Jose Quaresma [Thu, 29 Oct 2020 14:27:59 +0000 (14:27 +0000)]
gstreamer1.0-vaapi: upgrade to version 1.18.0
* 0001-vaapsink-downgrade-to-marginal.patch
Removed since these are included in 1.18.0
https://gitlab.freedesktop.org/gstreamer/gstreamer-vaapi/-/commit/13d090b5660a2e2e2c997db44ba9a391101e8bd6
API and plugin documentation are no longer built with gtk_doc.
The gtk_doc documentation has been removed in favour of a new unified
documentation module built with hotdoc.
Distributors should use the documentation release tarball instead of trying
to package hotdoc and building the documentation from scratch.
Signed-off-by: Jose Quaresma <quaresma.jose@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jose Quaresma [Thu, 29 Oct 2020 14:27:47 +0000 (14:27 +0000)]
gstreamer1.0-plugins-base: upgrade to version 1.18.0
* 0001-meson-build-gir-even-when-cross-compiling-if-introsp.patch
* 0002-meson-Add-variables-for-gir-files.patch
Removed since these are included in 1.18.0
License-Update: remove file://common/coverage/coverage-report.pl
Upstream dont have it any more.
API and plugin documentation are no longer built with gtk_doc.
The gtk_doc documentation has been removed in favour of a new unified
documentation module built with hotdoc.
Distributors should use the documentation release tarball instead of trying
to package hotdoc and building the documentation from scratch.
Signed-off-by: Jose Quaresma <quaresma.jose@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
API and plugin documentation are no longer built with gtk_doc.
The gtk_doc documentation has been removed in favour of a new unified
documentation module built with hotdoc.
Distributors should use the documentation release tarball instead of trying
to package hotdoc and building the documentation from scratch.
Signed-off-by: Jose Quaresma <quaresma.jose@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Drop 0001-Do-not-hardcode-lib-as-location-for-site-packages-an.patch,
0001-configure.ac-fix-LIBPL.patch and 0001-python3-Do-not-hardcode-lib-for-distutils.patch
as they are all replaced by the new --platlibdir option to ./configure
Rename 0001-Lib-sysconfig.py-fix-another-place-where-lib-is-hard.patch to
0001-Lib-sysconfig.py-use-libdir-values-from-configuratio.patch
and describe the changes better.
License-Update: documentation now dual license under PSF & BSD
(not relevant for the recipe. Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-Update: copyright years Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Build tweaks:
- drop separate dri option and the patch that adds it; just setting -Ddri-drivers='' is enough
- strip leading/trailing commas from drivers' lists, as upstream no longer allows blank entries if the list
is not empty (e.g. -Ddri-drivers=,driver1,driver2 is now producing an error).
- drop the long-obsolete llvm 3.3/3.2 logic
License-Update: reformatted into rst Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Drop backports.
Drop 0001-rpmfc.c-do-not-run-file-classification-in-parallel.patch as
enabling compression support in libmagic finally allows us to use
parallel file classification.
Add a backported patch that fixes musl builds.
License-Update: formatting Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This allows re-enabling parallel file classification in rpm
in particular, as otherwise libmagic calls out to external
executables, which isn't implemented in a thread-safe way.
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yann E. MORIN [Wed, 28 Oct 2020 15:47:56 +0000 (16:47 +0100)]
recipes-core/busybox: fixup licensing information
Commit 7d32417b4d (busybox: Correct the name of the bzip2 license)
changes the licesne from 'bzip2' to 'bzip2-1.0.6' on the rationale
that the 'bzip2 license was renamed from "bzip2" to "bzip2-1.0.6"
[...] to match the official SPDX identifier.'
Though the above is true for the bzip2 and pbzip2 packages, the bzip2
code bundled in busybox is a copy from the bzip2 1.0.4 version, not the
1.0.6 version.
As such, using bzip2-1.0.6 is wrong.
Unfortunately, there is no official SPDX license identifier for this
bzip2 1.0.4 version, so we just mimick the existing ones (bzip2-1.0.5
and bzip2-1.0.6) by using bzip2-1.0.4.
Also, there is a license file attached to that, so we add it to the
list.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr> Cc: Peter Kjellerstedt <peter.kjellerstedt@axis.com> Cc: Richard Purdie <richard.purdie@linuxfoundation.org> Cc: Alexandre BELLONI <alexandre.belloni@bootlin.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yann E. MORIN [Wed, 28 Oct 2020 15:47:55 +0000 (16:47 +0100)]
common-licenses: add bzip2-1.0.4
The bzip2 license changes with each version; the changes are subtle, but
that makes it a different license everytime:
- copyright year
- authorship identification and address
- version of the release
- date of the release
Although we currently only have bzip2 and pbzip2 packages, we're going
to need this license for busybox, which uses code from bzip2-1.0.4.
Richard Leitner [Wed, 28 Oct 2020 09:43:21 +0000 (10:43 +0100)]
xcb-proto: update to 1.14.1
This fixes [1] the following build error for nativesdk on Fedora 33
which is caused by the removal of fractions.gcd() in favor of math.gcd()
in python 3.9 [2]:
ImportError: cannot import name 'gcd' from 'fractions' (/usr/lib64/python3.9/fractions.py)
Furthermore the SRC_URI is adapted as the old one doesn't provide the
1.14.1 archive.
Khem Raj [Tue, 27 Oct 2020 17:58:37 +0000 (10:58 -0700)]
musl: Update to latest master
Changes in this revision bump are
* 0b87551b (upstream/master) lift sigaction abort locking to fix posix_spawn child deadlock
* 99d5098a update crypt_blowfish to support $2b$ prefix
* 957c2763 remove unused weak definition of __tl_sync in membarrier.c
* 55767360 move aio implementation details to a proper internal header
* 37337660 remove long-unused struct __timer from pthread_impl.h
* 85e16aec drop use of pthread_once in timer_create
* 6ae2568b remove unused SIGTIMER handler in timer_create
* 47baa030 remove incorrect fflush from assert failure handler
* da845d52 fix getgrouplist when nscd reports an empty list
* b7bc9665 fix posix_spawn interaction with fork and abort by taking lock
* 25ea9f71 fix unintended observability of SIGABRT disposition reset via sigaction
* bd153422 implement _Fork and refactor fork using it
* e1e98d86 rename fork source file
* 50716702 ldso: use pthread_t rather than kernel tid to track ctor visitor
* 1efc8eb2 fix stale lock when allocation of ctor queue fails during dlopen
* 69a1b390 drop use of pthread_once in mutexattr kernel support tests
* b115bee4 fix missing synchronization of fork with abort
* 3cd3de61 move __abort_lock to its own file and drop pointless weak_alias trick
* 34904d83 fix fork of processes with active async io contexts
Jose Quaresma [Mon, 26 Oct 2020 11:39:04 +0000 (11:39 +0000)]
shaderc: add receipe
A collection of tools, libraries and tests for shader compilation
This receipe is needed to build the gstreamer vulkan plugin
as it provides the binary glslc.
It is based on arch linux shaderc packge including the patches
https://github.com/archlinux/svntogit-packages/blob/91f0fa6ee3a220264a448527c1a8cf037caaad8e/trunk/PKGBUILD
Jose Quaresma [Mon, 26 Oct 2020 11:39:03 +0000 (11:39 +0000)]
glslang: add receipe
Glslang is the official reference compiler front end for the
OpenGL ES and OpenGL shading languages. It implements a strict interpretation
of the specifications for these languages. It is open and free for anyone to use,
either from a command line or programmatically.
This receipe is needed to build the new shaderc
Signed-off-by: Jose Quaresma <quaresma.jose@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Randy Witt [Mon, 19 Oct 2020 21:42:14 +0000 (14:42 -0700)]
rt-tests: Update recipes to use 1.8
John Kacur(the maintainer of rt-tests), recommends "that they use
unstable/devel/latest as the maintained stable branch". I received this
information on IRC, I couldn't find it documented anywhere else.
This means that both rt-tests and hwlatdetect can be updated to be 1.8,
the latest release on unstable/devel/latest.
0001-gzip-with-n-for-build-reproducibilty.patch was removed because
upstream now uses "gzip -n" by default.
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Randy Witt [Mon, 19 Oct 2020 21:42:13 +0000 (14:42 -0700)]
numactl: Skip the ptests when numa is not supported
When numa is not supported, there is no reason to run the ptests since
they will fail. While the best scenario would be for the upstream tests
to skip themselves when numa is not supported, for now skip them in
run-ptest.
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Randy Witt [Mon, 19 Oct 2020 21:42:11 +0000 (14:42 -0700)]
numactl: Add the recipe for numactl
This is a direct copy of numactl from meta-openembedded. numactl is
being moved to oe-core since the latest versions of rt-tests(which is in
oe-core) require libnuma.
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Thu, 29 Oct 2020 15:21:35 +0000 (15:21 +0000)]
scripts/buildhistory_analysis: Avoid tracebacks from file comparision code
We're seeing tracebacks from buildhistory analysing the python 3.8 -> 3.9
upgrade due to the significant file renames. Avoid these by checking before
removal as they can happen multiple times.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 29 Jan 2020 17:28:46 +0000 (17:28 +0000)]
sstatesig: Log timestamps for hashequiv in reprodubile builds for do_package
Currently if a task generates the same output with different timestamps,
hasequiv won't detect it but reproducibile builds will fail tests due
to the different timestamps.
Add do_package timestamps to the hash when reproducibile builds are enabled
to avoid this.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mingli Yu [Tue, 27 Oct 2020 01:29:39 +0000 (09:29 +0800)]
bitbake.conf: Exclude ${CCACHE_DIR} from pseudo database
When ccache enabled as below:
INHERIT += "ccache"
CCACHE_DIR = "/path/build/ccache"
There comes do_install failure for some recipes randomly, take
linux-libc-headerswhen as example.
$ cat /path/build/tmp/work/corei7-64-wrs-linux/linux-libc-headers/5.8-r0/pseudo/pseudo.log
[snip]
path mismatch [1 link]: ino 243004209 db '/path/build/ccache/6/stats' req '/path/build/ccache/7/stats.lock'.
[snip]
Exclude ${CCACHE_DIR} from pseudo database to fix the potential
do_install failure.
Signed-off-by: Mingli Yu <mingli.yu@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Mon, 26 Oct 2020 18:01:06 +0000 (18:01 +0000)]
glib-2.0: fix parsing of slim encoded tzdata
As of tzcode 2020b the timezone data is encoded using the 'slim' format
instead of the previous 'fat'. This exposes a number of bugs in GLib,
so backport the fixes to improve the parser.
[ YOCTO #14106 ]
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In glibc 2.32, the POSIX thread functions are in both
libc and libpthread, causing valgrind to misinterpert
test behaviour. This patch tells valgrind to intercept
both the libc and the libpthread functions, resolving
these ptest failures.
Here are the latest test results on qemux86-64:
=== Test Summary ===
TOTAL: 728
PASSED: 688
FAILED: 1
SKIPPED: 39
The failed test is in helgrind:
FAIL: helgrind/tests/pth_destroy_cond
This change contains races as it will start poking into do_package task
directories from do_populate_sysroot. If we want to do this for native
recipes, we need to add guards around the package code and only make
this happen for native in populate_sysroot, not target in
populate_sysroot too. Backtrace from an example problem below:
ERROR: openssl-1.1.1g-r0 do_populate_sysroot: Error executing a python function in exec_python_func() autogenerated:
The stack trace of python calls that resulted in this exception/failure was:
File: 'exec_python_func() autogenerated', lineno: 2, function: <module>
0001:
*** 0002:buildhistory_emit_pkghistory(d)
0003:
File: '/home/ross/Yocto/poky/meta/classes/buildhistory.bbclass', lineno: 319, function: buildhistory_emit_pkghistory
0315:
0316: write_pkghistory(pkginfo, d)
0317:
0318: # Create files-in-<package-name>.txt files containing a list of files of each recipe's package
*** 0319: bb.build.exec_func("buildhistory_list_pkg_files", d)
0320:}
0321:
0322:python buildhistory_emit_outputsigs() {
0323: if not "task" in (d.getVar('BUILDHISTORY_FEATURES') or "").split():
File: '/home/ross/Yocto/poky/bitbake/lib/bb/build.py', lineno: 256, function: exec_func
0252: with bb.utils.fileslocked(lockfiles):
0253: if ispython:
0254: exec_func_python(func, d, runfile, cwd=adir)
0255: else:
*** 0256: exec_func_shell(func, d, runfile, cwd=adir)
0257:
0258: try:
0259: curcwd = os.getcwd()
0260: except:
File: '/home/ross/Yocto/poky/bitbake/lib/bb/build.py', lineno: 503, function: exec_func_shell
0499: with open(fifopath, 'r+b', buffering=0) as fifo:
0500: try:
0501: bb.debug(2, "Executing shell function %s" % func)
0502: with open(os.devnull, 'r+') as stdin, logfile:
*** 0503: bb.process.run(cmd, shell=False, stdin=stdin, log=logfile, extrafiles=[(fifo,readfifo)])
0504: except bb.process.ExecutionError as exe:
0505: # Find the backtrace that the shell trap generated
0506: backtrace_marker_regex = re.compile(r"WARNING: Backtrace \(BB generated script\)")
0507: stdout_lines = (exe.stdout or "").split("\n")
File: '/home/ross/Yocto/poky/bitbake/lib/bb/process.py', lineno: 184, function: run
0180: if not stderr is None:
0181: stderr = stderr.decode("utf-8")
0182:
0183: if pipe.returncode != 0:
*** 0184: raise ExecutionError(cmd, pipe.returncode, stdout, stderr)
0185: return stdout, stderr
Exception: bb.process.ExecutionError: Execution of '/yocto/ross/build/tmp/work/neoversen1-poky-linux/openssl/1.1.1g-r0/temp/run.buildhistory_list_pkg_files.4158804' failed with exit code 2:
/yocto/ross/build/tmp/work/neoversen1-poky-linux/openssl/1.1.1g-r0/temp/run.buildhistory_list_pkg_files.4158804: 183: cd: can't cd to /yocto/ross/build/tmp/work/neoversen1-poky-linux/openssl/1.1.1g-r0/packages-split/openssl-engines
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
imagefeatures: New test case, test_empty_image, added
An empty image build file exists under the meta-selftest folder, test-empty-image.bb,
which builds an image with no additional packages. However, there were no further
selftest created to verify its emptiness
This change consists of the selftest related to the said image to verify its
emptiness and the 'import glob' moved as global import instead of local import.
The expected outcome of the test should be TRUE or 1 if the .manifest file content
is empty.
[YOCTO #8455]
Signed-off-by: Khairul Rohaizzat Jamaluddin <khairul.rohaizzat.jamaluddin@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mingli Yu [Thu, 22 Oct 2020 03:27:25 +0000 (11:27 +0800)]
tcl: adapt to potential pseudo changes
It failed as below when rerun do_install for tcl:
$ bitbake tcl
$ bitbake tcl -f -cinstall
[snip]
| NOTE: make -j 72 DESTDIR=/build/tmp/work/core2-64-wrs-linux/tcl/8.6.10-r0/image install
| abort()ing pseudo client by server request. See https://wiki.yoctoproject.org/wiki/Pseudo_Abort for more details on this.
| /build/tmp/work/core2-64-wrs-linux/tcl/8.6.10-r0/temp/run.do_install.3490694: line 184: 3499214 Aborted (core dumped) make -j 72 "$@"
By default the S is ${WORKDIR}/${BPN}-${PV}, but after unpack,
the tcl source [1] unpacked to ${WORKDIR}/${BPN}${PV} and all the
files under ${WORKDIR}/${BPN}${PV} are acutally the source files.
But the the main Makefile.in is under ${WORKDIR}/${BPN}${PV}/unix
for tcl, so there is below logic in tcl recipe:
S = "${WORKDIR}/${BPN}${PV}/unix"
To adapt the potential pseudo changes, there is a general logic to
exclude ${S} from pseudo database in base.bbclass in [2]. That's to
say, just the dir ${WORKDIR}/${BPN}${PV}/unix is excluded from the
pseudo database.
But it's not enough for tcl, we need to exclude the actual source dir
${WORKDIR}/${BPN}${PV} from pseudo database specifically to fix the
above do_install failure.
Richard Purdie [Mon, 19 Oct 2020 12:50:19 +0000 (13:50 +0100)]
oeqa: Add sync call to command execution
We previously put a sync call into devtool to try and combat the bitbake
timeout issues on the autobuilder. It isn't enough as the timeouts occur
mid test. They are also occurring on non-devtool tests.
Add in sync calls around command execution instead.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Victor Kamensky [Mon, 19 Oct 2020 22:21:46 +0000 (15:21 -0700)]
qemu: change TLBs number to 64 in 34Kf mips cpu model
Replace OE private qemu patch with one that got upstreamed
and solves the same problem: increase qemumips CI performance
by increasing number of TLBs in CPU model and reduce need to
run software TLB refill code.
Signed-off-by: Victor Kamensky <kamensky@cisco.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
