Age | Commit message (Collapse) | Author | Files |
|
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
When the function to set PSTAGING_ACTIVE is an anon python function OVERRIDES
is not set so we cannot set things to disabled in the context of BBCLASSEXTEND
recipes. To fix this we need to have native/nativesdk set OVERRIDES outside of
the anon function context. We then make a COW of the data at this point in
packaged-staging's anon function and evaluate so we know what will be in
OVERRIDES and check. While we're in here drop a duplicate test in
populate_sysroot_postamble and check nativesdk along with the rest of the
classes that need mangling.
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
* Changed the command that packages the contents of the
source ipk to ar instead of tar to match how other ipks
are built.
Signed-off-by: Chase Maupin <chase.maupin@ti.com>
Signed-off-by: Koen Kooi <koen@openembedded.org>
|
|
* Added License field to the control file.
Signed-off-by: Chase Maupin <chase.maupin@ti.com>
Signed-off-by: Koen Kooi <koen@openembedded.org>
|
|
This has some unintended consequences, revert for now.
This reverts commit 9a0e95f37639d9856ad41b8fc301f9b00369a797.
|
|
When the function to set PSTAGING_ACTIVE is an anon python function
overrides are not set so we cannot set things to disabled in the context
of BBCLASSEXTEND recipes. Call this function before we check
PSTAGING_ACTIVE. In the case of packagedstaging_fastpath it's cleaner
to convert to a full python function. While we're in here drop a
duplicate test in populate_sysroot_postamble and make do_package_stage
check PSTAGING_ACTIVE for == 0 rather than != 1, like the rest of the
functions do.
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
With the changes to move ${CROSS_DIR} into sysroots/ we now
$ORIGIN-mangle cross packages (and already were native) so change
the comment from cross.bbclass while we're at this and then put
these recipes into the usual packaged-staging cache.
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
- Don't supply site files for native.
- Split up the site information, so we don't need to add every single
combination of os and architecture to the siteinfo.bbclass.
- Drop the ${FILE_DIRNAME}/site-<bits> stuff, there are other ways to do it.
- SITEINFO_ENDIANESS -> SITEINFO_ENDIANNESS.
Signed-off-by: Chris Larson <clarson@kergoth.com>
|
|
Will revisit.
This reverts commit 87dc166bade8722af4076c8644cac6a0a437826f.
|
|
- Don't supply site files for native
- Split up the site information, so we don't need to add every single
combination of os and architecture to the siteinfo.bbclass
- Instead of specifically looking for '${FILE_DIRNAME}/site-${PV}/<sitefile>',
leverage FILESPATH. Now it searches the paths in both BBPATH and FILESPATH
for "site/<sitefile>".
- SITEINFO_ENDIANESS -> SITEINFO_ENDIANNESS
Signed-off-by: Chris Larson <clarson@kergoth.com>
|
|
bumped PR for the main perl recipes so this change takes effect
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
Acked-by: Koen Kooi <koen@openembedded.org>
Acked-by: Roman I Khimov <khimov@altell.ru>
|
|
ConfigParsed is often utilized to do programmatic changes to the configuration
metadata. If these other event handlers set any variables which
sanity.bbclass needs to check, there's a problem, since we can't ensure those
handlers run before this one. So, we move the sanity checking to when the
build is about to start, this ensures that the configuration metadata is
entirely ready to go.
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
* tmpdir layout is changed bump ABI version to force a rebuild.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Acked-by: Tom Rini <tom_rini@mentor.com>
Acked-by: Richard Purdie <rpurdie@rpsys.net>
Ackef-by: Chris Larson <clarson@kergoth.com>
Acked-by: Martin Jansa <Martin.Jansa@gmail.com>
Acked-by: Koen Kooi <koen@openembedded.org>
|
|
sysroot
* Update references to CROSS_DIR in classes and bitbake.conf.
* This patch is based on Joshua Lock's work in poky.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Acked-by: Tom Rini <tom_rini@mentor.com>
Acked-by: Richard Purdie <rpurdie@rpsys.net>
Ackef-by: Chris Larson <clarson@kergoth.com>
Acked-by: Martin Jansa <Martin.Jansa@gmail.com>
Acked-by: Koen Kooi <koen@openembedded.org>
|
|
GNU make will look for GNUmakefile, makefile and Makefile. GNUmakefile
isn't recommended but used by some programs, e.g. canutils.
Signed-off-by: Vitus Jensen <vjensen@gmx.de>
Acked-by: Chris Larson <chris_larson@mentor.com>
Acked-by: Denys Dmytriyenko <denis@denix.org>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
clamav 0.96.1 makes use of this flag
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
This patch introduces COMPATIBLE_TARGET_SYS
It is similar to COMPATIBLE_MACHINE but where COMPATIBLE_MACHINE
is used to specify that a certain recipe is for a certain machine
COMPATIBLE_TARGET_SYS can be used to specify that a certain recipe
is for a certain architecture.
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
Acked-by: Tom Rini <tom_rini@mentor.com>
|
|
Some programs such as libxml-parser-perl-native can have junky
RPATH directories. Found by Khem Raj.
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
in bitbake
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Acked-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
|
|
Rather than calculate the relative path ourself use a helper function
we've got. This also results in shorter rpaths being used.
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
* Created the sourceipk.bbclass class that can be used to
package the patched sources and recipe for an application
into an ipk for installation.
* Allows specification of the installation directory for the
sources using the SRCIPK_INSTALL_DIR variable.
* Can be enabled per package and controlled through the
CREATE_SRCIPK variable.
* Creates a README file in the sources directory that
explains the contents of the package.
* User can specify the package architecture for the source ipk
using the SRCIPK_PACKAGE_ARCH variable. The default value is
"all".
* Using the -L flag to the cp command to copy files pointed to by
symlinks and not the symlinks themselves.
* Adds source package to list of dynamic packages.
Signed-off-by: Chase Maupin <chase.maupin@ti.com>
Signed-off-by: Koen Kooi <k-kooi@ti.com>
|
|
bb.decodeurl()
* This patch now uses the path which essentially is the path and
filename relative to recipe dir. It will create the directories
as expected. Suppose if the is says file://pam.d/login in the
SRC_URI then it will be copied to $WORKDIR/pam.d/login as it
was happening before the fix to decodeurl in bitbake.
* Do not copy the .patch and .diff files into $WORKDIR
they are already symlinked in patches/ dir which will
be used by quilt.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Acked-by: Roman I Khimov <khimov@altell.ru>
Acked-by: Chris Larson <clarson@kergoth.com>
|
|
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
|
|
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
|
|
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
|
|
installed of unpacked
* silly solution to filter multiple versions shown in installed-packages.txt where one is with Status not-installed and the right one with installed
* someone with better idea?
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
Same reasoning as the {autotools,base}_dep_prepend rename.
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
When we start including methodpool funcs in the metadata, these blow up, since
the _prepend gets evaluated at finalize time.
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
Currently, we make all pstage packages be locked to BUILD_SYS which causes
annoying warnings in some cases and doesn't allow for sharing between say 32
and 64bit (which is valid for target bits).
We now change to ${PACKAGE_ARCH}-${HOST_OS} for everything except for cross
where we need to encode HOST_SYS into the name as well.
Acked-by: Khem Raj <raj.khem@gmail.com>
Acked-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
sorted entries alphabetically
indented everything properly
removed a tab
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
|
|
Sorted the entries alphabetically (except for x86_64 which I kept with the intel group).
Made sure the tables are all indented the same way.
replaced some tabs with spaces
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
|
|
* otherwise it will pick 'The' ie from line 'Description: The Enlightenment Window Manager Version 17' and mess installed-packages.txt
fails: Version_shr-version_1.0-20100512-r0.4_om-gta02.ipk
fails: The_e-wm_0.16.999.060+svnr48617-r13.4_armv4t.ipk
* this is also fix for missing entries in installed-package-sizes.txt
(when find cannot find ipk files for messed package names)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
On some host distributions the provided linux kernel headers are too old to
compile utilities we need. Given that we need these utilities to run things on
the target the best solution is to provide linux-libc-headers-native. Rather
than get things into an inconsistent state, we make linux-libc-headers-native
be a default dependency.
Acked-by: Leon Woestenberg <leon@sidebranch.com>
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
|
|
Libtool in some packages (e.g. curl) tries to relink libraries during
installation stage. If the Icecc environment isn't set in the do_install
such relinking fails. Fix this by adding do_install_prepend to
icecc.bbclass with set_icecc_env() call.
Signed-off-by: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com>
|
|
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
PSTAGE_TMDPDIR_STAGE
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
* otherwise fails on hosts without desktop-file-utils installed, because cannot find binaries from desktop-file-utils-native
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
instead setting PATH)
|
|
* host objdump can support multiple targets and instead of output it shows list of targets to choose from
see http://thread.gmane.org/gmane.comp.handhelds.openembedded/33294/focus=33357
* PATH set in cmd as suggested by kergoth
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Acked-by: Khem Raj <raj.khem@gmail.com>
|
|
|
|
Currently with srctree, we get errors like the following:
| + srcuri=
| + '[' '' == '' ']'
| + srcuri=OpenEmbedded
| + echo 'Source: '
| ++ find /scratch/oe/oe-build/oe/build/angstrom-2008.1/tmp/work/armv5te-angstrom-linux-gnueabi/autotools-demo-1.0-28-g3a96627-r0/staging-pkg '(' -name '*.la' -o -name '*-config' ')' -type f
| + stage-manager-ipkg-build -o 0 -g 0 /scratch/oe/oe-build/oe/build/angstrom-2008.1/tmp/work/armv5te-angstrom-linux-gnueabi/autotools-demo-1.0-28-g3a96627-r0/staging-pkg /scratch/oe/oe-build/oe/build/angstrom-2008.1/tmp/pstage/angstromglibc
| *** Error: CONTROL/control is missing field Source
| The Source field contain the URL's or filenames of the source code and any patches
| used to build this package. Either gnu-style tarballs or Debian source packages
| are acceptable. Relative filenames may be used if they are distributed in the same
| directory as the .ipk file.
|
| ipkg-build: Please fix the above errors and try again.
| ERROR: Function staging_packager failed
NOTE: package autotools-demo-1.0-28-g3a96627-r0: task do_package_stage: Failed
ERROR: TaskFailed event exception, aborting
This patch fixes that issue.
Acked-by: Chris Larson <chris_larson@mentor.com>
Acked-by: Koen Kooi <k-kooi@ti.com>
Acked-by: Khem Raj <raj.khem@gmail.com
|
|
package_update_index_ipk
We add a lockfile of do_populate_sdk and do_rootfs (in rootfs_ipk) that is also
the lockfile that package_update_index_ipk uses. With enough threads it is
possible that one image (or meta-toolchain) is attempting to use the package
index while another has only reached the point of generating the index leading
to an empty index being seen later on.
Acked-by: Chris Larson <chris_larson@mentor.com>
Acked-by: Roman I Khimov <khimov@altell.ru>
Signed-off-by: Tom Rini <tom_rini@mentor.com>
|
|
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Acked-by: Martin Jansa <Martin.Jansa@gmail.com>
|