Patrick Ohly [Thu, 3 Sep 2015 18:42:26 +0000 (20:42 +0200)]
runqemu: support full-disk images
This makes it possible to boot images with multiple partitions (the
ones ending in .hddimg or .hdddirect) in several ways:
runqemu qemux86 core-image-minimal hddimg
runqemu tmp/deploy/images/qemux86/core-image-minimal-qemux86.hddimg
VM=tmp-glibc/deploy/images/qemux86/iot-os-image-qemux86.hddimg FSTYPE=hddimg runqemu
Same for hdddirect.
This is useful for testing initramfs scripts, secure boot (when
switching to UEFI), or boot loaders like syslinux. For testing the
content of the rootfs, the ext4 image is better because that approach
is faster (no need to create another large image during build, rootfs
can be read directly instead of reading boot.img through loop device).
When booting a live image, the kernel, initramfs (if any) and kernel
parameters are taken from the image by the virtual machine's BIOS, so any
additional kernel parameters given to runqemu are ignored. This can be
avoided (already without this change) in a slightly hacky runqemu setup:
ROOTFS=tmp/deploy/images/qemux86/core-image-minimal-qemux86.hddimg \
FSTYPE=ext4 \
KERNEL=tmp/deploy/images/qemux86/bzImage-initramfs-qemux86.bin \
MACHINE=qemux86 \
runqemu serial kvm nographic 'bootparams=root=/dev/ram0'
The additional bzImage-initramfs-qemux86.bin kernel here was created
by adding this to local.conf:
INITRAMFS_IMAGE = "core-image-minimal-initramfs"
INITRAMFS_IMAGE_BUNDLE = "1"
In the code, the new FSTYPE=hddimg resp. hdddirect behaves almost
exactly like the older vmdk FSTYPE. New types were chosen because it
seemed cleaner than using FSTYPE=vmdk when the actual image pointed to
by VM is not in that format. The downside is that several checks for
FSTYPE=vmdk had to be duplicated for FSTYPE=hddimg.
The VM variable now gets interpreted as "virtual machine disk image"
instead of "vmdk image".
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
oprofileui uses gettext during the configuration task so should be inherit
gettext. This issue appears when an older version of gettext is used do to
pinning to the older non-gplv3 version.
[YOCTO #7795]
Signed-off-by: Saul Wold <sgw@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
oeqa/decorators: Fixed a problem with tests having the same names.
When two or more tests had the same name but different classes then
the decorator log whould have the output all wrong. This was because
a comparison which was made only between method names but now it
compares classes too.
[YOCTO #8029]
Signed-off-by: Lucian Musat <george.l.musat@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
An rmdir call in do_install_append was trying to remove the
sbindir, however in a system with a merged usr this directory
is not empty and therefore failing to rm it causes an ERROR.
Instead check that sbindir != base_sbindir before trying to
remove the directory.
Signed-off-by: Joshua Lock <joshua.lock@collabora.co.uk> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The Makefile for mount and osd_login utilities forces
/sbin as asbindir, however on a merged /usr system this directory
might not exist. Instead sed in the system sbindir.
Signed-off-by: Joshua Lock <joshua.lock@collabora.co.uk> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
pcmciautils: handle udev dir being 2 levels below /
When building with a merged /usr dir the udev directory
lives at /usr/lib/udev - update the FILES pattern to also
pick up udev files installed two levels below the / to
ensure a merged /usr works.
Signed-off-by: Joshua Lock <joshua.lock@collabora.co.uk> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When building with a merged /usr dir the udev directory
lives at /usr/lib/udev - update the FILES pattern to also
pick up udev files installed two levels below the / to
ensure a merged /usr works.
Signed-off-by: Joshua Lock <joshua.lock@collabora.co.uk> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
pulseaudio: handle udev dir being 2 levels below /
When building with a merged /usr dir the udev directory
lives at /usr/lib/udev - update the FILES pattern to also
pick up udev files installed two levels below the / to
ensure a merged /usr works.
Signed-off-by: Joshua Lock <joshua.lock@collabora.co.uk> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
alsa-utils: handle udev dir being 2 levels below /
When building with a merged /usr dir the udev directory
lives at /usr/lib/udev - update the FILES pattern to also
pick up udev files installed two levels below the / to
ensure a merged /usr works.
Signed-off-by: Joshua Lock <joshua.lock@collabora.co.uk> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 4 Sep 2015 15:59:38 +0000 (16:59 +0100)]
oeqa: Test failure/cleanup improvements
Currently, if qemu segfaults, the tests merrily continue trying to execute
which takes time for them to timeout and is a bit silly. Worse, no logs about
the segfault are shown to the user, its silent!
This patch tries to unravel the tangled web of issues and ensures that we:
* install a SIGCHLD handler which tells the user qemu exited
* check if qemu is running, if it isn't fail the test outright
* don't leave processes behind in sshcontrol which would hold
bitbake.lock and block shutdown
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 4 Sep 2015 15:27:41 +0000 (16:27 +0100)]
tcmode-default: Set gcc 5.2 as the default
gcc 5.X is now working in all the places we test it in, its been in
testing for quite some time. Time to make it the default (we have some
room in M4 for any other bugfixes). Its easy to switch back to 4.9, we
should really remove 4.8 at this point (to meta-oe?).
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If a test case is decorate by oeqa.utils.decorators.tag, this case will
by add a tag, testrunner will filter these tags by TEST_SUITES_TAGS
[YOCTO #7849]
Signed-off-by: zjh <junhuix.zhang@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
gcc-target.inc: Remove non-related gcc headers from include_fixed folder
Without this patch, the D's include_fixed folder may change after building it
(due to the gcc's fixinc.sh script, executed on the do_compile task) and changes
depend on the current sysroot headers, making the gcc's builds non-deterministic.
[YOCTO #7882]
Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Wed, 26 Aug 2015 14:13:46 +0000 (17:13 +0300)]
devtool: run kernel dependencies
The kernel package needs "kern-tools-native" in order for it's
do_kernel_metadata. Thus, devtool extract for kernel in a pristine
environment fails. With the current bb.tinfoil implementation it is not
possible to run arbitrary bitbake commands - e.g. run
"bitbake kern-tools-native -c populate_sysroot" in our case. This patch
implements an ugly workaround for that problem, basically by hardcoding
this dependency and running the required bitbake task(s) before tinfoil
is initialized.
[YOCTO #6658]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
Markus Lehtonen [Wed, 26 Aug 2015 12:29:41 +0000 (15:29 +0300)]
devtool: extract: correct initial rev for kernel packages
Change handling of kernel packages so that the "initial rev" is parsed
correctly. Also, the devtool-specific git tags (devtool-base and
devtoo-patched) are now generated for kernel packages as well.
[YOCTO #6658]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
Markus Lehtonen [Tue, 18 Aug 2015 14:13:12 +0000 (17:13 +0300)]
devtool: make required tasks be run in kernel build
Set SRCTREECOVEREDTASKS appropriately in the workspace .bbappend file
for kernel recipes. This tries to ensure that all needed tasks (esp.
configure and patch) are run when building the kernel - tasks which
would normally be disabled by externalsrc.bbclass.
[YOCTO #6658]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
libunwind: Invalid dwarf opcodes can cause references beyond the end of
the array
Off-by-one error in the dwarf_to_unw_regnum function in include/dwarf_i.h
in libunwind 1.1 allows local users to have unspecified impact via
invalid dwarf opcodes.
Signed-off-by: Li Zhou <li.zhou@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jian Liu [Tue, 25 Aug 2015 08:29:01 +0000 (16:29 +0800)]
sdk.py: fix conflicts of packages
If packages are conveyed to smart to install at the same time,
conflicts will not happen.
Try to install packages into sdk image at the same time.
This patch is not so perfect. For example,
IMAGE_INSTALL += "lib32-ncurses"
IMAGE_INSTALL += "ncurses-dev"
ncurses-dev and lib32-ncurses-dev will have conflicts during packages installation.
Signed-off-by: Jian Liu <jian.liu@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This function is intended to be used in ROOTFS_POSTPROCESS_COMMAND, and checks
for any paths outside of /home which are owned by the user running bitbake.
Signed-off-by: Christopher Larson <chris_larson@mentor.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
binutils: Fix symbols with --dynamic-list when using gold linker
There are random crashes and hangups, seen in libQtCore (QT4) as mentioned
in the binutils bugs as well, the reason is that
gold ignores all other symbols except specified in --dynamic-list which
is different in behavior as compared to bfd linker. The patch is a
backport from upstream master. This patch implements the bfd linker's
behaviour into gold.
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
qemurunner: Handle lack of data on run serial gracefully
This changes the behavior when data was not received over
the serial console when a command is run. With this the
socket is no longer closed but it throws and exception that
can handled in upper layers. With this the test can continue
without throwing errors for not having the socket anymore.
[YOCTO #8118]
Signed-off-by: Mariano Lopez <mariano.lopez@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ed Bartosh [Wed, 2 Sep 2015 10:58:17 +0000 (13:58 +0300)]
image.py: rename _write_env -> _write_wic_env
Renamed this function as it's too generic name for it.
It writes variables, which are used by wic to .env file,
so _write_wic_env is better name for it.
Thanks Christopher Larson for poining out to this.
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ed Bartosh [Wed, 2 Sep 2015 10:58:15 +0000 (13:58 +0300)]
wic: get rid of listing properties
Functionality of listing and using properties of wic images
does not exist in the wic code. However, there are plenty of
help and usage content about it, which is very confusing.
Removed everything regarding image properties from wic codebase.
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ed Bartosh [Wed, 2 Sep 2015 10:58:03 +0000 (13:58 +0300)]
oe-selftest: wic: configure a build
Added MACHINE_FEATURES and IMAGE_FSTYPES variables to the build config
to ensure that bootimg and efi artifacts are built. This is needed to
build canned wic images that are using those artifacts.
Introduced class variable Wic.image_is_ready to avoid building image
more than once. It would be better to build image in setUpClass, which
is called only once, but add_config method can't be called in class method,
so we have to use this trick with class variable in setUp method.
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Li Zhou [Tue, 25 Aug 2015 06:49:34 +0000 (14:49 +0800)]
valgrind: enable rt_sigpending syscall on ppc64 linux
When running one application under Valgrind on ppc64 arch, we got a
missing syscall error. Get upstream patch from valgrind website to
enable rt_sigpending syscall on ppc64 linux.
Signed-off-by: Li Zhou <li.zhou@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Kevin Hao [Tue, 25 Aug 2015 10:36:07 +0000 (18:36 +0800)]
kernel.bbclass: add the runtime dependency on kernel-vmlinux for kernel-image
When a BSP uses vmlinux for boot, the kernel-image package is just
empty. But by default the kernel-vmlinux is not installed. Then the
pkg_postinst_kernel-image() would create a symlink to a non-existent
file. Fix this by adding the runtime dependency on kernel-vmlinux for
kernel-image if the KERNEL_IMAGETYPE is "vmlinux".
Signed-off-by: Kevin Hao <kexin.hao@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Patrick Ohly [Wed, 26 Aug 2015 16:10:45 +0000 (18:10 +0200)]
mtd-utils: add xattr PACKAGECONFIG and fix acl dependency
The unconditional removal of -DWITHOUT_XATTR accidentally introduced a
compile-time dependency on "acl", because "sys/acl.h" gets
included. This caused random compile failures.
To fix this, we introduce a proper PACKAGECONFIG for the "xattr"
support, with the distro's "xattr" feature determining the
default. Setting the define must be done with custom code because
PACKAGECONFIG can only influence configure parameters.
The "acl" distro feature is not checked because although enabling
"xattr" support now triggers a build of acl, nothing from it will not
get included in the resulting target image. This seems more suitable
than not enabling xattr support in mtd-utils when acl support in the
distro is disabled.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Those code fragments date back to ancient times. EXTRA_OECONF is not
required anymore and we should give gcc another try to do it right.
Testing on cortex-a8 with thumb and -O2 reveals a performance boost of
82 percent during encoding in comparison to -O0. -O4 gives another 7
percent.
Signed-off-by: Stefan Müller-Klieser <s.mueller-klieser@phytec.de> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Tue, 25 Aug 2015 13:48:32 +0000 (16:48 +0300)]
package_manager: support for signed RPM package feeds
This change makes it possible to create GPG signed RPM package feeds -
i.e. package feed with GPG signed metadata (repodata). All deployed RPM
repositories will be signed and the GPG public key is copied to the rpm
deployment directory.
In order to enable the new feature one needs to define four variables in
bitbake configuration.
1. 'PACKAGE_FEED_SIGN = "1"' enabling the feature
2. 'PACKAGE_FEED_GPG_NAME = "<key_id>"' defining the GPG key to use for
signing
3. 'PACKAGE_FEED_GPG_PASSPHRASE_FILE = "<path_to_file>"' pointing to a
file containing the passphrase for the secret signing key
4. 'PACKAGE_FEED_GPG_PUBKEY = "<path_to_pubkey>"' pointing to the
corresponding public key (in "armor" format)
The user may define "GPG_BIN" in the bitbake configuration in order to
specify a specific the gpg binary/wrapper to use for signing.
[YOCTO #8134]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Tue, 25 Aug 2015 11:49:14 +0000 (14:49 +0300)]
os-release: add the public package-signing key
Adds the public package-signing key into this package. It will be
installed under /etc/pki/rpm-gpg if the RPM signing feature is used. The
key file is not currently directly used by anything in the target
system. It is merely there for possible later use.
[YOCTO #8134]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Fri, 21 Aug 2015 14:21:57 +0000 (17:21 +0300)]
package_rpm: support signing of rpm packages
This patch adds a new bbclass for generating rpm packages that are
signed with a user defined key. The packages are signed as part of the
"package_write_rpm" task.
In order to enable the feature you need to
1. 'INHERIT += " sign_rpm"' in bitbake config (e.g. local or
distro)
2. Create a file that contains the passphrase to your gpg secret key
3. 'RPM_GPG_PASSPHRASE_FILE = "<path_to_file>" in bitbake config,
pointing to the passphrase file created in 2.
4. Define GPG key name to use by either defining
'RPM_GPG_NAME = "<key_id>" in bitbake config OR by defining
%_gpg_name <key_id> in your ~/.oerpmmacros file
5. 'RPM_GPG_PUBKEY = "<path_to_pubkey>" in bitbake config pointing to
the public key (in "armor" format)
The user may optionally define "GPG_BIN" variable in the bitbake
configuration in order to specify a specific gpg binary/wrapper to use.
The sign_rpm.bbclass implements a simple scenario of locally signing the
packages. It could be replaced by a more advanced class that would
utilize a separate signing server for signing the packages, for example.
[YOCTO #8134]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Thu, 27 Aug 2015 06:32:47 +0000 (09:32 +0300)]
createrepo: disable RPM signature validation
Disable RPM signature validation so that it is possible to create
package feeds of signed RPM packages without importing the public part
of the signing key into the RPM database. In any case, the signatures
are validated when the packages in the feed are used (e.g. in image
generation of manually installing packages from the feed).
The original idea idea of this patch is from Mark Hatle
<mark.hatle@windriver.com>.
[YOCTO #8134]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently it is not possible to run a exported test,
but this patch will allow to use the HosDumper class
when running a exported test, otherwise the HostDumper
class will break runexpored test.
[YOCTO #8118]
Signed-off-by: Mariano Lopez <mariano.lopez@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
qemurunner: Added host dumps when there are errors
This adds an instance of HostDumper to qemurunner,
with this instance now is possible to get dumps
from the host when there is an error.
This adds dump points in the next cases:
- runqemu exits before seeing qemu pid
- Fail to get qemu process arguments
- Not reach login banner before timeout
- qemu pid never appears
This also modifies the constructors of BaseDumper,
HostDumper and TargetDumper, they don't require
the datastore anymore, but the feature to replace
datastore variables has been lost (never used)
[YOCTO #8118]
Signed-off-by: Mariano Lopez <mariano.lopez@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changes required:
- Rename opkg-cl to opkg
- Add libarchive dependency
- Drop backport patches
- Drop obsolete directory options
- Add patch to handle empty index files
Based on initial work by Paul Barker.
Signed-off-by: Alejandro del Castillo <alejandro.delcastillo@ni.com> CC: Paul Barker <paul@paulbarker.me.uk> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
check_snprintf.patch has been dropped, because it seems to fix a problem
that doesn't anymore exist, and doesn't have any description of what the
problem was and how was it fixed.
tarfix.patch has been merged upstream.
The rest of the patches have been rebased to the new upstream release
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
LICENSE checksum has changed because the copyright year was changed
from 2014 to 2015 in it:
https://github.com/mkj/dropbear/commit/19e1afbd1ca6d306166ce74bcd6c6889f8d196f3
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Martin Jansa [Fri, 28 Aug 2015 16:07:16 +0000 (18:07 +0200)]
pulseaudio: improve PACKAGECONFIG for dbus/consolekit dependencies
* the problem is that consolekit module is built whenever dbus is enabled
and consolekit is available only in distributions with x11 in DISTRO_FEATURES
* many distributions want to enable dbus support (required for bluez support),
but without consolekit because they aren't using x11
* allow to completely disable dbus (and consolekit)
* add consolekit runtime dependency only for x11 in DISTRO_FEATURES
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jussi Kukkonen [Thu, 27 Aug 2015 13:05:59 +0000 (16:05 +0300)]
gmp: Bring back version 4.2.1 (LGPL 2.1+)
gmp 4.2.1 was removed in f181c6ce8b apparently accidentally: It
was not noticed that 4.2.1 is LGPL 2.1 (and not GPL) so provides
a useful alternative to the newer "GPLv2 | LGPLv3" version.
* Reintroduce 4.2.1. The source includes files that are GPL but the
library package is LGPL 2.1+
* Also reintroduce the two patches removed in f181c6ce8b.
* Refactor gmp.inc: gmp 6.0.0 build should not be affected in any way.
* Update 6.0.0 license from "GPLv2 | LGPLv3" to "GPLv2+ | LGPLv3+".
[YOCTO #8197]
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jussi Kukkonen [Thu, 27 Aug 2015 13:05:58 +0000 (16:05 +0300)]
nettle: Fix binary license to LGPLv2.1+
* Set the nettle binary package license to LGPLv2.1+:
There are GPL files in the sources but none of these are used
to produce the files we ship.
* Remove the useless package specific licenses: none of the named
packages are actually produced and the licenses do not affect
the overall license of either the sources or the binary package.
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
