Ross Burton [Fri, 10 Jan 2020 17:24:21 +0000 (17:24 +0000)]
perl: improve reproducibility
Occasionally the reproducibility selftest fails because perl-pod differs,
specifically that the perltoc.pod file was sometimes missing modules.
Debugging revealed that there are missing dependencies so there is a build race:
building perltoc.pod from an clean build tree results in no modules being listed
at all.
A bug has been filed at https://github.com/arsv/perl-cross/issues/86 to solve
this properly, but for now we can just delete perltoc.pod after make has
finished and re-generate it.
[ YOCTO #13726 ]
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Wed, 8 Jan 2020 19:48:07 +0000 (13:48 -0600)]
runqemu: Add network bridge support
Qemu supports attaching the virtual machine to an existing network
bridge interface via the qemu-bridge-helper program (as long as the
system is correctly configured to give the user permissions). Add
support for runqemu to do this also via the "bridge=<INTERFACE>"
argument. Note that for this to work correctly, the host
qemu-bridge-helper must be used, not the one that might have been built
by qemu-native. In order for qemu to correctly find this program, a
qemu-oe-bridge-helper program has been added to qemu-helper-native, and
runqemu will use this helper as the bridge helper. The helper will look
for the host qemu-bridge-helper first by looking in the
QEMU_BRIDGE_HELPER environment variable, then by search common paths
where the helper is installed.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Diego Rondini [Wed, 8 Jan 2020 16:48:55 +0000 (17:48 +0100)]
cups: fix crossscripts
Fix crossscripts to report the correct "serverbin" value.
While the packaged "cups-config --serverbin" reported
"/usr/libexec/cups" the crossscripts version reported
"/usr/lib/cups", causing packaging issues when building for example
cups-filters.
Also fix FILES_${PN} to use ${libexecdir}; previously it was working
just because "${libexecdir}/*" was part of the default values in
bitbake.conf.
Signed-off-by: Diego Rondini <diego.rondini@kynetics.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Khem Raj [Wed, 8 Jan 2020 16:42:54 +0000 (08:42 -0800)]
opensbi: Use -ffreestanding compiler options
This usually helps compiler to not use built-in optimizations
for string functions and mem* functions unnecessarily, since OE's
compiler is hosted for linux cross compile, this also means we instruct
it rightly for building a baremetal application
Refresh the patches to apply without fuzz
Fixes linking errors e.g.
platform/qemu/virt/lib/libplatsbi.a(sbi_string.o): in function `sbi_memset':
| /mnt/b/yoe/build/tmp/work/riscv64-yoe-linux/opensbi/0.5-r0/git/lib/sbi/sbi_string.c:102: undefined reference to `memset'
| collect2: error: ld returned 1 exit status
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tanu Kaskinen [Mon, 6 Jan 2020 09:18:05 +0000 (11:18 +0200)]
alsa-utils: 1.1.9 -> 1.2.1
I moved the ALLOW_EMPTY line next to the line that sets FILES to "".
That makes it easier to see that the package is not only allowed to be
empty, but forced to be empty.
Signed-off-by: Tanu Kaskinen <tanuk@iki.fi> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tanu Kaskinen [Mon, 6 Jan 2020 09:18:04 +0000 (11:18 +0200)]
alsa-topology-conf: new recipe, version 1.2.1
The topology configuration files were moved from the alsa-lib repository
to a new alsa-topology-conf repository. The move was accompanied by
a license change from LGPL2.1 to BSD-3-Clause.
Signed-off-by: Tanu Kaskinen <tanuk@iki.fi> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tanu Kaskinen [Mon, 6 Jan 2020 09:18:03 +0000 (11:18 +0200)]
alsa-ucm-conf: new recipe, version 1.2.1.2
The UCM configuration files were moved from the alsa-lib repository to a
new alsa-ucm-conf repository. The move was accompanied by a license
change from LGPL2.1 to BSD-3-Clause.
Signed-off-by: Tanu Kaskinen <tanuk@iki.fi> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This upgrade brings support for the new "UCM2" configuration format.
Another big change is that the UCM and topology configuration files were
moved to separate repositories, so those are removed from alsa-lib and
will need to be packaged separately.
The topology API was moved from libasound to a separate libatopology
library. This seems like a recipe for regressions for applications that
used to use libasound for the topology stuff, but I don't think there
are many applications in existence that use the topology API. I hope the
upstream knows what they are doing.
Patches added:
0001-configure.ac-remove-an-unnecessary-libtool-fix.patch
This fixes a failure during configure.
0001-ucm-Use-strncmp-to-avoid-access-out-of-boundary.patch
0002-ucm-return-always-at-least-NULL-if-no-list-is-availa.patch
Backports. I don't know if these are regression fixes, but since the UCM
code was heavily worked on in this upgrade, it seems likely that these
patches fix regressions.
0003-namehint-correct-the-args-check.patch
0004-namehint-improve-the-previous-patch-check-the-return.patch
Backports. A change in alsa-plugins caused a regression in the pulse
plugin, but that was actually due to a bug in alsa-lib, which is fixed
by these patches.
0005-ucm-Do-not-fail-to-parse-configs-on-cards-with-an-em.patch
Backport. Fixes a regression with some Intel hardware.
0001-Fix-alsa-sound-.h-for-external-programs.patch
0001-uapi-Move-typedefs-from-uapi-to-sound.patch
Backports. These fix build failures in alsa-tools.
I removed alsa-doc from PACKAGES, because no files were assigned to it.
Signed-off-by: Tanu Kaskinen <tanuk@iki.fi> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
populate_sdk_ext.bbclass: No longer needed to clean away conf/sanity_info
Since the sanity_info file has moved from the conf directory to the
cache directory, there is no longer any need to clean it away
explicitly in clean_esdk_builddir() since the whole cache directory is
already cleaned away anyway.
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
sanity.bbclass: Move sanity_info from conf to cache
Since this file is written during recipe parsing, having it in the
${BUILDDIR}/conf directory, which is covered by an inotify watcher,
will trigger a re-parse the next time bitbake is run and the resident
bitbake server is enabled. This causes the sanity_info file to be
updated again, which triggers a new parse the next time bitbake is run
ad infinitum. Moving it to ${BUILDDIR}/cache should avoid this.
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Barker [Tue, 7 Jan 2020 11:27:33 +0000 (11:27 +0000)]
archiver.bbclass: Add new mirror archiver mode
We define a new method of populating a source mirror using the archiver
bbclass instead of simply copying the contents of the downloads
directory. This allows the archiver features such as copyleft license
filtering and recipe type filtering to be used when preparing a source
mirror.
This new archiver mode is selected by setting `ARCHIVE_MODE[src]` to
'mirror'.
The source mirror mode can either be 'split' (default) or 'combined',
controlled by `ARCHIVER_MODE[mirror]`. Additionally, sources can be
excluded as needed by setting `ARCHIVER_MIRROR_EXCLUDE` to a list of
URI prefixes. These options are described in more detail in the new
entries in the header of archiver.bbclass.
New oeqa selftest cases are added to cover the mirror archiver mode.
Signed-off-by: Paul Barker <pbarker@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yi Zhao [Wed, 8 Jan 2020 06:51:17 +0000 (14:51 +0800)]
maintainers.inc: cleanup entries
Cleanup the following entries because these recipes have been removed
from oe-core:
cve-check-tool
gcc-cross-initial
gcc-crosssdk-initial
gcc-source-8.3.0
gccmakedep
gnome-themes-standard
gtk+
gtk-icon-utils-native
libnewt-python
mkfontdir
python3-file-utils
Signed-off-by: Yi Zhao <yi.zhao@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Barker [Wed, 8 Jan 2020 11:25:46 +0000 (11:25 +0000)]
wic: Add --include-path argument
This option adds the contents of the given path to a partition built
with the rootfs source plugin. The path is relative to the directory in
which wic is running not the rootfs itself so use of an absolute path
is recommended. This option is most useful when multiple copies of the
rootfs are added to an image and it is required to add extra content to
only one of these copies. This option only has an effect with the
rootfs source plugin.
Signed-off-by: Paul Barker <pbarker@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Barker [Wed, 8 Jan 2020 11:25:45 +0000 (11:25 +0000)]
wic: Preserve intermediate files
When debugging issues or trying to add new features in wic it's useful
to examine the intermediate files created by the tool such as images of
the individual partitions. Previously these files were created in a
subdirectory of IMGDEPLOYDIR and, deleted by wic itself and then the
working subdirectory completely removed at the end of IMAGE_CMD_wic.
Instead we should work in a staging directory named build-wic and leave
behind intermediate files for analysis if necessary.
This has the added benefit of allowing individual partition images to
be preserved by adding commands to IMAGE_CMD_wic_append() in an image
recipe if needed. This is useful where an update system requires exact
sized and pre-formatted partition images for deployment.
Signed-off-by: Paul Barker <pbarker@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Particularly remove a broken detection of skipped tests
(which was marking tests that actually passed), and install
additional artefacts needed for testing.
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
perl: install typemap and other extutils metadata as part of perl-core
Modules actually make use of these files, so they belong in perl-core
other than perl-doc (the immediate failure was ptests for
libmodule-build-perl failing).
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently if a ptest does not produce PASS or FAIL, but simply
errors out, this is not caught or reported; I think some ptests
may have silently regressed due to this.
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
dummy-sdk-package.inc: do multilib expanesion for RREPLACES as well
Otherwise, multilib packages such as lib64-perl aren't properly replaced:
Error:
Problem: package target-sdk-provides-dummy-1.0-r0.sdk_provides_dummy_target conflicts with lib64-perl provided by lib64-perl-5.30.1-r0.x86_64
- package target-sdk-provides-dummy-1.0-r0.sdk_provides_dummy_target conflicts with lib64-perl-module-strict provided by lib64-perl-5.30.1-r0.x86_64
- package target-sdk-provides-dummy-1.0-r0.sdk_provides_dummy_target conflicts with lib64-perl-module-warnings provided by lib64-perl-5.30.1-r0.x86_64
- package target-sdk-provides-dummy-1.0-r0.sdk_provides_dummy_target conflicts with lib64-perl-module-vars provided by lib64-perl-5.30.1-r0.x86_64
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
lib/oe/package_manager.py: put the sdk_provides_dummy_target upfront
Otherwise, the package utilizing this architecure does not get
properly prioritized by rpm/dnf. This isn't seen single-lib builds
(as the architecture ends up upfront anyway), but is triggered in
multulib where the architecture may end up in the middle of the list
due to multilib expansion/trimming duplicates.
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Stefano Babic [Fri, 3 Jan 2020 12:33:22 +0000 (13:33 +0100)]
libubootenv: Add libubootenv which replaces u-boot-fw-utils
libubootenv is a replacement for u-boot-fw-utils. It is
hardware-independent and provides fw_printenv and fw_setenv tools that
are full compatible with the ones provided by U-Boot. A library is
provided to access the environment from an own application.
License is LGPL-2.1 and this allow to link the library to proprietary
code. The user of the tools should install the configuration file
"fw_env.config", as he is already used to with u-boot-fw-utils. The
configuration file is compatible with u-boot-fw-utils.
A full discussion about issues on current u-boot-fw-utils can be read on
U-Boot's ML at:
Maxim Uvarov [Sat, 4 Jan 2020 20:08:57 +0000 (23:08 +0300)]
runqemu: add virtio block device
Implement new QB_DRIVE_TYPE="/dev/vdb" which falls back
to virtio block device emulated by qemu.
That is needed to support following qemu command line:
-drive id=disk0,file=%s,if=none,format=%s \
-device virtio-blk-device,drive=disk0'
Real use case is arm64 emulation with qemu with atf+optee+uboot,
so that uboot driver sees this disk and able to load from it.
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Mon, 6 Jan 2020 12:47:26 +0000 (12:47 +0000)]
glib: set missing cross property
GLib 2.60 onwards checks for /proc/self/cmdline instead of assuming that
it exists on Linux (because Cygwin also has it). As this is a runtime
test it doesn't execute in cross builds so defaults to false, but under
Linux this should be true.
Restore the previous behaviour and silence the warning by setting the
cross property appropriately.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Mon, 6 Jan 2020 10:08:46 +0000 (10:08 +0000)]
sstate: Avoid races over rewritten siginfo files
Try and avoid errors like:
Exception: bb.process.ExecutionError: Execution of 'build/tmp/work/x86_64-linux/libxslt-native/1.1.34-r0/temp/run.sstate_unpack_package.12343' failed with exit code 1:
touch: setting times of 'build/sstate_devtool/80/4d/sstate:libxslt-native::1.1.34:r0::3:804d33f3f8d032b01ae92207669dd0a8f95a84917d563d122a77df19e786d73c_populate_lic.tgz.siginfo': Stale file handle
WARNING: exit code 1 from a shell command.
by not overwriting existing files.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Mon, 6 Jan 2020 13:14:17 +0000 (13:14 +0000)]
scripts/gen-lockedsig-cache: Only look at new format sstate objects
We can have sstate directories which are mixtures of old and new layout entries.
Only use the new format ones since these are the only ones bitbake itself will
look at.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sun, 5 Jan 2020 14:49:03 +0000 (14:49 +0000)]
sstate: Handle sstate filenames longer than 255 characters
Many filesystems can't cope with filenames longer that 255 characters.
Add code to detect this and truncate non-essential elements of the filename
to stay within the limit.
[YOCTO #13268]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 11 Dec 2019 12:06:54 +0000 (12:06 +0000)]
sstate: Improve SSTATE_PKG handling
Move the task handling code into the SSTATE_PKGNAME variable using a temporary variable.
This makes the code more understandable as as well as allowing the length of the
final sstate filename to be more easily accesses for following patches.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 3 Jan 2020 18:03:35 +0000 (18:03 +0000)]
sstate: Add extra directory level
We're having speed issues on the autobuilder due to the numbers of files in sstate
directories. We previously split these by the first two characters of the hash.
This change extends this to split by the next two characters as well, creating
more layers of directories.
This should signifiantly speed up eSDK builds on the autobuilder as the current
sstate layout simply isn't scaling there but addresses a general complaint.
gen-lockedsig-cache needed to be updated for the new split level sstate.
Also update tests for new layout.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Khem Raj [Fri, 3 Jan 2020 07:21:22 +0000 (23:21 -0800)]
musl: Build outside of source tree
musl does not use cmake/autotools etc. but plane old makefile basded
build system, which means it does not get B defined to be outside S
therefore define B explicitly to be outside S
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Thu, 2 Jan 2020 16:37:58 +0000 (16:37 +0000)]
sstatesig: Avoid resetting taskhash within siggen for locked sigs
Since get_unihash uses taskhash as a key internally, changing it means
different bebahour when locked sigs are active verses not active. Under
corner cases this leads to a signature mismatch.
Avoid this by by adding a wrapper for the place its externally exposed
and then not changing the internals.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Mon, 30 Dec 2019 14:44:01 +0000 (14:44 +0000)]
sstatesig: Fix locked signature handling with unihashes
get_taskhash will call get_unihash internally in the parent class. We
need to disable our filter of it whilst this runs else incorrect hashes
can be calculated.
This is believed to be causing the locked signatures test to fail under
some circumstances (depending on whether earlier hashes are being
remapped).
[YOCTO #13605]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sat, 28 Dec 2019 00:02:46 +0000 (00:02 +0000)]
sstatesig: Test cross/native hashserv method extension
We can have one taskhash which represents multiple native/cross sstate objects
since they're stored by BUILD_ARCH or possibly host distro (or host gcc version).
We need to put these into separate namespaces on hashserv since their outhashes
will never match and we need deterministic lookups for the different namespaces.
Use this extramethod option to handle this. This fixes several problematic
failures on the autobuilder.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Kalle Lampila [Tue, 31 Dec 2019 15:10:49 +0000 (17:10 +0200)]
wic/filemap: If FIGETBSZ iotctl fail, failback to os.stat
Some file systems don't support fetching the block size (notably the
file system Docker uses for containers), so if iotctl() fail, try to use
failback via os.stat() to get block size.
Signed-off-by: Kalle lampila <kalle.lampila@lempea.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-Update: URLs changed from http to https Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Mon, 23 Dec 2019 01:18:37 +0000 (17:18 -0800)]
buildtools-extended-tarball: add recipe with build-essentials
* For some aging distros, such as CentOS 7, the native version
of gcc is simply too ancient and is a constant source of
headaches for moving forward.
* Add an extended version of buildtools-tarball which adds all
of build-essential, so that the host is now modernized and
capable of compiling the latest versions of components.
Fixes [YOCTO #13714]
Signed-off-by: Tim Orling <timothy.t.orling@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sat, 28 Dec 2019 23:37:37 +0000 (23:37 +0000)]
bitbake.conf: Simplify TARGET_SYS/SDK_SYS
As pointed out by Masahiro Yamada <masahiroy@kernel.org>, ('' or 'custom')
equates to "custom" and this code seems unecesaarily complicated and unused.
It was likely intended to supress the value if XXX_OS was set to '' but the
code doesn't actually do that. Simplify it as we shouldn't have this indirection
if we don't need it, its horrible.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Python uses features of glibc that require it to dynamically load (i.e.
dlopen()) libgcc_s at runtime. However, since this isn't a link time
dependency, it doesn't get picked up automatically by bitbake so
manually add it to RDEPENDS.
There is an outstanding bug in Python to make it explicitly link against
libgcc at link time which would remove the need for this. See:
https://bugs.python.org/issue37395
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Mon, 30 Dec 2019 20:59:50 +0000 (14:59 -0600)]
runqemu: Add multiconfig support
Users may want to run qemu against a specific multiconfig instead of the
base configuration, so give them the ability to specify which config
should be used with the MULTICONFIG environment variable.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Mon, 30 Dec 2019 20:59:29 +0000 (14:59 -0600)]
classes/image_types_wic: Depend on do_image_complete tasks
do_image_wic now depends on the do_image_complete tasks from all it's
dependencies. This allows images recipes specified in WKS_FILE_DEPENDS
(for example, if a machine wants to put a initrd into a boot partition)
to correctly trigger rebuilds for the wic image.
Note that recrdeptask can't be used as was done for do_deploy because
this would cause circular task dependency errors. This shouldn't be
necessary anyway if the dependencies specified in WKS_FILE_DEPENDS are
complete.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
André Draszik [Mon, 30 Dec 2019 20:36:55 +0000 (20:36 +0000)]
systemd: fix ${PN}-journal-remote systemd unit
Commit d31ff84a5f47 ("systemd: fix PN-container package splitting"), d0b2cedfb099 in poky, created a new package for the remote journal,
but introduced a typo during declaration of the systemd unit part
of that package.
Fix it.
Signed-off-by: André Draszik <git@andred.net> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
While installing perl-doc and tcl-doc to image, there is a file conflicts
at do_rootfs
...
|file /usr/share/man/man3/Thread.3 conflicts between attempted installs of
perl-doc-5.30.1-r0.core2_64 and tcl-doc-8.6.10-r0.core2_64
...
Use update-alternatives to fix
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Mon, 23 Dec 2019 01:18:36 +0000 (17:18 -0800)]
nativesdk-buildtools-perl-dummy: add dependencies for autoconf and automake
* For buildtools-extended-tarball, where we are adding all of build-essentials
to the nativesdk, we need additional perl modules for autoconf and automake.
Signed-off-by: Tim Orling <timothy.t.orling@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
