When installing a kernel package, the symlinks created by
update-alternatives should point to a path relative to KERNEL_IMAGEDEST
and not an absolute path to '/boot'.
Failing to do so causes problems when resolving the symlink inside some
bootloaders which mount the boot partition elsewhere.
(From OE-Core rev: c7bc46b9bc29dd0953ab8d63b50fa105bb66892e)
Signed-off-by: David Vincent <freesilicon@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When calculating PKGSIZE we sum the size of files after doing lstat() so we
don't count directory metadata overhead, but were not correctly handling
hardlinks. This results in packages such as e2fsprogs-mke2fs having PKGSIZE of
1.5M when it's actually a single 300K binary with five hardlinks.
[ YOCTO #10423 ]
(From OE-Core rev: e82a7f879419828b42b5cc71e6229c8235090f21)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
getVarFlag() now defaults to expanding by default, thus remove the
True option from getVarFlag() calls with a regex search and
replace.
Search made with the following regex:
getVarFlag ?\(( ?[^,()]*, ?[^,()]*), True\)
(From OE-Core rev: 2dea9e490a98377010b3d4118d054814c317a735)
Signed-off-by: Joshua Lock <joshua.g.lock@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
getVar() now defaults to expanding by default, thus remove the True
option from getVar() calls with a regex search and replace.
Search made with the following regex: getVar ?\(( ?[^,()]*), True\)
(From OE-Core rev: 7c552996597faaee2fbee185b250c0ee30ea3b5f)
Signed-off-by: Joshua Lock <joshua.g.lock@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some c++ libraries fail to build if uninative is built
with gcc 5.x and host gcc version is either 4.8 or 4.9.
The issue should be solved by making separate uninative sstate
directory structure sstate-cache/universal-<gcc version> for host gcc
versions 4.8 and 4.9. This causes rebuilds of uninative if host gcc
is either 4.8 or 4.9 and it doesn't match gcc version used to build
uninative.
[YOCTO #10441]
(From OE-Core rev: d36f41e5658bbbb6080ee833027879c119edf3e0)
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The initramfs image recipes changed to use PACKAGE_INSTALL
so they will not be affected by IMAGE_INSTALL, and will cause
error when inherit extrausers:
| ERROR: core-image-minimal-initramfs-1.0-r0 do_rootfs:
core-image-minimal-initramfs: usermod command did not succeed.
So use PACKAGE_INSTALL as well in extrausers.bbclass to fix it.
(From OE-Core rev: fa541362e2d2cc0494a86a413b7b52dfe3eee908)
Signed-off-by: Jackie Huang <jackie.huang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Move patch_path(), src_patches() and should_apply() to oe.patch, making
them easier to call from elsewhere (particularly across the
UI/server boundary).
(From OE-Core rev: 2724511e18810cc8082c1b028e3b7c8a8b5def56)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If PATCHTOOL is "git", and PATCH_COMMIT_FUNCTIONS is set to "1", for
additional tasks between do_unpack and do_patch, make a git commit. This
logic was previously implemented in devtool itself, but it makes more
sense for it to be implemented in the patch class since that's where the
rest of the logic is for this (or in lib/oe/patch.py). It also makes
it possible for this to work with tinfoil2.
(From OE-Core rev: f24f59ea1d8bc335ea8576f6a346d0935f4a3548)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
With the changes to the code for extracting source for a recipe, we are
properly executing the tasks for a recipe, which means their stamps (and
therefore signatures) are important. When running devtool extract on
the lsof recipe I noticed that do_fetch and do_unpack were executing a
second time when we called for do_patch, and this turned out to be
because LIC_FILES_CHKSUM in that recipe contains an entry which
is an absolute path (has ${S} at the start). Normally this wouldn't be
an issue since S is under TMPDIR and thus the existing code would ignore
it, however devtool's extraction code extracts to a temporary directory
which is not under TMPDIR; the result was the path to this file was not
being ignored and the second time around when the license file had been
extracted it was incorporated into the signature. We don't want this, so
explicitly exclude S as well as B and WORKDIR for good measure.
(From OE-Core rev: 1c99d74a862f25e23ea6465fab7ddc9ce74d6974)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If you printed a warning through bb.warn() / bbwarn or an error through
bb.error() / bberror, this was also being picked up by our log_check
mechanism that was designed to pick up warnings and errors printed by
other programs used during do_rootfs. This meant you saw not only the
warning or error itself, you saw it a second time through log_check,
which is a bit ugly. Use the just-added BB_TASK_LOGGER to access the
logger and add a handler that we can use to find out if any warning or
error we find in the logs is one we should ignore as it has already been
printed.
Fixes [YOCTO #8223].
(From OE-Core rev: fb37304d27857df3c53c0867e81fbc8899b48089)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
To trap errors and halt processing, do_kernel_metadata was recently
switched to exit on any non zero return code. While the concept is
sound, there are subcommands that have legitimate non-zero return
codes.
Instead of removing set +e, we'll explicitly check the return code
of the commands that can error, and throw a bbfatal to alert the
user.
(From OE-Core rev: a4705e62d0973c290011fc0d250501d358b659e8)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We need to add do_write_qemuboot_conf to the list of tasks to be wiped as
otherwise the second time an image is built it will fail.
[ YOCTO #10758 ]
(From OE-Core rev: 3206f408e908870629eb47afac6249a3c9497a66)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There's no need to chdir() as tarfile.add() can be told what name to use in the
archive.
(From OE-Core rev: d0b282bce34db44dde4dd7f53a64dfaafe6789de)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a new variable CVE_PRODUCT for the product name to look up in the NVD
database. Default this to BPN, but allow recipes such as tiff (which is libtiff
in NVD) to override it.
(From OE-Core rev: ba330051570a4c991885ee726cb187e0c911bd4f)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This change allows to pass QA for packages built with sh4-oe-linux-musl
toolchain, the problem is reproted while building core-image-minimal target:
ERROR: readline-7.0-r0 do_package_qa:
Error executing a python function in exec_python_func() autogenerated
(From OE-Core rev: dff25cb16ffca0876e5a747644bd6223129df402)
Signed-off-by: Vladimir Zapolskiy <vz@mleia.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In hybrid systemd/sysvinit builds, if the recipe inherits systemd and systemd
is installed, we can safely assume that the service gets stopped by the prerm
script fragment from systemd.bbclass. This fixes deinstallation of packages
with initscripts returning errors when no running service was found. The preinst
shouldn't run the initscript either, because postinst will call systemctl restart.
(From OE-Core rev: f2cc0fdfd97b097145e40bc52674e17378cec863)
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If GLIBC_SPLIT_LC_PACKAGES is set to a non-zero value, convert
glibc-binary-localedata-XX-YY to be a meta package depending on
glibc-binary-localedata-XX-YY-lc-address and so on. This enables
saving quite some space if someone doesn't need LC_COLLATE for
example.
Some regex code was removed from output_locale_binary_rdepends,
because legitimize_package_name already converts to lowercase.
(From OE-Core rev: 40b4b31092058f8833ada3dfc254d362dd3139c1)
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixed:
MACHINE = "qemuarm"
IMAGE_FSTYPES += "ext3.bz2.u-boot"
[snip]
No IMAGE_CMD defined for IMAGE_FSTYPES entry 'ext3.bz2.u-boot' - possibly invalid type name or missing support class
[snip]
This is because image_types_uboot is not inherited, inherit it when
needed will fix the problem.
(From OE-Core rev: 742a22ab7fd333e99d8701220d5a1db28347b1af)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
These files cause warnings because they don't get packaged, now that they
don't land in kernel-vmlinux anymore.
(From OE-Core rev: a49569e3a7534779bbe3f01a0647fd076c95798d)
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
With multiple kernel images enabled starting with 'vmlinux', e.g.
vmlinux.gz and vmlinux.bin, all files landed inside the
kernel-vmlinux package.
On top of that, even initramfs images were included, e.g.
vmlinux.gz-initramfs-*.
(From OE-Core rev: b7f4133b44b740e8ac8e758b0d4a3ee32d326332)
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When iterating over kernel image types to set up their packaging
variables, don't use make targets but the real names.
It was surprising if both vmlinux.bin and vmlinux.gz were enabled
and only the latter had its filename extension removed from the
package name.
(From OE-Core rev: aa189f183e10588f7e8d642f351bd9b8d69f3ea9)
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The code failed to copy the initramfs in case it was a plain
cpio archive.
(From OE-Core rev: 7dbdb4ea91aa027866da2bd46c65fe65a25c848f)
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This hasn't been used in years and isn't referenced anywhere, drop
the definition.
(From OE-Core rev: 68258f856f08f35813964e58ef761471e29373ad)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Historically the processing of kernel meta data contained some
commands that exited with a non-zero return code. Special processing
was required to properly deal with their exit.
That is no longer true, and instead of handling all return codes
and doing an explicit 'exit' call, we can remove set -e from the
routine and have all errors be trapped and stop processing.
(From OE-Core rev: 476ffd57cf5b6fba40d4e3f5dd913824ab8a8d3d)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The do_deploy runs depmodwrapper which requires
pkgdata/kernel-depmod/kernel-abiversion, which is generated by
PACKAGEFUNCS. So we need let do_deploy depend on do_packagedata.
This can fix the errors which happens sometimes when kernel upgrades:
DEBUG: Executing shell function do_deploy
Error: Kernel version 4.8.8-WR9.0.0.1_standard does not match kernel-abiversion (4.8.8-WR9.0.0.0_standard)
And we only see this error when kernel upgrades and rebuild, but doesn't
see it in a normal build, this is becuase depmodwrapper doesn't exit
error when kernel-depmod/kernel-abiversion doesn't exit, it just prints
an error which should go into log.do_deploy:
if [ ! -r /path/to/sysroots/qemux86-64/pkgdata/kernel-depmod/kernel-abiversion ]; then
echo "Unable to read: /path/to/sysroots/qemux86-64/pkgdata/kernel-depmod/kernel-abiversion" >&2
else
[foo]
fi
We can see that there is no "exit 1", I guess it was designed to let it
can run without kernel-abiversion
(From OE-Core rev: 64148a30397b8c92414262c0d414d103d2b97ddd)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Hooks into the new monitordisk.py event and records the used space for
each volume. That is probably the only relevant value when it comes to
visualizing the build and recording more would only increase disk
usage.
(From OE-Core rev: 21a5b569370f47cc02291e1d8b76fe43faa04ea6)
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
/proc/[diskstats|meminfo|stat] get sampled and written to the same
proc_<filename>.log files as during normal bootchat logging. This will
allow rendering the CPU, disk and memory usage charts.
Right now sampling happens once a second, triggered by the heartbeat
event.That produces quite a bit of data for long builds, which will be
addressed in a separate commit by storing the data in a more compact
form.
(From OE-Core rev: 6f4e8180b5b4857eaf6caf410fd3a4a41ed85930)
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The current class works fine when a recipe uses SYSTEMD_AUTO_ENABLE
'enable' and has no on device pkg_postinst(), ie when the postinst is
run as part of rootfs creation. However, when there is a component of
pkg_postinst() that is run on device the 'systemctl restart' is run as
part of the run_postinsts.service at boot. This results in the boot
spinning indefinitely with:
[ *** ] A start job is running for Run pending postinsts (7s / no limit)
The issue could potentially be that the packages service has an
'After' clause which comes later in the boot, beyond
run_postinsts.service, creating a chicken before the egg
scenario. Even service files without an 'After' clause cause this
situation however. Despite this not being the cause of the issue this
fix will prevent this scenario from happenning.
Using strace we are able to find that during boot, when
run_postinsts.service is running attempting to start or restart any
service will result in the call get stuck on poll(). Since the
run_postinsts.service does not monitor the outcome of the call to
restart we can work around this by using '--no-block'.
(From OE-Core rev: 6ad6a0084a73088fc2a27ab9958e5c46d6e094fc)
Signed-off-by: Mark Asselstine <mark.asselstine@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Writing qemuboot.conf in write_qemuboot_conf() does not modify the
rootfs and thus conceptually shouldn't be executed as part of rootfs
creation.
Running it as separate task is cleaner and fixes the problem of
missing qemuboot.conf files for meta-swupd virtual images; those
images replace do_rootfs and ROOTFS_POSTPROCESS_COMMANDs don't run at
all.
The task gets added such that it runs roughly at the same time as
before. Probably it doesn't actually need to depend on do_rootfs, but
this way we don't write a useless qemuboot.conf in cases where
do_rootfs fails.
(From OE-Core rev: c8260447ed115bc5be7df9b25e449a7744d9f2df)
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Instead of being executed for every file in every package, this is now just
called for each package. It is also now correctly called for packages which
don't have any content but do have postinst scripts.
[ YOCTO #10711 ]
(From OE-Core rev: afda72b2424528eaff9054327530bdf5654bec66)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
QAPATHTEST defines a function that is executed for every file in every package.
For tests which just need to look at the datastore this is massive overkill.
Add QAPKGTEST, which is invoked for each package in the recipe.
(From OE-Core rev: acc3cc26099c77e4eeb44c75bc7167ab58ef1147)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Pull the test matrix processing out as a function so it can be reused.
(From OE-Core rev: 3caccd3f6079b7e284d32e1eb0217107425e7bf8)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This test should be looking for the expanded value of ${D} in the expanded value
of pkg_postinst and so on, but one of the getVar() calls was passing
expand=False so the test would never be true.
(From OE-Core rev: cc545044cba51317bee32e3bf674723e422e3a8a)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add mappings for aarch64-elf and aarch64_be-elf to binary lookup table
which allows for the generation of baremetal toolchains.
(From OE-Core rev: e90a1c4b8fd7baa738eb4683e5eac60905e04296)
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
gnome-common is deprecated upstream and an increasingly smaller fraction of
GNOME packages still use it. To avoid adding it redundantly to DEPENDS, remove
it from the class.
(From OE-Core rev: 1d18c2b7bfaed77b30e02d14ae581d920e4eb446)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The change to move C++ ABI tweaks to bitbake.conf should make this redundant, so
revert it.
This reverts commit c56cd49a12.
(From OE-Core rev: af280990bff4a484fd8a18e3442b56a0a39b7611)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Create a common include file for compiler flags which allow native binaries to
be interoperable on a wide range of hosts. In particular the C++ ABI is
problematic so choose the CXX11 version to allow interoperation between gcc4 and
gcc5 based hosts. Moving this to a common include instead of uninative.bbclass
allows uninative to be configured later and used in the eSDK (where its
mandatory) even if the base configuration doesn't enable uninative by default
(e.g. nodistro in OE-Core).
[ YOCTO #10645 ]
(From OE-Core rev: 60c912ae9306532bdd4c5e09a65863ee77c12f43)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
All lines except one (the one containing the 'Elapsed Time') follows the format
'stat: value'. Fix that so post parsing the stats is simpler.
(From OE-Core rev: 47dfd172754440a7ab28eb1b59d9bd484cdad22a)
Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When attempting to run devshell, if no terminal is available, the
error being thrown was not very specific. This adds a list of
commands that failed, informing the user of what they can install to
fix the error.
[ YOCTO #10472]
(From OE-Core rev: c077f4aab2fc956408d4ad45c4e2e2ea6e480624)
Signed-off-by: Stephano Cetola <stephano.cetola@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The old code:
network_enabled = not d.getVar('BB_NO_NETWORK', True)
It is True only when BB_NO_NETWORK is not set (None),
but BB_NO_NETWORK = "0" should also be True while "1" means no network,
"0" means need network in a normal case.
(From OE-Core rev: 3d8db6cb992f96023a0486f64fe6b0f1ead04184)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This can help fix the problem when the error happens.
Now the error message is:
Fetcher failure for URL: 'https://www.example.com/'. URL https://www.example.com/ doesn't work.
Please ensure your host's network is configured correctly,
or set BB_NO_NETWORK = "1" to disable network access if
all required sources are on local disk.
(From OE-Core rev: 0b31a03c59efea3a8742762772e39dcbd6f8ed25)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The TOOLCHAINEXT_OUTPUTNAME is different from TOOLCHAIN_OUTPUTNAME, it
is used for eSDK only, so that it doesn't mix with SDK, use "?=" for it
so that other conf file can define it.
If we don't use "?=" here, then we need use forcevariable to redfine it:
TOOLCHAINEXT_OUTPUTNAME_forcevariable = "foo"
(From OE-Core rev: 933539a07af3f1cca79e49b0ea92cb7699095f96)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
It is possible to use gzip or bzip2 to compress patches and still refer
to them in compressed form in the SRC_URI value within a recipe. If you
run "devtool modify" on such a recipe, make changes to the commit for
the patch and then run devtool update-recipe, we need to correctly
associate the commit back to the compressed patch file and re-compress
the patch, neither of which we were doing previously.
Additionally, add an oe-selftest test to ensure this doesn't regress in
future.
Fixes [YOCTO #8278].
(From OE-Core rev: e47d21624dfec6f71742b837e91da553f18a28c5)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We strip the TOOLCHAIN_OPTIONS and HOST_CC_ARCH from CC/CXX in cmake.bbclass
whereas CFLAFS and CXXFLAGS assume that TOOLCHAIN_OPTIONS are
part of CC/CXX variables, this causes compile failures when cmake
is running compiler tests during configure on some architectures
especially armhf, because hf ABI information -mfloat-abi is part
of TOOLCHAIN_OPTIONS, so what happens is that testcase gets compiled
without hard-float, howver, during linking the float ABI option
is passed via LDFLAGS, now linker rejects this and fails like
/mnt/a/build/tmp-glibc/sysroots/x86_64-linux/usr/libexec/arm-oe-linux-gnueabi/gcc/arm-oe-linux-gnueabi/6.2.0/ld: error: cmTC_27947 uses VFP register arguments, CMakeFiles/cmTC_27947.dir/src.cxx.o does not
mnt/a/build/tmp-glibc/sysroots/x86_64-linux/usr/libexec/arm-oe-linux-gnueabi/gcc/arm-oe-linux-gnueabi/6.2.0/ld: failed to merge target specific data of file CMakeFiles/cmTC_27947.dir/src.cxx.o
collect2: error: ld returned 1 exit status
This means that CMake now fails the configure time test too
which is not right, e.g. it might disable features which actually do exist
and should be enabled e.g. in case above it is resulting as below
Performing C++ SOURCE FILE Test HAS_BUILTIN_SYNC_SUB_AND_FETCH failed with the following output:
Its actually a bug in CMake see
https://gitlab.kitware.com/cmake/cmake/issues/16421
CMake is ignoring CMAKE_CXX_FLAGS when using CHECK_CXX_SOURCE_COMPILES
function.
Until it is fixed upstream, we add HOST_CC_ARCH and TOOLCHAIN_OPTIONS
to CFLAGS and CXXFLAGS, so that we can ensure that compiler invocation
remains consistent.
(From OE-Core rev: 826f3cdb474b5728b22f08d2342fc90235ca9e7d)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
