Ensure the fallback value if used is written to the SDE file
and hence stored in sstate, reducing any confusion within the
code over '0' values.
Bump the HASHEQUIV_VERSION since we've had a ton of trouble
with ensuring this rolls out correctly on the autobuilder so
others may too, take a clean slate for it.
(From OE-Core rev: 53ffffa32b76330835287dfc05d4e4796841af08)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This makes sure that we can compile glibc for powerp9 based machines
irrespective of endianness or bitness
(From OE-Core rev: 6201018650fe8966b99860bbda24c5903e7d7a60)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
musl is configured to use /lib and /usr/lib for ppc64 unlike glibc where
it is expected to have libs in /lib64 and /usr/lib64
(From OE-Core rev: 998d15a86282d9cf57fbd0a4eb76a56b9938a348)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If we handle the or-later licences separately (which we should),
we need to add in the missing name mappings for the code to
function correctly.
(From OE-Core rev: ccc785f9d428b279c8aa9a1ca00dc685342fdb1e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* Updated mappings of license fields of meta/conf/licenses.conf to match
latest SPDX naming.
* Add mappings to the old names
* Renamed license files to match the new preferred names.
* Added "or later" versions of license mappings
* Added "or later" versions of common license files eg GPL-2.0-or-later
Fixes: [YOCTO #13320]
(From OE-Core rev: 5ecf139a31fa7bd813855f1235ea9f434fbcb2e0)
Signed-off-by: Meh Mbeh Ida Delphine <idadelm@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Rather than trying to set the umask of every task to the correct
value for determinism, set one value globally. This uses a new
bitbake variable so bump the minimum version to match.
This fixes strange determinism issues in at least quilt-ptest,
valgrind-ptest and kernel-devsrc.
(From OE-Core rev: b07b7deeae3d519d9998d583592f3e4f8f2802b3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Adding -O can be troublesome in some packages where it may override the
O<n> specified by CFLAGS, this can be due to configure processing of
CFLAGS and munging them into new values in Makefiles, which is
contructed from CC and CFLAGS passed by bitbake environment. Problem
arises if the sequence is altered, which seems to be the case in some
packages e.g. ncurses, where the value from CC variable is added last
and thus overrides -O<n> coming from CFLAGS,
Therefore grok the value from SELECTED_OPTIMIZATION and append the
appropriate -O<level> flag to lcl_maybe_fortify so the level does not
change inaderdantly.
Since we do not use -O0 anymore there is no point of checking for
DEBUG_BUILD since it uses -Og now which works fine with
-D_FORTIFY_SOURCE=2, so check for optlevel O0 instead
(From OE-Core rev: 9571a18f7d15b3bffafc2e277ab90a21d6763697)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Cc: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
compiler can only use fortify options when some level of optimization is
on, otherwise it ends up sending some warnings.
warning: _FORTIFY_SOURCE requires compiling with optimization (-O) [-W#warnings]
this is usually OK, since -O<level> would be added via CFLAGS to
compiler cmdline in normal compile stages, however during configure
there are problems when CC,CPP,CXX are probed alone in configure tests
which results in above warning, which confuses the configure results and
autotools 2.70+ detects it as error e.g.
configure:17292: error: C preprocessor "riscv32-yoe-linux-clang -target riscv32-yoe-linux -mlittle-endian -mno-relax -Qunused-arguments -fstack-protector-strong -D_FORTIFY_SOURCE=2 -Wformat -Wformat-security -Werror=format-security --sysroot=/mnt/b/yoe/master/build/tmp/work/riscv32-yoe-linux/ndpi/3.4-r0/recipe-sysroot -E" fails sanity check
See `config.log' for more details
therefore adding a -O ( which actually is -O1 ) to lcl_maybe_fortify
means we can properly test these configure tests and real -O<level> will
still override -O added here, so overrall behavior improves
(From OE-Core rev: b6113dd68caa46d56cf3c8293119f2b9d8b137fd)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Drop backported patches which are already present in 2.36 release
(From OE-Core rev: 897afa95ba340f1124decac5753e1d1e1283b515)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This looks like it was from before the recipe was split, we'd expect
the system qemu mode for running the images so the dependency should be
updated.
(From OE-Core rev: 3a4fed4ae0e8a0d1bd62ea5fa1ef12925e1f20f5)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We never recommend setting RDEPENDS or RPROVIDES without a package name
against them. The default in bitbake.conf is legacy only, drop it.
The python recipe was trying to add to the empty variable in the native case
fix that too.
(From OE-Core rev: b8bbc1bbe282cce2ea4d0ff293f931f6caf6153b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This reverts commit f5b29367af4d8e5daea5771264774aa49519f9a8.
Will use ccache-native which is more reliable.
(From OE-Core rev: 1b659623430e1a6e6dd266e65bab7ca8155a7138)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This involves some pretty major changes for qemu. In particular, they
switched to meson+ninja so we have to adapt to that.
Patch changes:
* CVE patches - dropped as backports
* cflags fix - upstream code changed significantly, need new patch if still issues
* mips TLB entries - dropped as merged upstream
* usb fix - dropped as merged upstream
* find_datadir - dropped as code no longer present that I could find
A patch was added to allow us to force the configure script into "cross" mode
without setting cross_prefix which has other effects we don't need/want.
Dependencies on meson/ninja were added.
Specifying the python interpreter causes the internal meson copy to be
built/used which is undesireable for us so don't do that. The correct
python is in PATH anyway.
Acked-by: Alistair Francis <alistair.francis@wdc.com>
(From OE-Core rev: 181c635567aafb9b4787d8d6d0bcd4a615ceae80)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Builds were failing on WSL2 which turns out to be due to accesses to
/run/ on those systems. Add this to PSEUDO_IGNORE_PATHS to fix
builds on WSL2.
[YOCTO #14175]
(From OE-Core rev: 1d1bf51217e8b4d54af28739d3271484ee5a7974)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is about to be needed by the psplash recipe and is a standard part
of coreutils so usually present on hosts.
(From OE-Core rev: 99206612a58e6b69718910e0806d9d36997b2fba)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
On kernel's v5.11+, we get the following warning after boot:
[ 47.287826] vmap allocation for size 3149824 failed: use vmalloc=<size> to increase size
It can be fixed by passing the vmalloc page size explicitly for our ARM
machines.
(From OE-Core rev: 5c6064e1a38f4dbb1ab3fa04adb1c772817826af)
Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As the reference kernel is going to v5.10, we also move the libc headers
to match. This is also the latest LTS kernel, so is appropriate for a
libc-headers version bump.
This has been tested against all supported architectures for both glibc
and musl.
(From OE-Core rev: 1f1412c9a52c4fa3ded4f87a0c7ffdc4a243d632)
Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
ccache 4.x has hard dependencies on cmake-native (used as
build system) and zstd, which means inserting ccache-native
as DEPENDS into everything creates circular dependencies which
are impossible to break.
ccache 3.x did not have this problem as it used plain makefiles
and an in-tree copy of zlib.
(From OE-Core rev: f5b29367af4d8e5daea5771264774aa49519f9a8)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Zstd is becoming increasingly popular as the modern compression
standard, and there are several recipes in core now that have an option
for it.
(From OE-Core rev: b132e9e8647c74fba439c689ec1409993e8590ed)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently there's a limitation that Extensible SDKs can only be build on a
machine with the same architecture as the SDK.
However, SDKMACHINE defaults to x86-64, which means eSDKs cannot be built
out of the box on arm64 systems.
By defaulting SDKMACHINE to ${BUILD_ARCH} we generate SDKs for the build
host, so they'll always work out of the box.
(From OE-Core rev: 539cd4a5fbb9e4507d90bfb1165fcb06907b46d8)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This also effectively reverts commit b6d30c21b0:
bitbake.conf: Extend PSEUDO_IGNORE_PATHS to ${COREBASE}/meta
The contents of ${COREBASE}/meta were ignored as pyc files could be
generated for the contents of the lib subdirectory if python modules
were imported within a pseudo context. However this doesn't protect us
from pyc files being generated in the lib directories for other layers.
It's far better to tell python not to produce pyc files when running
under pseudo (by setting the PYTHONDONTWRITEBYTECODE variable) as this
will cover any location where pyc files could possibly be created. This
variable is set in FAKEROOTBASEENV so that it applies to the
bitbake-worker instance for fakeroot tasks, preventing pyc files from
being generated for imports in python tasks themselves.
Also add a test case to ensure that pyc files are not created in tasks
which are executed under pseudo.
(From OE-Core rev: 73d538f20743017a44cea4c20dbe09a0327cfc71)
Signed-off-by: Paul Barker <pbarker@konsulko.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Use oe.path.canonicalize() to canonicalize the paths in
PSEUDO_IGNORE_PATHS before passing them to pseudo. This is needed since
pseudo will compare them to paths that are canonicalized.
(From OE-Core rev: 1f2a497daa09cf3459e27ad6e0e8513938b52c79)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
qemuriscv enables hvc0 along with ttyS0, however its not enabled in
/proc/consoles, getty tries to enable it in inittab and erroring out
Fixes below message with sysvinit
INIT: Id "hvc0" respawning too fast: disabled for 5 minutes
(From OE-Core rev: 8a6559f1561ca6b7719bb46fc446db46d8086ea3)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Refactor EFI_BOOT_IMAGE so that the EFI name for the architecture
is exposed as EFI_ARCH, so that other recipes (such as bootloaders)
can reuse it.
(From OE-Core rev: d3bf2d0fbd50309d7cf8f066c258ba1ebdb05f47)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This reverts commit cc30537bb4d0815c4d69f0b1a1a8a200085902d5.
There are issues with this change where the build directory may be included
within the paths included in BBLAYERS, which is particularly problematic
where people are using kas.
An alternative to this to handle the pyc file issue is being worked on but
revert this for now whilst that fix is developed.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The OCTEON TX2 core is custom designed by Marvell and is compliant
with Arm V8.2 architecture.
(From OE-Core rev: ad4f82742c6f32b35e2b965a32cccc96aad02d19)
Signed-off-by: Kevin Hao <kexin.hao@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Instead of ignoring ${COREBASE}/meta in PSEUDO_IGNORE_PATHS (which may
or may not ignore all layers depending on how they are named and placed
under ${COREBASE}), ignore all layers.
(From OE-Core rev: cc30537bb4d0815c4d69f0b1a1a8a200085902d5)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If files are copied into /usr/src/debug directly from WORKDIR (e.g. makedevs)
we'd get the permissions from the checkout which would depend on the host umask.
Avoid this and be deterministic by setting the file modes consistently. Core
code copies the files in so we're responsible for the permissions.
Unfortunately to force this change to apply we need to invalidate both
the package tasks and the hash equivalance mappings since file mode
'corruption' already made it into the output hashes (both input options
were mapped to the output hashes).
(From OE-Core rev: 1f958bcd6c9cd12ec76d80586cba15f4d6ed17a7)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Current version of Qemu in OE-core now works fine in rv32/user-mode the
said nvalid instruction errors are gone, so we can enable it now
(From OE-Core rev: f3fa54f91eef5b1b967a6a14b53a07de052dd17a)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Commit 05a87be51b44608ce4f77ac332df90a3cd2445ef introduced
a Python conditional expression when updating PATH that
generates syntax warnings in bitbake-cookerdaemon.log:
Var <PATH[:=]>:1: SyntaxWarning: "is not" with a literal. Did you mean "!="?
Fix this by using the more appropriate '!=' comparison
operator.
(From OE-Core rev: 2e753a12cf6bb98f9e0940e5ed6255ce8c538eed)
Signed-off-by: Matt Madison <matt@madison.systems>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
webkit fails to link with LTO on at this point. Some of the issues
are being worked on upstream and disabling LTO should be re-evaluated as
we update this change.
(From OE-Core rev: fb24d6504c30d3302fda42b6e6a918d674323a8f)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wayland-utils contains wayland-info utility which deprecates
and replaces weston-info from weston.
(From OE-Core rev: bb8fa9950e716fdcee818a38dc4df3a19b1f3c2f)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
perf invokes kernel sources, and currently fails in multiple places for
now therefore disable LTO for perf
(From OE-Core rev: f59a2db89a245d3ef9acd094f2055fb07dd4f42b)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
