix building with uClibc, taken from OpenEmbedded.]
Joris Lijssens [Tue, 22 Dec 2015 10:36:36 +0000 (11:36 +0100)]
emlog: new package
[Thomas:
- update to the latest github version, which has a number of
fixes/improvements compared to the latest tag.
- add a patch to fix the build of the kernel module with recent
kernel versions (>= 3.19).
- use $(INSTALL) -D with a full destination path.
- rewrap Config.in help text, lines were too long.]
Gustavo Zacarias [Tue, 22 Dec 2015 10:46:11 +0000 (07:46 -0300)]
rsync: security bump to version 3.1.2
Fixes:
Make sure that all transferred files use only path names from inside the
transfer. This makes it impossible for a malicious sender to try to
make the receiver use an unsafe destination path for a transferred file,
such as a just-sent symlink.
Ezequiel García [Mon, 30 Nov 2015 15:30:58 +0000 (12:30 -0300)]
openbox: new package
Add support for the Openbox Windows Manager.
[Thomas:
- don't mention MMU in the Config.in comment.
- wchar was forgotten in the Config.in comment.
- the dependency of the comment on thread should be "depends on
!BR2_TOOLCHAIN_HAS_THREADS" so that the comment is shown when
thread support is *not* available.
- add missing select of BR2_PACKAGE_XLIB_LIBXFT and
BR2_PACKAGE_XLIB_LIBXRENDER which are needed for pango to have xft
support, otherwise the build fails with "configure: error: Package
requirements (pango >= 1.8.0 pangoxft >= 1.8.0) were not met:"
- remove useless empty newline in the hash file
- add missing dependency on host-pkgconf, which is needed since
the configure.ac uses PKG_CHECK_MODULES.]
toolchain-external: add support for musl toolchain on ARM EABIhf
Since a few releases, the pre-built musl external toolchain has added
an ARM EABIhf variant, built for ARMv5T. This commit allows this
additional external toolchain to be used.
Thomas Petazzoni [Sun, 18 Oct 2015 21:05:02 +0000 (23:05 +0200)]
linux: don't build appended DTB image in place and support multiple images
Currently, the linux.mk logic for appended DTB image does the
appending of the DTB in place, directly at the end of the zImage using
a >> sign. This is incorrect because if you run "make linux-rebuild"
multiple times, you get the DTB appended over and over again to the
image.
Since keeping the 'zImage' or 'uImage' name for the appended DTB image
is not very clear, this commit moves to using the 'zImage.<dtb>' and
'uImage.<dtb>' format. This way, we can clearly distinguish the
original image from the appended one.
In addition, this naming scheme easily allows to generate *multiple*
appended DTB images: from one zImage, you can generate multiple
zImage.<dtb> for several DTBs, and then generate (if requested) the
corresponding uImage.<dtb>.
To achieve this, this commit:
- Changes the definition of LINUX_APPENDED_DTB to iterate over
$(KERNEL_DTS_NAME), and generate a zImage.<dtb> image for each of
them.
- Changes the addition of LINUX_APPENDED_DTB for appended uImage to
also iterate over $(KERNEL_DTS_NAME).
- Provide a different implementation of LINUX_INSTALL_IMAGE which
installs all the appended DTB images (but not the bare image)
- Remove the checks that verified that only one DT name is passed
when appended DTB is used, since we now support generating multiple
DT images.
Thomas Petazzoni [Sun, 18 Oct 2015 21:05:01 +0000 (23:05 +0200)]
linux: only install the DTBs when not in appended DTB mode
When you're using the "appended DTB" mode, the Device Tree blob gets
appended to your kernel image, so there is no point in installing both
the DTB and the kernel image to the images or target directories,
installing the kernel image itself is sufficient.
Therefore, this commit disables the definition of LINUX_INSTALL_DTB
when appended DTB is used.
Thomas Petazzoni [Sun, 18 Oct 2015 21:05:00 +0000 (23:05 +0200)]
linux: de-duplicate DTB and Linux image installation
Currently, the LINUX_INSTALL_DTB and LINUX_INSTALL_DTB_TARGET macros
are exactly the same, except for the target directory.
Similarly, LINUX_INSTALL_KERNEL_IMAGE_TO_TARGET and
LINUX_INSTALL_IMAGES_CMDS are copying the kernel image, just to a
different place (and with a different strategy).
As a preparation for future additions, this commit de-duplicate this
code:
- LINUX_INSTALL_DTB becomes a make macro that takes one argument: the
destination directory.
- LINUX_INSTALL_IMAGE is a new make macro that also takes on
argument: the destination directory.
Both macros are used by LINUX_INSTALL_KERNEL_IMAGE_TO_TARGET and
LINUX_INSTALL_IMAGES_CMDS to respectively install to the target
directory and the images directory.
Jörg Krause [Sun, 1 Nov 2015 08:35:18 +0000 (09:35 +0100)]
package/wpa_supplicant: add config option to build wpa_client library
Version 2.5 of wpa_supplicant introduces the config option
CONFIG_BUILD_WPA_CLIENT_SO to enable building wpa_client as shared library.
We need to add a patch from upstream to fix a runtime error with
libwpa_client.so:
Error relocating /usr/lib/libwpa_client.so: bin_clear_free: symbol not found
Romain Naour [Sat, 19 Dec 2015 16:39:14 +0000 (17:39 +0100)]
package/elementary: rename libelementary to elementary
The upstream name is elementary so use it.
Even if this package provide libelementary.so, it also provide
elementary_codegen, elementary_run and elm_prefs_cc tools.
Romain Naour [Sat, 19 Dec 2015 18:14:46 +0000 (19:14 +0100)]
toolchain-external: Ti Arago ARM: support only one version
See the conclusion about external toolchains during the Buildroot
meeting [1]:
"In the future, we stick to a single external toolchain version. The
Kconfig symbol should not encode the version (avoid legacy handling)"
Romain Naour [Sat, 19 Dec 2015 18:14:45 +0000 (19:14 +0100)]
toolchain-external: Synopsys ARC: support only one version
See the conclusion about external toolchains during the Buildroot
meeting [1]:
"In the future, we stick to a single external toolchain version. The
Kconfig symbol should not encode the version (avoid legacy handling)"
Romain Naour [Sat, 19 Dec 2015 18:14:44 +0000 (19:14 +0100)]
toolchain-external: ADI Blackfin: support only one version
See the conclusion about external toolchains during the Buildroot
meeting [1]:
"In the future, we stick to a single external toolchain version. The
Kconfig symbol should not encode the version (avoid legacy handling)"
Romain Naour [Sat, 19 Dec 2015 18:14:43 +0000 (19:14 +0100)]
toolchain-external: CodeSourcery x86: support only one version
See the conclusion about external toolchains during the Buildroot
meeting [1]:
"In the future, we stick to a single external toolchain version. The
Kconfig symbol should not encode the version (avoid legacy handling)"
Romain Naour [Sat, 19 Dec 2015 18:14:42 +0000 (19:14 +0100)]
toolchain-external: CodeSourcery SH: support only one version
See the conclusion about external toolchains during the Buildroot
meeting [1]:
"In the future, we stick to a single external toolchain version. The
Kconfig symbol should not encode the version (avoid legacy handling)"
Romain Naour [Sat, 19 Dec 2015 18:14:41 +0000 (19:14 +0100)]
toolchain-external: CodeSourcery PowerPC: support only one version
See the conclusion about external toolchains during the Buildroot
meeting [1]:
"In the future, we stick to a single external toolchain version. The
Kconfig symbol should not encode the version (avoid legacy handling)"
Romain Naour [Sat, 19 Dec 2015 18:14:40 +0000 (19:14 +0100)]
toolchain-external: CodeSourcery ARM: support only one version
See the conclusion about external toolchains during the Buildroot
meeting [1]:
"In the future, we stick to a single external toolchain version. The
Kconfig symbol should not encode the version (avoid legacy handling)"
André Hentschel [Sun, 20 Dec 2015 02:24:12 +0000 (03:24 +0100)]
wine: Bump to version 1.8
Bumping to brand new stable version 1.8
Upstreamed patches are dropped, sane patch is rebased.
Adding libpcap, pulseaudio and samba4 as soft dependencies.
sfnt2fon moved to its own subfolder and needs to be built in host-wine.
[Thomas:
- replace patch 0001 fixing autoreconf by a simpler patch that simply
drops the -Wall and -Werror flags passed to automake.
- reformat the two patches using Git
- drop the "uclibc || glibc" dependency inherited from libv4l, since
libv4l has now been fixed to build on musl.
- drop the hook work arounding a build issue around man pages, and
replace by a patch that completely disables the build/installation
of documentation.
- change license to LGPLv2.1+, as noted in the COPYING file.
- drop --enable-shared=yes, since it is already passed by the
autotools-package infrastructure]
projects / buildroot-mgba.git / log