Joshua Watt [Mon, 12 Feb 2018 16:52:02 +0000 (10:52 -0600)]
icecc-create-env: Symlink alternate names
Instead of renaming files to a new path in the toolchain archive, keep
the files with their original paths and create a relative symbolic link
from the new path to the original file.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Mon, 12 Feb 2018 16:52:01 +0000 (10:52 -0600)]
icecc-create-env: Use program interpreter for deps
ldd cannot always be used to determine a program's dependencies
correctly, particularly when the program specifies an alternate program
interpreter (dynamic loader). This commonly happens when using a
uninative tarball. Instead, determine the program's requested
interpreter, and ask it to list the dependencies.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Mon, 12 Feb 2018 16:51:59 +0000 (10:51 -0600)]
icecc-create-env: Reformat
The environment script used an annoying mix of tabs and spaces and no
mapping of tabs to spaces would produce pleasant indentation. Reformat
to eliminate tab characters and settle on 4 spaces for indentation
(which matches the upstream icecream script from which this is derived)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Mon, 12 Feb 2018 16:51:58 +0000 (10:51 -0600)]
icecc.bbclass: Skip canadian-cross compiles
icecc.bbclass will no longer attempt to distribute cross-canadian
compiles. While it is technically possible to generate a toolchain that runs
on the build system and generates executables for the host system, this
is not the normal way that icecc operates. There are so few of these
recipes that it is probably not worth maintaining a distinct code path
for them.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Mon, 12 Feb 2018 16:51:57 +0000 (10:51 -0600)]
icecc.bbclass: Fix STAGING_BINDIR_TOOLCHAIN usage
STAGING_BINDIR_TOOLCHAIN is actually a path list, not a single path. Fix
icecc.bbclass to try all the paths in the variable instead of treating
it as a single path.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Mon, 12 Feb 2018 16:51:56 +0000 (10:51 -0600)]
icecc.bbclass: Move to shared work directory
Generate the icecc toolchains in a shared work directory. This class was
already setup to correctly synchronize creating the toolchains in a
shared location before the RSS changes, so return to that behavior
instead of generated the toolchains in each recipe's sysroot.
Additionally, it makes no sense for each recipe to generate a toolchain,
only to find it was already generated and uploaded to the compile server
by another recipe.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 14 Feb 2018 14:43:42 +0000 (14:43 +0000)]
ltp: Improve package stripping
Just exclude the specific tests which have an issue with being stripped
rather than the whole package. This reduces the disk footprint by around
400MB.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 14 Feb 2018 14:40:55 +0000 (14:40 +0000)]
package: Add INHIBIT_PACKAGE_STRIP_FILES to allow files to remain unstripped
There are cases where its useful to allow only a select few files
to be excluded from the package stripping mechanism. Currently
this isn't possible so add a variable to allow this.
This is to be used sparingly as in general the core code should be
doing the right thing. This is better than the alternative of leaving the whole
package unstripped.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This upgraded version of eudev includes:
- Headers to build under glibc-2.25 and above
- Bug fix on ata_id
- Misc: rules and hardware database update
Signed-off-by: Chang Rebecca Swee Fun <rebecca.swee.fun.chang@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yeoh Ee Peng [Thu, 8 Feb 2018 00:01:04 +0000 (16:01 -0800)]
oe-selftest: meta_ide: add tests for meta-ide-support
QA team were testing meta-ide-support manually. Add automated
tests to test that bibtake meta-ide-support will create the
toolchain and environment setup script. Also test that after
using environment setup script, one can compile c program
and build cpio project.
Signed-off-by: Yeoh Ee Peng <ee.peng.yeoh@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Remove recipe for sato-icon-theme that was obsolete since YP 2.2
release and change ownership for eudev recipe.
Signed-off-by: Chang Rebecca Swee Fun <rebecca.swee.fun.chang@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
64 bit entry point should be passed in 2 literals ( "0x1 0x00008000"
).ENTRYPOINT is assigned with first half only and erroring out as
'command not found' for the second half. Adding quotes while assignment fixes the
issue.
Signed-off-by: Vineeth Chowdary Karumanchi <vineethchowz.chowdary@xilinx.com> Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bruce Ashfield [Tue, 6 Feb 2018 16:21:59 +0000 (11:21 -0500)]
linux-yocto/4.9: drop recipe
As was previously announced, anything older than 4.14 is being
dropped in the master/release branches to better support newer
processors and to ensure that safe/secure kernels are the
defaults for all builds. The time required to update the older
kernels with constant updates (more than just CVEs) is not
justified in new releases.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bruce Ashfield [Tue, 6 Feb 2018 16:21:58 +0000 (11:21 -0500)]
linux-yocto/4.4: drop recipe
As was previously announced, anything older than 4.14 is being
dropped in the master/release branches to better support newer
processors and to ensure that safe/secure kernels are the
defaults for all builds. The time required to update the older
kernels with constant updates (more than just CVEs) is not
justified in new releases.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bruce Ashfield [Tue, 6 Feb 2018 16:21:57 +0000 (11:21 -0500)]
linux-yocto/4.10: drop recipe
As was previously announced, anything older than 4.14 is being
dropped in the master/release branches to better support newer
processors and to ensure that safe/secure kernels are the
defaults for all builds. The time required to update the older
kernels with constant updates (more than just CVEs) is not
justified in new releases.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bruce Ashfield [Tue, 6 Feb 2018 16:21:55 +0000 (11:21 -0500)]
linux-yocto/4.12: pinctrl backports
Backporting the following pinctrl commits to enable controllers on
Intel Cannon Lake:
4b7a5c1b4ec5 pinctrl: intel: Add Intel Cannon Lake PCH-H pin controller support 044631ce1937 pinctrl: intel: Add Intel Cannon Lake PCH pin controller support 2054b0ea59a7 pinctrl: intel: Make it possible to specify mode per pin in a group 55b9053b5a5a pinctrl: intel: Add support for variable size pad groups
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Richard Purdie [Fri, 9 Feb 2018 09:48:48 +0000 (09:48 +0000)]
utility-tasks: Drop fetchall and checkuriall tasks
The same thing can now be done with "bitbake <target> --runall=fetch"
or "bitbake <target> --runall=checkuri".
Dropping the tasks takes "bitbake core-image-sato -g" from 22s to 8s
since it no longer has to resolve the recursive dependencies (it
doesn't know if any given target will touch them or not until it
computes them). That is a significant enough win that its worth any
impact this may have on the small number of users using the tasks.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jackie Huang [Fri, 2 Feb 2018 07:10:33 +0000 (15:10 +0800)]
e2fsprogs: only quiet debugfs for native package
Many test cases expect the output of running commands for debugfs,
the patch quiet-debugfs.patch suppress the output so ptest failed:
| 292 tests succeeded 47 tests failed
The patch was meant to suppress the output in do_rootfs according
to the original commit:
======================
commit 13a8d3869b0fe0ebdddcee37fb48c5b9cdf718ba
Author: Ross Burton <ross.burton@intel.com>
Date: Mon Dec 23 13:38:34 2013 +0000
e2fsprogs: silence debugfs
debugfs echos the commands when it is executing a script, but as the scripts
used at rootfs time are long this massively inflates the do_rootfs log.
Comment out the echo so that the rootfs isn't 20K lines longer than it needs to
be.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
======================
So only apply the patch for native package, then the ptest will
be pass for target package.
Signed-off-by: Jackie Huang <jackie.huang@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ricardo Salveti [Fri, 2 Feb 2018 04:27:30 +0000 (02:27 -0200)]
grub-efi: add configfile as default built-in
The module configfile allows loading a custom grub configuration file
from block devices, which can be used to customize the default grub
menu when creating images with wic (via configfile argument).
Signed-off-by: Ricardo Salveti <ricardo@opensourcefoundries.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Rework 0002-Don-t-set-up-m32-m64-we-do-that-ourselves.patch to
remove the offending bits from the function instead of removing
calls to the function all over the place.
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Huang Qiyu [Thu, 1 Feb 2018 18:01:58 +0000 (20:01 +0200)]
libnl: 3.2.29 -> 3.4.0
1) Upgrade libnl from 3.2.29 to 3.4.0.
2) Add one patch "0001-PATCH-fix-libnl-3.4.0-musl-compile-problem.patch", for musl compile.
3) Delete one patch "fix-pktloc_syntax_h-race.patch", since upstream has refactored the makefiles, and the problematic code is now absent.
Signed-off-by: Huang Qiyu <huangqy.fnst@cn.fujitsu.com> Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Hongxu Jia [Fri, 2 Feb 2018 09:06:11 +0000 (04:06 -0500)]
apr: 1.6.2 -> 1.6.3
1. Improve inappropriate patches:
- Drop inappropriate configure_fixes.patch
Use setting variable ac_cv_file__dev_zero and
ac_cv_sizeof_struct_iovec to replace
- Drop cleanup.patch
Aassign variable libtool at ./buildconf executing and
use 0001-build-buildcheck.sh-improve-libtool-detection.patch
to replace. Submitted it to upstream.
- Rename configfix.patch to
0002-apr-Remove-workdir-path-references-from-installed-ap.patch
Add its original comments and author, explain why it is inappropriate
- Drop upgrade-and-fix-1.5.1.patch
Use 0003-Makefile.in-configure.in-support-cross-compiling.patch
to replace. And submitted it to upstream.
2. Fix build path issue to improve reproducibility
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Yi Zhao [Fri, 2 Feb 2018 02:16:09 +0000 (10:16 +0800)]
libbsd: update to 0.8.7
Drop 0001-Fix-for-older-GCCs-not-supporting-__has_include.patch as it
had been merged upstream.
Rebase 0001-Replace-__BEGIN_DECLS-and-__END_DECLS.patch.
Signed-off-by: Yi Zhao <yi.zhao@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Hongxu Jia [Thu, 1 Feb 2018 06:17:10 +0000 (01:17 -0500)]
gpgme: 1.9.0 -> 1.10.0
1. Disable test at build time to workaround corss-compile
Add 0007 to add option to disable python test and submit
it to upstream
2. Rework patch 0002, since the following commit of upstream.
https://git.gnupg.org/cgi-bin/gitweb.cgi?p=gpgme.git;a=commit;h=7309ce6f5f7c86570953a141965d4f54cd9ad9a0
3. Add patch 0006 to fix build path issue to improve reproducibility
and submit it to upstream
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Otavio Salvador [Thu, 1 Feb 2018 10:28:19 +0000 (08:28 -0200)]
glide: Add 0.13.1 version
Glide is a widely used Vendor Package Management for Golang,
especially in private projects as it allows for internal repositories
clone and other fancy things.
This recipe has been originally done as part of updatehub development.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Ross Burton <ross.burton@intel.com>
Liwei Song [Thu, 1 Feb 2018 06:40:49 +0000 (01:40 -0500)]
linux-firmware: package all ibt-17-x-x.sfi/ddc firmware
All ibt-17-x-x.sfi/ddc firmware are use to support Intel Bluetooth 9560
they are needed in different version of Bluetooth driver since
4.14 kernel version.
commit b77bb7afe513 ("linux-firmware: package ibt-17-16-1 firmware")
only package one of the ibt-17 series firmware.
As the Bluetooth driver's update, to avoid packaging the ibt-17 firmware
one by one, install them in one package ibt-17.
Signed-off-by: Liwei Song <liwei.song@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Huang Qiyu [Wed, 31 Jan 2018 08:56:34 +0000 (16:56 +0800)]
tar: 1.29 -> 1.30
1.Upgrade tar from 1.29 to 1.30.
2.Modify musl_dirent.patch, since the data has been changed.
3.Delete CVE-2016-6321.patch, since it is integrated upstream.
Signed-off-by: Huang Qiyu <huangqy.fnst@cn.fujitsu.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ross Burton [Wed, 31 Jan 2018 11:05:11 +0000 (11:05 +0000)]
i2c-tools: rewrite packaging
For unknowable reasons this recipe inherits autotools-brokensep despite not in
fact using autotools.
Remove the inherit, add a do_install() that calls the makefile directly.
Extend EXTRA_OEMAKE to pass bindir/sbindir/incdir/libdir/mandir so the files are
all packaged in the correct location (and remove the now redundant sed of
Makefile).
Pass EXTRA=eeprog to EXTRA_OEMAKE to build eeprog, and remove the now redundant
edit of Makefile.
Backport a patch from upstream to remove linux/i2c-dev.h instead of installing
it with a different name (which nobody will be using).
Signed-off-by: Ross Burton <ross.burton@intel.com>
Anders Roxell [Tue, 30 Jan 2018 13:12:42 +0000 (14:12 +0100)]
ltp: remove patches that already have fixes upstream
The following patches has already fixes for the issues they are solving
in upstream.
- 0001-add-_GNU_SOURCE-to-pec_listener.c.patch
- 0011-Rename-sigset-variable-to-sigset1.patch
- 0025-mc_gethost-include-sys-types.h.patch
- 0027-sysconf01-Use-_SC_2_C_VERSION-conditionally.patch
- 0033-shmat1-Cover-GNU-specific-code-under-__USE_GNU.patch
Suggested-by: Yi Zhao <yi.zhao@windriver.com> Signed-off-by: Anders Roxell <anders.roxell@linaro.org> Signed-off-by: Ross Burton <ross.burton@intel.com>
Derek Straka [Tue, 30 Jan 2018 03:04:39 +0000 (22:04 -0500)]
python-native: add dependency for gdbm and db native packages
These two packages are required to ensure the manifest files contain
all of the generated packages. Without this, the db and gdbm packages
will not contain the .so files as they are skipped during the compilation steps
Signed-off-by: Derek Straka <derek@asterius.io> Signed-off-by: Ross Burton <ross.burton@intel.com>
Huang Qiyu [Tue, 30 Jan 2018 08:58:38 +0000 (16:58 +0800)]
ruby: 2.4.2 -> 2.5.0
1.Upgrade ruby form 2.4.2 to 2.5.0.
2.Update the checksum of LIC_FILES_CHKSUM.
3.Delete ruby-CVE-2017-9224.patch, ruby-CVE-2017-9227.patch, ruby-CVE-2017-9229.patch, since it is integrated upstream.
4.Modify ruby-CVE-2017-9226.patch, since the data has been changed.
Signed-off-by: Huang Qiyu <huangqy.fnst@cn.fujitsu.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Paul Eggleton [Wed, 31 Jan 2018 19:32:07 +0000 (08:32 +1300)]
devtool: set up git repos so that singletask.lock is ignored
singletask.lock is written out while certain tasks execute for recipes
that have externalsrc.bbclass enabled - this includes recipes in
devtool's workspace. It appears that there's a race where
singletask.lock will be there one minute and then when we try to get the
file checksum of it (since we want to know if anything in the source
tree has changed) it will be gone, and git chokes. To fix that, add
singletask.lock to .git/info/exclude in the repository, regardless of
whether we created the repository or not. In any case singletask.lock
should never be tracked by git, so this is a good thing to be doing for
that reason as well.
This fixes oe-selftest failures in test_devtool_modify that we've seen
on the Yocto Project autobuilder:
bb.data_smart.ExpansionError: Failure expanding variable
do_compile[file-checksums], expression was ${@srctree_hash_files(d)}
which triggered exception CalledProcessError: Command
'['git', 'add', '-A', '.']' returned non-zero exit status 128.
Note that this only fixes this issue for devtool; if you are using
externalsrc independently of devtool there's a chance this will still
be an issue unless you add singletask.lock to your .gitignore.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
npm.bbclass: Node module name and recipe name can be different
Some NPM modules have the same name as their low level dependencies.
To prevent recipe naming conflicts, allow node module recipe names
to start with the "node-" prefix.
Signed-off-by: Zoltán Böszörményi <zboszor@pr.hu> Signed-off-by: Ross Burton <ross.burton@intel.com>
npm cache clear throws an error with npm@5 and suggests to
use npm cache verify instead. But our cache is actually empty,
so use npm cache clear --force.
npm install in the source directory creates symlinks with npm@5.
Use a combination of npm pack and npm install module-version.tgz
that works the same way with older and new npm versions and is
guaranteed to create actual copies instead of directory symlinks.
This change allows using nodejs 8.x LTS, tested with 8.9.4.
Signed-off-by: Zoltán Böszörményi <zboszor@pr.hu> Signed-off-by: Ross Burton <ross.burton@intel.com>
Joshua Watt [Wed, 31 Jan 2018 19:49:56 +0000 (13:49 -0600)]
waf.bbclass: cd to ${S} before checking version
waf requires that the current working directory be the project root (in
this case ${S} when it is invoked. The check to get the waf version was
being executed as a prefunc for do_configure, which meant it was
executed before the current working directory was switched to ${S}, and
thus would fail with some recipes. Fix this by changing to ${S} before
executing "waf --version"
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Max Krummenacher [Sat, 27 Jan 2018 13:54:08 +0000 (14:54 +0100)]
image_types: tar with --numeric-owner
If --numeric-owner is neither used when creating or extracting the archive
containing the rootfs then tar tries to change the numeric uid/gid of the
files based on user/group names of the host used to extract the archive.
Create the archive with --numeric-owner to remove the burden of having to
use --numeric-owner when extracting.
Signed-off-by: Max Krummenacher <max.krummenacher@toradex.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
We need to ensure that all xorg modules are linked with
SECURITY_X_LDFLAGS to ensure that they will be able to resolve their run
time dependencies. The approach of listing each driver in
security_flags.inc lets less frequently used drivers be run-time
broken. Move the flag logic into xorg-driver-common.inc so that all
xorg modules from all layers will have the correct security flags used.
Cc: Khem Raj <raj.khem@gmail.com> Signed-off-by: Tom Rini <trini@konsulko.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Niko Mauno [Wed, 31 Jan 2018 16:55:10 +0000 (18:55 +0200)]
bitbake.conf: Add xattr to DISTRO_FEATURES_NATIVE
Commit db1f1adace58763c35774e3fdfeaac5c3ca646fd disabled 'xattr' from
DISTRO_FEATURES when native packages are built, whereas for target
packages it was retained. This changed eg. mtd-utils-native to build
without extended attributes (capabilities) support from there on.
Thus even setcap was called succesfully for a binary during pkg_postinst
on build host, the capabilities did not exist on target rootfs due to
now xattr-less host-side mkfs utility.
Adding xattr to DISTRO_FEATURES_NATIVE restores previous behaviour,
fixing the discontinuity in capability propagation to target
device rootfs.
Signed-off-by: Niko Mauno <niko.mauno@vaisala.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Martin Jansa [Tue, 30 Jan 2018 16:41:23 +0000 (16:41 +0000)]
tcmode-default.inc: drop preferred version of gzip-native
* gzip was recently upgraded from 1.8 to 1.9, now all the builds show:
NOTE: preferred version 1.8 of gzip-native not available (for item gzip-native)
NOTE: versions of gzip-native available: 1.9
* drop the setting, because nobody is probably going to use older
gzip-native than 1.4 when there is only 1.9 available in oe-core
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Tue, 23 Jan 2018 00:54:13 +0000 (00:54 +0000)]
gettext: beat library detection into shape
For reasons I can't explain gettext uses several hundred lines of convoluted m4
to find the paths to a library.
If we don't tell it where to find a library it will hunt around and potentially
have host contamination as /usr/lib on the host is explicitly searched.
If we tell it the prefix to a library then we get bad RPATHs in the binaries
(such as /usr/lib/../lib), and the search assumes that it knows best about what
the library directories are under that prefix (even when it's wrong).
So, replace the lookup where possible with pkg-config calls (libxml2, glib,
libcroco). libunistring doens't have a pkgconfig file so just don't use the
system libunistring: the library is tiny anyway.
Martin Jansa [Tue, 30 Jan 2018 11:21:37 +0000 (11:21 +0000)]
tcmode-default.inc: bump LINUXLIBCVERSION to 4.14%
* linux-libc-headers were updated without updating PREFERRED_VERSION causing
following messages being shown in every single build:
NOTE: preferred version 4.12% of nativesdk-linux-libc-headers not available (for item nativesdk-linux-libc-headers)
NOTE: versions of nativesdk-linux-libc-headers available: 4.14.13
NOTE: preferred version 4.12% of linux-libc-headers not available (for item linux-libc-headers)
NOTE: versions of linux-libc-headers available: 4.14.13
NOTE: preferred version 4.12% of linux-libc-headers not available (for item linux-libc-headers-dev)
NOTE: versions of linux-libc-headers available: 4.14.13
NOTE: preferred version 4.12% of nativesdk-linux-libc-headers not available (for item nativesdk-linux-libc-headers-dev)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
package_rpm.bbclass: run pre/post installation scriptlets using sh -e
This allows catching errors in the scriptlets which would otherwise
go unnoticed, e.g. this sequence:
====
bogus_command
proper_command
====
would work just fine. Note that this patch needs all of the preceding
patches, as otherwise running failing scriptlets with -e would defer
them to first boot, instead of properly reporting failure and aborting
the package installation.
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
meta/lib/oe/package_manager.py: deprecate 'exit 1' as a way to defer to first boot
'exit 1' is not optimal for two reasons:
1) Code is hard to read; it is not obvious that it means 'defer what follows to first boot'.
2) Worse, this hides actual errors in the scriptlets; there is no difference between scriptlet
failing because it's intended to be run on target and scriptlet failing because there's a bug or
a regression somewhere.
The new, supported way is to place the code that has to run on target into pkg_postinst_ontarget(),
or, if a more fine-tuned control is required, call 'postinst-intercepts defer_to_first_boot' from
pkg_postinst() to explicitly request deferral to first boot.
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
package.bbclass: add support for pkg_postinst_ontarget()
This function is a convenient and more readable shortcut for situations
when the postinst code always needs to run on target. All commands that
cannot be executed during cross-install and can only be run on target
should go into this function. They will only be executed on first boot
(if package was cross-installed) or immediately during package installation
on target.
Plain pkg_postinst() works as before: it is run during cross-install time,
it can contain a request to defer to first boot, and it is also run
during package installation on target.
Also fix the oeqa test for this functionality to use the new function
where appropriate.
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
meta/lib/oe/rootfs.py: do not execute defer_to_first_boot when processing postinst_intercept hooks
That hook is empty, and doesn't need to be executed; it merely indicates that packages
that have used it are requesting to defer their postinst scripts to first boot
unconditionally.
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
