Currently kernel-fitimage gets the dtb from KERNEL_DEVICETREE, however
there are instances when the DTB file can be generated using other
recipes, provide a way to include the DTB instead of inkernel device
tree.
Add support for external device tree during the creation of fitimage. If
virtual/dtb is set using PREFERRED_PROVIDER_virtual/dtb = "devicetree", which
inherits the devicetree.bbclass then use the path provided in
EXTERNAL_KERNEL_DEVICETREE else use KERNEL_DEVICETREE during fitimage process
(From OE-Core rev: 084f4de4dbaf9821516fc0254d35f4fb04311d27)
Signed-off-by: Manjukumar Matha <manjukumar.harthikote-matha@xilinx.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The AArch64 little-endian ABI requires that the dynamic loader is always
available at /lib/ld-linux-aarch64.so.1. Similarly, the big-endian ABI
requires that the dynamic loader is always available at
/lib/ld-linux-aarch64_be.so.1.
glibc-package.inc contains code that tries to ensure this, but
unfortunately it is defeated by the combination of multilib and usrmerge
because it does not take into account that /lib is the same as /usr/lib
with usrmerge when it adds the loader path to libc_baselibs and when it
attempts to show that /usr/lib is empty in do_poststash_install_cleanup.
This results in the symlink not being included in the package and a build
failure due to rmdir failing.
Richard Purdie also suggested[1] that ${nonarch_base_libdir} should not be
used as a synonym for /lib in this case.
This hopefully-fixed version always sets ARCH_DYNAMIC_LOADER and then uses
${root_prefix}/lib/${ARCH_DYNAMIC_LOADER} to refer to the dynamic loader
which works with both multilib and usrmerge. Since ARCH_DYNAMIC_LOADER is
only non-empty if the symlink is required, the code to create it can move
to do_install_append. Then do_poststash_install_cleanup needs to be taught
that ${exec_prefix}/lib may not be empty if the dynamic loader symlink is
there.
It appears not to be possible to specify the name of the loader via a
variable with an override, since the _aarch64 override is applied even for
_aarch64-be, so I've set the loader name using ${TARGET_ARCH} instead.
Build-tested and inspected core-image-minimal rootfs with:
* AArch64 no multilib (real loader in correct place)
MACHINE = "qemuarm64"
* AArch64 multilib (symlink in correct place)
MACHINE = "qemuarm64"
MULTILIBS = "multilib:lib32"
DEFAULTTUNE_virtclass-multilib-lib32 = "armv7at-neon"
require conf/multilib.conf
* AArch64 usrmerge (real loader in correct place)
DISTRO_FEATURES += "usrmerge"
MACHINE = "qemuarm64"
* AArch64 multilib usrmerge (symlink in correct place)
DISTRO_FEATURES += "usrmerge"
MACHINE = "qemuarm64"
MULTILIBS = "multilib:lib32"
DEFAULTTUNE_virtclass-multilib-lib32 = "armv7at-neon"
require conf/multilib.conf
* big-endian versions of all of the above by also setting
DEFAULTTUNE = "aarch64_be".
(building glibc only.)
* x86_64 (real loader in /lib as before)[2]
MACHINE = "qemux86"
* x86_64 multilib (real loader in /lib64 as before)
MACHINE="qemux86-64"
MULTILIBS = "multilib:lib32"
DEFAULTTUNE_virtclass-multilib-lib32 = "x86"
require conf/multilib.conf
I also tested leaving an unwanted file in ${exec_prefix}/lib for
do_poststash_install_cleanup to detect, and I believe the detection always
worked correctly.
[1] http://lists.openembedded.org/pipermail/openembedded-core/2018-November/276120.html
(From OE-Core rev: a705c0782c863ee960d65b5109168a4587a0a7b7)
Signed-off-by: Mike Crowe <mac@mcrowe.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
For multilib to work correctly, BASE_LIB overrides must be provided for
each new tune added in this file.
(From OE-Core rev: e39c5ec90ebbc37064c9cd59eba12603317740cd)
Signed-off-by: Mike Crowe <mac@mcrowe.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changelog:
http://alsa-project.org/main/index.php/Changes_v1.1.6_v1.1.7
The two patches have now been upstreamed. Apart from those changes, the
only thing that changed was the hwmixvolume tool, which got ported from
PyGTK to PyGObject, and from GTK 2 to GTK 3.
When testing hwmixvolume, I found that it depends on pyalsa, which has
not been packaged for OE. I believe hwmixvolume has never worked on OE.
It certainly didn't work before this patch, trying to build it failed
due to python-pygtk not being available. Even if python-pygtk was
available at some point in the past, hwmixvolume has always used pyalsa,
but the alsa-tools recipe has never had that dependency declared.
(From OE-Core rev: b1426844235a64246d46d71cf826e871c92d7ed4)
Signed-off-by: Tanu Kaskinen <tanuk@iki.fi>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
A revamp of the syscalls/setregid tests made to use a new
library [LTP:7a1b4427] required a conversion to test messages.
This test in particular did not include a formatted string,
but a just a string element:
testcases/kernel/syscalls/setregid/setregid01.c
This patch allows LTP to be built with -Wformat,
-Wformat-security, and -Werror=format-security.
(From OE-Core rev: ecee899dec53d324053112ed8764268fef930408)
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
To avoid conflicts with other forks of glibc remove the diff from the
ChangeLog.
(From OE-Core rev: 695d79af1edcc76a01055b01922f0d106c8291ca)
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Perl and its dependencies have a decent footprint impact. On my
xz compressed filesystem:
634880: /usr/lib/libperl.so.5.24.4
Put c_rehash in the openssl-misc package so the dependency can be
avoided where it isn't needed.
Change-Id: Iae9bccabfb1c8cfa1401ca6785abc39713d3fdf0
(From OE-Core rev: d2b1a889ef8fb9e6a2fa3d9bfc3eaf6113db9b1f)
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In Ubuntu 16.04 LTS userspace is build for PowerPC 32-bit while kernel
selected by the installer depending on PowerPC machine type:
* 32-bit for PowerMac G4 (ppc7400) and below
* 64-bit for PowerMac G5 and above
Thus uname(2) returns ppc64 for 64-bit kernels and 32-bit userspace
making build impossible due to missing some of lib64 multilib
equivalents in Ubuntu repository.
Using setarch(8) override to make whole host look as PowerPC 32-bit
can actually help with build but requires mapping for ppc target to
their libgpg-error equivalent to fix native build.
Build tested on Ubuntu 16.04 LTS host on PowerMac G5 with command:
MACHINE=qemuppc setarch ppc bitbake core-image-full-cmdline
(From OE-Core rev: e81a5a640a2a00c43796cd47e168c93bb389a6c8)
Signed-off-by: Serhey Popovych <serhe.popovych@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Building musl for powerpc64 enforces elfv2 ABI since commit 68c964185519
("arch-powerpc64.inc: Use elfv2 ABI when building with musl") that is
not compatible with valgrind.
(From OE-Core rev: d4eb90b5a4bc5fd8619120f783bc78b0e7dc829e)
Signed-off-by: Serhey Popovych <serhe.popovych@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Use commit 4f07ffa8f5ab ("Use #if instead of #ifdef for ENABLE_GSHADOW")
from upstream to check ENABLE_GSHADOW correctly that is defined as 0
in case of musl.
While there replace specific patch with one from upstream that does
exactly the same commit 66a5b5ce9b99 ("basic/user-util: properly protect
use of gshadow").
(From OE-Core rev: c9580ef0810196f6703567d9db458b73dbbfb35f)
Signed-off-by: Serhey Popovych <serhe.popovych@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Licence checksum updated because the copyright dates were changed.
(From OE-Core rev: f3896b69c78abeaefc1c60e7a6d7b2ed85eb7015)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
meson has been found to set it erroneously.
(From OE-Core rev: 920824125eb87928a329a072fa6c5a839efc932b)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
PRINTF settings no longer seem necessary (tested with mingw).
Add meson-specific bits to Enable-more-tests-while-cross-compiling.patch and
0001-Install-gio-querymodules-as-libexec_PROGRAM.patch
Add 0001-Set-host_machine-correctly-when-building-with-mingw3.patch
to allow 'mingw32' as target machine in addition to 'windows'.
Add 0001-meson.build-do-not-hardcode-linux-as-the-host-system.patch to
correct ARM build errors.
Drop configure-libtool.patch (autotools-specific).
Fix API docs generation
(From OE-Core rev: e185235dd97510bfdc621cef9c18d8d13b16006d)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Drop 0001-Disable-tests-in-native-builds.patch as upstream has fixed the issue.
Rework 0001-Fix-a-couple-of-decisions-around-cross-compilation.patch to
enable generation of loaders.cache in target builds (using a native tool),
because otherwise building tests becomes impossible.
Rework 0002-Work-around-thumbnailer-cross-compile-failure.patch into
0003-target-only-Work-around-thumbnailer-cross-compile-fa.patch
(which deals with substituting native tools in cross builds) and
0004-Do-not-run-tests-when-building.patch (which avoids running
test binaries during cross builds).
Rebase fatal-loader.patch.
License checksum updates as COPYING file had 2.0 version of LGPL and
has been replaced with 2.1 version.
Take meson's x11 and installed_tests options into use.
Install gdk-pixbuf-query-loaders also into $bindir, as we need the native
version during cross compile.
(From OE-Core rev: 8fb16df2e6977bb3508239eb8d447b8c5401d5ed)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
ltp doesn't build cleanly with the stringformat security flags, work around
this until upstream sorts out the issues.
(From OE-Core rev: ce0e5719c9b10121c32be67237b88b0289499f2f)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This patch has been refreshed:
* 0008-Check-if-__GLIBC_PREREQ-is-defined-before-using-it.patch
These patches have been dropped as they have been merged:
* 0001-netns_helper.sh-use-ping-6-when-ping6-is-not-avaliab.patch
* 0001-setrlimit05-Use-another-method-to-get-bad-address.patch
* 0001-sigwaitinfo01-recent-glibc-calls-syscall-directly.patch
* 0001-statx-fix-compile-errors.patch
* 0001-syscalls-fcntl-make-OFD-command-use-fcntl64-syscall-.patch
* 0001-getcpu01-Rename-getcpu-to-avoid-conflict-with-glibc-.patch
This patch has been added:
* define-sigrtmin-and-sigrtmax-for-musl.patch
Added these rdependencies:
* file: for ld01, file01 and logrotate
* quota: for quota_remount_test01
(From OE-Core rev: d198364c1007aab6523787f9e1bb6bb274cc89ad)
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
GTK_THEME instructs gtk to use its built-in theme. Otherwise
gtk attempts to use the theme from the host, which may be from
a totally mismatching gtk version.
On the other hand FONTCONFIG_PATH tells it to use the host fonts,
as providing fonts in the native sysroot and instructing the components
to use them is a lot more tricky.
GDK_PIXBUF_MODULE_FILE is set, because otherwise qemu works but
fills stdout with error messages, which eventually fill the pipe
they go into. That pipe is read from only when qemu exits (to
collect any error messages) by runqemu script. The pipe fill-up
causes the qemu process to lock up.
(From OE-Core rev: d5b8a65861adba298a9dc2c4e9a7a7da5490073a)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
| In file included from /home/alexander/development/poky/build-virgl-gtk-64/tmp/work/x86_64-nativesdk-pokysdk-linux/nativesdk-qemu/3.1.0-r0/qemu-3.1.0/include/ui/egl-context.h:5,
| from /home/alexander/development/poky/build-virgl-gtk-64/tmp/work/x86_64-nativesdk-pokysdk-linux/nativesdk-qemu/3.1.0-r0/qemu-3.1.0/ui/egl-context.c:3:
| /home/alexander/development/poky/build-virgl-gtk-64/tmp/work/x86_64-nativesdk-pokysdk-linux/nativesdk-qemu/3.1.0-r0/qemu-3.1.0/include/ui/egl-helpers.h:46:55: error: unknown type name 'Window'; did you mean 'minor'?
| EGLSurface qemu_egl_init_surface_x11(EGLContext ectx, Window win);
| ^~~~~~
| minor
(From OE-Core rev: 01324f5e741a106cdb4b2989379f71b0b727a224)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The warnings deal with two specific cases of failure:
1) unable to execute nativesdk postinsts for mingw SDKs because
they need to be run under wine
2) unable to execute target postinsts when there is no qemu
usermode support for the target
Neither of these should be a big problem as mingw issue was never
found to problematic, and target postinsts deal with things that
are needed at runtime and not at build time which is the purpose
of SDKs.
The specific reason to do this is to reduce the amount of warnings
shown by the Yocto autobuilder, to zero eventually.
(From OE-Core rev: cbc32fcd9b52e750600cce9dd84b33e3ce612eae)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Upgrade dpkg from 1.18.25 to 1.19.4:
* update SRC_URI which DEBIAN_MIRROR
* un-export PERL which causes script scripts/dpkg-architecture.pl fails
to get cpu type and package arch
* add configure option '--disable-update-alternatives' which causes
installed-vs-shipped issue
* add patch for dpkg-native to resolve host tar may not support
options --owner=NAME:ID and --group=NAME:ID
* update context of patches
* drop dpkg-configure.service that apt provides a systemd timer to do it
(From OE-Core rev: 5185b6fe3d4fa9b53baffa6ee419da60ac056da2)
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The license checksum changed due to a change in copyright years.
(From OE-Core rev: ee3057bf109b2d11ecb0089c2c424e9b6033932e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Instead of using a native specific packageconfig option, we now set
the swrast as the supported driver.
Currently the native version does not build any dri drivers at all.
This is intentional: mesa-native is used only by virgl, so was made
the most minimal possible configuration for mesa-native so that a)
virgl works; and b) build time is as short as possible.
There are two swrast drivers, one dri-based, another gallium-based.
Autotools lets you build both, meson forces you to choose.
Thus DRIDRIVERS = "swrast" is only set for _native, as for _target
we get the gallium driver instead.
See discussion here [1]
Based on https://patchwork.openembedded.org/patch/158748/
[1] https://patchwork.openembedded.org/patch/158806/
(From OE-Core rev: 223ee432b908180932bf70e77d4058964d7f0786)
Signed-off-by: Fabio Berton <fabio.berton@ossystems.com.br>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The perl-cross uses PERL_API_VERSION but not PERL_VERSION to define
libperl soname: libperl.so.$PERL_API_REVISION.$PERL_API_VERSION.$PERL_API_SUBVERSION.
For perl stable releases, the PERL_API_SUBVERSION is always 0. Specify
the so version for libperl to make sure the libperl.so can be linked to
the correct library.
(From OE-Core rev: 6257fcd86e63fe533aeaadbc65a54213b4885c16)
Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
bitbake gdk-pixbuf gdk-pixbuf-native
bitbake build-sysroots -c build_target_sysroot
can lead to tracebacks as gdk-pixbuf-native is being installed into the
target sysroot. The issue is that the x86_64 (common BUILD_ARCH) sysroot
components directory can contain a mix of native and target artefacts.
Differentiate by the "-native" in the recipe names. Should also trim
down the size of the sysroot used in eSDK.
(From OE-Core rev: 57f32836060bba0eaa6b36f53146dd6fd07b77ed)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When you install the eSDK in publish mode and then try to build a
derivative sdk using devtool build-sdk, the following error happens
| NOTE: Generating sstate task list...
| NOTE: Generating sstate-cache...
| NOTE: Generating sstate-cache...
| DEBUG: Python function copy_buildsystem finished
| NOTE: Executing install_tools ...
| DEBUG: Executing shell function install_tools
| install: cannot stat /core/meta/files/ext-sdk-prepare.py':
No such file or directory
This patch will fix the error by installing ext-sdk-prepare.py in
publish mode, so that derivative sdk can be generated.
(From OE-Core rev: aab4089a57a655ca5a7792f05eb5d8dcdb934ae1)
Signed-off-by: Manjukumar Matha <manjukumar.harthikote-matha@xilinx.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
