Without this, a linux-rebuild or bootloader-rebuild will not install
the rebuilt image in the images directory.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The changes in e055aea216 and
0849e8193e have broken the bootloader
support.
This is because the 'pkgparentdir' macro returns the string 'boot/'
for bootloaders, while the GENTARGETS_INNER macro was testing this
against 'boot' to know if the configuration option to check for this
package is BR2_PACKAGE_* or BR2_TARGET_*.
So, instead of testing against 'boot', we test against 'boot/'.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
For wget-downloaded packages, only the tarball name was displayed,
while for svn/git/bzr packages, the complete repository URL was
shown. As discussed on the list, what we want for 'make external-deps'
is only the tarball name even for git/svn/bzr downloads.
For local packages, we of course keep the full path since it's the
only information that we can provide.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
This can be used this way :
<pkg>_VERSION = 42
<pkg>_SITE = file:///some/local/directory
<pkg>_SOURCE = mypkg-$(<pkg>_VERSION).tar.bz2
Can be useful to integrate a home-made project or for testing purposes.
The default command to retrieve files is 'cp' but 'rsync' could also be used.
Through sshfs, it should also be possible to get non-public remote files on a
ssh server.
[ Thomas Petazzoni: use $(PKG)_SITE and $(PKG)_SOURCE variables
instead of $(1) and $(2) ]
[ Peter: don't append $(QUIET), cp doesn't handle -q]
Signed-off-by: David Wagner <david.wagner@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Thanks to the pkgparentdir and pkgname functions, we can rewrite the
GENTARGETS macro in a way that avoids the need for each package to
repeat its name and the directory in which it is present.
[Peter: pkgdir->pkgparentdir]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
This commit adds the functions "pkgname" and "pkgdir", which can then
be used to automatically find the name of the package and the
directory of the package which does the call to
GENTARGETS/AUTOTARGETS/CMAKETARGETS. These functions use the
$(MAKEFILE_LIST) variable, which make automatically makes available,
and which contains the list of included Makefiles, with the current
Makefile being last. Thanks to this variable and a little bit of
string manipulation, we can easily find out automatically the package
name and the directory it is part of.
These functions are used in later commits to simplify the GENTARGETS,
AUTOTARGETS and CMAKETARGETS calls.
[Peter: Rename to pkgmakefile->pkgdir, pkgdir->pkgparentdir as that's what it is]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The new override source directory mechanism allows to tell Buildroot
to use a particular directory as the source directory for a
package. However, this mechanism works with a local override makefile
and not directly within the package recipe itself.
For some use cases, it might be desirable to write a package recipe
which always refers to a local source directory (and not a http, git,
svn or bazaar download). This commit makes this possible by adding the
'local' site method. It allows to write package recipes as follows:
MYPKG_SITE = /tmp/mypkg-source-code
MYPKG_SITE_METHOD = local
[...]
$(eval $(call GENTARGETS,package,mypkg))
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
We are often asked "how can I restart the build of a package ?" or
"how can I restart the build of package from the configure part
?". Obviously, tweaking with stamp files is possible, but not very
user friendly.
Therefore this patch adds two new per-package targets: <pkg>-rebuild
and <pkg>-reconfigure. They will remove the right stamp files and
restart the complete build process (by using the 'all' target, so that
not only the package is reconfigured, recompiled and reinstalled, but
the root filesystem images are also regenerated).
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
When a variable <pkg>_OVERRIDE_SRCDIR is defined, then Buildroot will
no longer try to download, extract and patch the package. It will
simply use the value of this variable as the source directory for the
package. The contents of the package sources will be rsynced to the
package build directory in $(O)/build/pkg-version/.
This can be used to tell Buildroot that the sources for a given
package are inside some directory that you control, and which can be
versioned in Git/SVN, or handled in whichever way you want.
Those <pkg>_OVERRIDE_SRCDIR variables will be defined by a local
makefile included by Buildroot, which will be handled in a later
commit.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
This will be useful to integrate Xenomai and RTAI into Buildroot, so
that they can patch the kernel before the normal patching process
starts.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Currently, the dependency chain is :
configure -> patch
patch -> extract
extract -> depends
depends -> source + dependencies ready
This is a problem for RTAI, as :
* In linux/linux-ext-rtai.mk, we need the RTAI source code to be
extracted/patched by the RTAI package, so we need to have a
dependency on RTAI.
* The RTAI package cannot have its configure step executed before the
Linux kernel is built, because it needs the kernel source code to
be extracted and configured.
Therefore, we have a sort of circular dependency. We solve this
circular dependency by having the RTAI kernel part depending on
rtai-patch, and having the RTAI userspace part depending on the linux
package. However, for this to work, we have to modify the dependency
chain to:
configure -> patch + depends
patch -> extract
extract -> source
depends -> dependencies ready
Which means that dependencies are guaranteed to be built not when the
package is extracted, but when the package is configured. Which is
technically more correct that what we were doing in the past.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The name "patch-kernel.sh" is a bit stupid, since this script is used
to patch everything in Buildroot, not only kernel trees.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
993e51bc22 introduced a <pkg>_DL_VERSION
in addition to <pkg>_VERSION. This variable, purely internal to the
package infrastructure, allows to store the *raw* version name, as
defined in the package .mk file. However, as this version string can
contain slashes, it is not appropriate to be part of the build
directory name of the package. This is why <pkg>_VERSION is defined to
be the same thing as <pkg>_DL_VERSION, except that slashes are
replaced by underscores.
However, the initial implementation didn't take into account the case
of host packages. So 763822e874 was
committed to fix this. But unfortunately, this commit got it wrong
again: it forgot the case where <pkg>_VERSION is set, but not
HOST_<pkg>_VERSION. In this case, HOST_<pkg>_DL_VERSION remained
empty. This can be seen for example by building host-sstrip.
This problem has been noticed by Julien Boibessot
<julien.boibessot@free.fr>, who also helped in fixing the
issue. Thanks!
The fix is simply to ensure that <pkg>_DL_VERSION is properly set in
all cases: when HOST_<pkg>_VERSION is defined, when only <pkg>_VERSION
is defined, and when no version is defined.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Allan W. Nielsen <a@awn.dk>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
As the kernel and bootloaders do not use the normal BR2_PACKAGE_*
Kconfig options, their target name was not automatically added to the
global TARGETS variable. Each bootloader .mk and the linux.mk had to
add their own target manually to TARGETS, and the package
infrastructure was making tests on non-existing Kconfig variables.
This commit improves the package infrastructure so that it looks at
BR2_PACKAGE_<pkg> for packages, BR2_TARGET_<pkg> for bootloaders and
at the special BR2_LINUX_KERNEL for the linux package.
This allows to simplify a little bit the bootloaders and linux .mk
files.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
This will be needed for the Linux kernel package, which needs to
download patches in a custom way.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The patching procedure was using $($(NOHOSTPKG)_NAME) to find the name
of the package without the host- prefix for host packages, which is
needed to find the package directory and the prefix of the patches.
However, this $($(NOHOSTPKG)_NAME) variable is empty when there is
only an host variant for the package and no target variant for it.
Therefore, instead of using this $($(NOHOSTPKG)_NAME) variable, we now
use a new $(RAWNAME) variable, which is always the name of the package
without the host- prefix, even for host packages.
Thanks to Yann E. Morin for providing the tip on using patsubst
instead of subst.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
This will allow the bootloader makefiles and the kernel makefile to
rely on the GENTARGETS infrastructure.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
When using GENTARGETS (or macroes depending on it), there is no way of
specifying a custom extraction procedure. When using the patch one can
simply define $(PACKAGE_NAME)_EXTRACT_CMDS which will override the
default. If non is defined it will fall back to the default extract
procedure.
An example could look like this:
=== PACKAGE BEGIN ===
FGLRX_VERSION=11-3
FGLRX_SOURCE:=ati-driver-installer-$(FGLRX_VERSION)-x86.x86_64.run
FGLRX_SITE:=http://www2.ati.com/drivers/linux/
define FGLRX_EXTRACT_CMDS
sh $(DL_DIR)/$(FGLRX_SOURCE) --extract $(@D)
endef
$(eval $(call GENTARGETS,package,fglrx))
=== PACKAGE END ===
Signed-off-by: Allan W. Nielsen <a@awn.dk>
Cc: Allan W. Nielsen <a@awn.dk>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
There's no reason to make the source files writable for everyone.
Reported-by: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Commit ae2be8d (infrastructure: remove support for old style hooks)
forgot to update an uninstall rule.
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Daniel Hobi <daniel.hobi@schmid-telecom.ch>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
This will reduce the space used and speed up the clone as it is only
used to generate an archive, which doesn't need the git working tree.
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
For packages where the sources are included in buildroot (E.G. makedevs).
We unfortunately already use no <pkg>_SOURCE to mean <pkg>_<version>.tar.gz,
in several packages (and for git/svn/bzr support), so you need to
define <pkg>_SOURCE to the empty string to use it.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Without this patch we'd need separate versioned patch directories for
target and host packages, thus having unwanted duplication.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Instead of having to require svn, git or bzr unconditionally, only
require them when one package needs them to be downloaded.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Commit 993e51bc22 introduced a problem
to compute <PKG_NAME>_VERSION variable when PKG_NAME start with
'HOST_'.
Signed-off-by: Lionel Landwerlin <llandwerlin@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
This reverts commit 48cf66f1a2.
Signed-off-by: Maxime Petazzoni <maxime.petazzoni@bulix.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The new DL_MODE variable dispatches between the various download
implementations of each method (Git, Subversion, Wget) to deal with the
normal download (default mode, 'DOWNLOAD'), the source-check
('SOURCE_CHECK') and to show the external dependencies for external-deps
('SHOW_EXTERNAL_DEPS').
For the latter, the legacy script wget-show-external-deps.sh is no
longer required as $(WGET) isn't called directly anymore but always
through the DOWNLOAD helper.
Signed-off-by: Maxime Petazzoni <maxime.petazzoni@bulix.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Packages can now be sourced from Git and Subversion repositories. The
download method will be autodetected from the URI (git://, svn://, etc).
If the repository is accessed through http(s), you can force the
download method by setting a _SITE_METHOD variable to either 'git' or
'svn', respectively and without the quotes.
The package's _VERSION variable defines which commit, revision, tag or
branch should be checked out. For Git, it can be HEAD, a commit ID, a
tag name or branch name (anything that can be checked out with `git
checkout`). For Subversion, it must be a revision number, or HEAD.
Signed-off-by: Maxime Petazzoni <maxime.petazzoni@bulix.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
In preparation for the re-work of the DOWNLOAD helper to support
multiple download methods and protocols, the spider feature used with
wget is removed for now until it is re-implemented on top of the new
download methods.
Signed-off-by: Maxime Petazzoni <maxime.petazzoni@bulix.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Now that all old-style hooks occurences have been converted to
new-style hooks, let's get rid of the code needed to support old-style
hooks from the package infrastructure.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Make KCONFIG_{ENABLE,DISABLE,SET}_OPT match whole words only.
Fixes that BR_INET_IPV6 enables CONFIG_IPV6 in the kernel's
.config, but also nukes everything with "CONFIG_IPV6" in its name, e.g.
CONFIG_IPV6_PRIVACY. This in turn causes any IPv6 options in a custom
kernel .config to be lost and oldconfig going interactive.
Signed-off-by: Malte Starostik <m-starostik@versanet.de>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Paul Jones <paul@pauljones.id.au>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The KCONFIG_ENABLE_OPT, KCONFIG_SET_OPT and KCONFIG_DISABLE_OPT are
new make functions to respectively enable, set and disable options in
Kconfig-like files (as used by the kernel, uClibc or Busybox).
They can be used as follows :
$(call KCONFIG_ENABLE_OPT,CONFIG_FOOBAR,/path/to/.config)
$(call KCONFIG_SET_OPT,CONFIG_BARFOO,foobar,/path/to/.config)
$(call KCONFIG_DISABLE_OPT,CONFIG_FARBOO,/path/to/.config)
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The new <pkg>-show-depends simply outputs the list of dependencies for
the given package.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
As noticed by Thomas, we call the UPPERCASE macro a lot, and it slows down
startup quite a bit.
Optimize it by implementing it in make, rather than forking a shell + tr.
The implementation is heavily based on the 'up' macro from gmsl
(http://gmsl.sf.net)
With this in place, startup time is ~5 times lower.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Since the documentation cleanup is done by the main Makefile in a
global way, there's no need to do that on a per-package basis in the
generic package infrastructure.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This new infrastructure allows to write simpler .mk files for packages
not using the autotools as their build system, by factorizing many
common steps (download, extract, patching), and will more easily allow
Buildroot-wide changes in how the packages are handled.
The main macro is called GENTARGETS and works similarly to the
AUTOTARGETS macro that already exists for autotools-based
packages. However, the set of variables to be defined before calling
the macro is different. Refer to the documentation for details.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
