Khem Raj [Fri, 18 Mar 2022 01:03:25 +0000 (18:03 -0700)]
qemuriscv: Use virtio-tablet-pci for mouse
This helps in making mouse response better where transition between host
and guest mouse is abrupt and not precise and as a result its difficult
to access stuff near the edges.
Trevor Gamblin [Thu, 17 Mar 2022 19:22:27 +0000 (15:22 -0400)]
iptables: use nft backend with libnftnl PACKAGECONFIG
Currently, when the libnftnl (part of meta-networking) PACKAGECONFIG is
enabled for iptables, both legacy and nft-based binaries are built and
installed in the image. However, the "iptables" symlink in this case
still points to xtables-legacy-multi, rather than xtables-nft-multi.
This patch adds a conditional check to replace the symlink to point to
the latter if iptables is built with libnftnl support, which is
consistent with other major distros (e.g. Fedora). The "iptables-legacy"
symlink remains present and unmodified.
Richard Purdie [Thu, 17 Mar 2022 21:05:15 +0000 (21:05 +0000)]
oeqa/selftest/tinfoil: Improve tinfoil event test debugging
We still see occasional test failures for unknown reasons. Add some debugging to
show whether the matching files event was received even if the command complete wasn't.
Also ensure any commandfailed/commandexit event is shown.
This will hopefully aid debugging the next time the issue occurs.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Thu, 17 Mar 2022 17:12:08 +0000 (17:12 +0000)]
python3-sphinx: Work around reproducibility issue
The dict used in this recipe doesn't write into the cache consistently, work
around it until this issue is resolved in python upstream (similar to frozenset
issues).
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Neill [Wed, 16 Mar 2022 10:37:24 +0000 (10:37 +0000)]
systemd: Update 250.3 -> 250.4
The following security and bug-fix patches are included as part of the 250.4
update:
c6603da3ad boot: Properly check status code of console_key_read 2198c08d07 core: really skip automatic restart when a JOB_STOP job is pending 367041af81 pid1: set SYSTEMD_NSS_DYNAMIC_BYPASS=1 env var for dbus-daemon 160eeab224 virt: Fix Xen Dom0 detection logic to no longer report as VM 514a4c051c network: bridge: fix endian of vlan protocol 4dbc210124 resolve: fix possible memleak d82bd80cf4 resolve: fix potential memleak and use-after-free dcba78244e util: another set of CVE-2021-4034 assert()s 74dfb51f70 sd-dhcp6-client: fix sending prefix delegation request during rebind df59c65a23 mkdir: allow to create directory whose path contains symlink ae95ca27be sd-dhcp-lease: fix memleak 2b04d3b3fc sd-dhcp-lease: fix reading unaligned memory 1ef56ad928 network: xfrm: refuse zero interface ID 7dc0f80588 sd-dhcp-lease: fix a memory leak in dhcp_lease_parse_search_domains 426807c54b sd-dhcp-lease: fix an infinite loop found by the fuzzer 0456e3aaaa oomd: fix race with path unavailability when killing cgroups
Richard Purdie [Thu, 17 Mar 2022 00:32:50 +0000 (00:32 +0000)]
python3-cryptography: Fix ptest result handling
We run the tests correctly, we just don't get individual test result
reporting. Borrow the code from python3-bcrypt to handle this correctly
and ensure the test pass/fail scores and individual test results are
reported in a format we can read.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 16 Mar 2022 18:32:41 +0000 (18:32 +0000)]
classes/python_pep517: add more comments
Remove mention of prebuilt wheels, this is for the full PEP517 build
process and recipes that want to install prebuilt wheels can use
pypa/installer directly.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 16 Mar 2022 18:32:37 +0000 (18:32 +0000)]
classes/python_pep517: implement a standard do_compile
As all PEP517-compliant build systems have a universal API, we can ask
that users of this class set PEP517_BUILD_API to the class that implements
this API and call it ourselves, instead of users needing to implement
near-identical do_compile tasks themselves.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joe Slater [Tue, 15 Mar 2022 22:45:37 +0000 (15:45 -0700)]
expect: modify fixline1 script
fixline1 needs to be processed by fixline1, but we
short-circuit that. Enter the multi-line shebang which
I am a bit queasy about, but expect loves it.
Signed-off-by: Joe Slater <joe.slater@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wangmy [Wed, 16 Mar 2022 00:17:57 +0000 (08:17 +0800)]
dpkg: upgrade 1.21.1 -> 1.21.2
The following patches refreshed for new version:
0001-build.c-ignore-return-of-1-from-tar-cf.patch
0003-Our-pre-postinsts-expect-D-to-be-set-when-running-in.patch
0004-The-lutimes-function-doesn-t-work-properly-for-all-s.patch
0007-dpkg-deb-build.c-Remove-usage-of-clamp-mtime-in-tar.patch
noman.patch
remove-tar-no-timestamp.patch
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 15 Mar 2022 17:34:56 +0000 (17:34 +0000)]
license: Drop AVAILABLE_LICENSES
This variable is a performance liability and is highly dependent on which
layers are added to the configuration which can cause signature issues
for users. We have no users left in OE-Coreso remove it.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 15 Mar 2022 15:04:05 +0000 (15:04 +0000)]
create-spdx: Use function rather than AVAILABLE_LICENSES
We can directly call the function rather than using the variable indirection.
As this is the last user of the variable, it then allows removal of it in a
followup patch.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 15 Mar 2022 17:29:13 +0000 (17:29 +0000)]
perf-build-test/report: Drop phantomjs and html email reports support
phantomjs isn't reliable and we've moved to sharing the reports via a webserver.
Update the scripts to more match those being used in the autobuilder helper
where the html email support was removed.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ahmad Fatoum [Tue, 15 Mar 2022 17:23:53 +0000 (18:23 +0100)]
kernel-fitimage: allow overriding FIT configuration prefix
The prefix was recently changed to play ball with the CVE-2021-27138
workaround in U-Boot. Already deployed bootloaders though may still
expect a configuration name containing @ or they may not be affected
by this issue.
Also, uses may want to customize it beyond the [@-] issue: When
device trees are built from a recipe using kernel-devicetree.bbclass
the configuration nodes will contain the parent directories, e.g.
will become "conf-freescale_imx8mm-myboard.dtb". When moving to
devicetree.bbclass, this prefix go away. With this change here,
users can restore it easily by setting FIT_CONF_PREFIX =
"conf-freescale_", ensuring backwards compatibility.
Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Tue, 15 Mar 2022 15:37:21 +0000 (15:37 +0000)]
python3-setuptools-rust-native: use setuptools_build_meta
The wheel generation problems are solved now: a wheel is correctly built
to setuptools_rust-1.1.2-py3-none-any.whl, so we can use
setuptools_build_meta.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Tue, 15 Mar 2022 16:23:43 +0000 (16:23 +0000)]
devupstream: fix handling of SRC_URI
As the class handler runs before overrides are actually applied we need
to check both SRC_URI:class-devupstream and SRC_URI, otherwise the
automatic assignment of S="${WORKDIR}/git" for git repositories does
not work if the base recipe uses http: but :class-devupstream has a git://
entry instead.
Also, there's no need to set S:class-devupstream, we can just assign to
S here.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wangmy [Tue, 15 Mar 2022 11:22:19 +0000 (19:22 +0800)]
libuv: upgrade 1.43.0 -> 1.44.1
License-Update:
the following message removed from LICENSE:
android-ifaddrs.h, android-ifaddrs.c, copyright Berkeley Software Design
Inc, Kenneth MacKay and Emergya (Cloud4all, FP7/2007-2013, grant agreement
n° 289016). Three clause BSD license.
Changes since version 1.43.0:
=============================
* process: simplify uv__write_int calls
* macos: don't use thread-unsafe strtok()
* process: fix hang after NOTE_EXIT
* darwin: remove EPROTOTYPE error workaround
* doc: fix v1.43.0 changelog entries
* win: replace CRITICAL_SECTION+Semaphore with SRWLock
* darwin: translate EPROTOTYPE to ECONNRESET
* android: use libc getifaddrs()
* unix: fix STATIC_ASSERT to check what it means to check
* unix: ensure struct msghdr is zeroed in recvmmsg
* test: test with maximum recvmmsg buffer
* unix: don't allow too small thread stack size
* bsd: ensure mutex is initialized
* doc: add gengjiawen as maintainer
* process: monitor for exit with kqueue on BSDs
* test: fix flaky uv_fs_lutime test
* build: fix cmake install locations
* thread,win: fix C90 style nit
* build: rename CFLAGS to AM_CFLAGS
* doc/guide: update content and sample code
* process,bsd: handle kevent NOTE_EXIT failure
* test: remove flaky test ipc_closed_handle
* darwin: bump minimum supported version to 10.15
* win: return fractional seconds in uv_uptime()
* build: export uv_a for cmake
* loop: add pending work to loop-alive check
* win: use GetTickCount64 for uptime again
* win: restrict system DLL load paths
* win,errors: remap ERROR_ACCESS_DENIED to UV_EACCES
* bench: add 'uv_queue_work' ping-pong measurement
* build: fix error C4146 on MSVC
* test: fix benchmark-ping-udp
* win,fs: consider broken pipe error a normal EOF
* document the values of enum uv_stdio_flags
* win,loop: add missing uv_update_time
* win,fs: avoid closing an invalid handle
* fix oopsie from
* doc: clarify android api level
* win: fix style nits [NFC]
* test: fix flaky udp_mmsg test
* test: fix ipc_send_recv_pipe flakiness
* doc: checkout -> check out
* core: change uv_get_password uid/gid to unsigned
* hurd: unbreak build on GNU/Hurd
* freebsd: use copy_file_range() in uv_fs_sendfile()
* test: use closefd in runner-unix.c
* Reland "macos: use posix_spawn instead of fork"
* android: fix build error when no ifaddrs.h
* unix,win: add uv_available_parallelism()
* process: remove OpenBSD from kevent list
* zos: fix build breakage
* process: only use F_DUPFD_CLOEXEC if it is defined
* win,poll: add the MSAFD GUID for AF_UNIX
* unix: simplify uv__cloexec_fcntl()
* doc: add secondary GPG ID for vtjnash
* unix: remove uv__cloexec_ioctl()
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wangmy [Tue, 15 Mar 2022 11:22:16 +0000 (19:22 +0800)]
libcgroup: upgrade 2.0 -> 2.0.1
module.patch
musl-decls-compat.patch
removed since they're included in 2.0.1
Changelog:
==========
- Allow reading/writing from cgroup.* files in cgroup v2
- Add support for cgroup v2's cgroup.threads file
- Fix issue where libcgroup/pam wasn't working properly when
cgrulesengd is disabled
- Fix a bug where the cgroup version wasn't initialized in a
named cgroup v1 hierarchy
- Various automake bug fixes
- Build PAM module as unversioned DSO
- Fix build issues with musl libc
- Fix potential TOCTOU race in cgroup_get_procs()
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wangmy [Tue, 15 Mar 2022 12:12:29 +0000 (20:12 +0800)]
lttng-modules: upgrade 2.13.1 -> 2.13.2
The following patches removed since they're included in 2.13.2
0001-fix-block-remove-GENHD_FL_SUPPRESS_PARTITION_INFO-v5.patch
0002-fix-block-remove-the-rq_disk-field-in-struct-request.patch
0003-fix-mm-compaction-fix-the-migration-stats-in-trace_m.patch
0004-fix-btrfs-pass-fs_info-to-trace_btrfs_transaction_co.patch
0005-fix-random-rather-than-entropy_store-abstraction-use.patch
0006-fix-net-skb-introduce-kfree_skb_reason-v5.17.patch
0007-fix-net-socket-rename-SKB_DROP_REASON_SOCKET_FILTER-.patch
Changelog:
=========
* Fix: incorrect in/out direction for syscall exit
* fix: net: socket: rename SKB_DROP_REASON_SOCKET_FILTER (v5.17)
* fix: net: skb: introduce kfree_skb_reason() (v5.17)
* fix: random: rather than entropy_store abstraction, use global (v5.17)
* fix: btrfs: pass fs_info to trace_btrfs_transaction_commit (v5.17)
* fix: mm: compaction: fix the migration stats in trace_mm_compaction_migratepages() (v5.17)
* fix: block: remove the ->rq_disk field in struct request (v5.17)
* fix: block: remove GENHD_FL_SUPPRESS_PARTITION_INFO (v5.17)
* Copyright ownership transfer
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wangmy [Tue, 15 Mar 2022 11:22:17 +0000 (19:22 +0800)]
libdnf: upgrade 0.65.0 -> 0.66.0
0001-Add-WITH_TESTS-option.patch
0001-Look-fo-sphinx-only-if-documentation-is-actually-ena.patch
removed since they're included in 0.66.0
0001-Get-parameters-for-both-libsolv-and-libsolvext-libdn.patch
0004-Set-libsolv-variables-with-pkg-config-cmake-s-own-mo.patch
modified for new version
Changelog:
=========
Use rpmdbCookie from librpm, remove hawkey.Sack._rpmdb_version
Fix handling transaction id in resolveTransactionItemReason (RhBug:2010259,2053014)
Remove deprecated assertions (RhBug:2027383)
Skip rich deps for autodetection of unmet dependencies (RhBug:2033130, 2048394)
Increase required rpm version since we use rpmdbCookie()
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wangmy [Tue, 15 Mar 2022 10:51:04 +0000 (18:51 +0800)]
harfbuzz: upgrade 4.0.0 -> 4.0.1
Changelog:
=========
- Update OpenType to AAT mappings for hist and vrtr features.
- Update IANA Language Subtag Registry to 2022-03-02.
- Update USE shaper to allow any non-numeric tail in a symbol cluster, and remove
obsolete data overrides.
- Fix handling of baseline variations to return correctly scaled values.
- A new experimental hb_subset_repack_or_fail() to repack an array of objects,
eliminating offset overflows. The API is not available unless HarfBuzz is
built with experimental APIs enabled.
- New experimental API
+hb_link_t
+hb_object_t
+hb_subset_repack_or_fail()
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Mon, 14 Mar 2022 14:39:23 +0000 (14:39 +0000)]
python_pep517: use installer instead of pip
Instead of battling pip to install a wheel, use installer. Installer
does one thing, so it's faster and easier to work with.
This means setuptools, pip, and wheel are no longer part of the
bootstrap phase, so they can be built normally. To avoid sysroot file
conflicts these three recipes can't install .pyc files to the native
sysroot.
We currently patch pypa/installer to allow us to override the interpreter
used, which means we can drop the interpreter seding.
We don't need to recompile any Python which is found in $bindir as
Python doesn't actually load those files.
Across a build of oe-core, the only differences between using pip and
installer are:
- the .dist-info/RECORD files are ordered differently
- the .dist-info/REQUESTED and INSTALLER files are not created
- the hashbang in native scripts is "/usr/bin/env nativepython" instead
of pointing directly at the native sysroot python3.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
projects / thirdparty / openembedded / openembedded-core-contrib.git / log
