Add support for generating python3 recipes using the recipetool / devtool.
Drop python2 support at the same time.
Tested with:
oe-selftest -r recipetool.RecipetoolTest
[YOCTO #13264]
(From OE-Core rev: d8b2f58974482b3b1ccc65c5f93104d0d7ba87bc)
Signed-off-by: Maciej Pijanowski <maciej.pijanowski@3mdeb.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wic images are handled as vmtype images. Starting qemu with "-kernel"
parameter and an image of type wic is not supported. Especially for
"-machine virt" the combination of wic with -kernel parameter would
be beneficial.
The new parameter QB_FSINFO allows to pass image type specific flags to
runqemu. QB_FSINFO is a space separated list of parameters. Parameters are
structured according to the following pattern: image-type:flag.
For now two parameters are supported:
- wic:no-kernel-in-fs
The wic image is treated as rootfs only image. A -kernel option is
passed to qemu.
- wic:kernel-in-fs
The wic image is treated as VM image including a bootloader and a
kernel. This is still the default behavior.
Example:
QB_DEFAULT_FSTYPE = "wic"
QB_FSINFO = "wic:no-kernel-in-fs"
QB_KERNEL_ROOT = "/dev/vda1"
QB_SYSTEM_NAME = "qemu-system-aarch64"
QB_MACHINE = "-machine virt"
...
[YOCTO #13336]
(From OE-Core rev: 2aa79a67affd22dfa37e4c2945c6ab0c86321f98)
Signed-off-by: Adrian Freihofer <adrian.freihofer@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This allows "multiconfig:" targets to continue to work by internally
mapping them to the new "mc:" naming, allowing older builds to work
as before.
(Bitbake rev: c4d90890547af642e99cc541af3415df3559563e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
After real world use its clear the "multiconfig:" prefix to multiconfig tasks,
whilst clear, is also clumbersome. Switch to use the short version instead.
mcdepends will continue to work with "multiconfig:" for now as well. The commandline
will only accept mc: going forward.
[YOCTO #11168]
(Bitbake rev: 821daf093b76504067a8b77dfa4b181af6ec92b4)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Adds an initial test for reproducible builds to the OE selftest. This
initial test builds core-image-minimal using sstate, then does a clean
build without sstate in another build directory, and finally does a
binary comparison of the resulting package files between the two builds.
The test is currently always skipped since it doesn't pass yet, but it
can easily be enabled locally
(From OE-Core rev: 2e591bdf93ec9e59b900562263dfe8e72b163baa)
Signed-off-by: Joshua Watt <JPEWHacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
OE_IMPORTS is not intended to be touched by users, but there are cases
in which layers might want to make additional Python modules available
to Python functions. For example, Python modules defined in the layer
themselves (under meta-layer/lib).
(From OE-Core rev: 00fa8391365863fa7805ad61b2d1a8425b9ea040)
Signed-off-by: Chris Laplante <chris.laplante@agilent.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
testimage.bbclass starts qemu with the first image type found in
the IMAGE_FSTYPES list. It's weird: this ['wic', 'tar'] works but
this ['tar'. 'wic'] does not. If QB_DEFAULT_FSTYPE is defined,
this fstype is booted.
(From OE-Core rev: aedb6bf9b6ccf37f69372642bc4c5dbbca92d0d9)
Signed-off-by: Adrian Freihofer <adrian.freihofer@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
While hacking on this I got an Exception. It's better to define
variables also in python.
Signe:-off-by: Adrian Freihofer <adrian.freihofer@siemens.com>
(From OE-Core rev: 1ea225a86cdee4ed932ede509d3351d5aecae497)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In minimal images the agent-ptrace test is skipped unless gdb is
installed which explains the difference in test counts. We don't
want a build dependency on gdb and the test isn't critical so just
document the dependency.
(From OE-Core rev: d3f29e6e52367e124e3f543f970038c0332ad3e1)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In minimal images all tests pass due to a missing dependency on make.
Add the missing dependecy.
The test list created by the run-ptest script is incorrect as it includes
entries like "fi". Simplify it and correct it.
Some tests are skipped due to mpfr not being enabled. Correctly mark these
as SKIP.
Some tests need the "../gawk" script, create a link to it so the tests
work correctly.
Also print more information in the failed test cases.
(From OE-Core rev: 8392d60e05c4eed0b08273c1a934eeea54215079)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We don't really care about kernels this old anymore, so drop the patch.
(From OE-Core rev: b9630048451342a0ca1080955b0df35b0c94ce46)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Several of the parted fixes are actually upstream in slightly different form.
(From OE-Core rev: 90e032a4dc8e7d56db667d35a15cbf30b9fcdf06)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Drop autotools-specific parts from patches, as all autotools files
have been removed upstream (meson is now the only option).
Remove autotools-specific 0001-Revert-Use-absolute-paths-in-pkg-config-files.patch
Remove backported fix-nl-abaltmon.patch
Add a hunk to Enable-more-tests-while-cross-compiling.patch which comments out
the part of the tests that hard-codes native ld for building.
(From OE-Core rev: c7479bfd5fad4b9ba2e4299048a0c432200a5204)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some file systems don't support fetching the block size (notably the file system
Docker uses for containers), so handle the iotctl() failing and raise the
expected error.
(From OE-Core rev: 3757073726a00c5250556aae3d0daac76b88085e)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
I left out the word "not" in a key sentence in the new
"Checking on the Upgrade Status of a Recipe" section.
Little words can make big differences.
(From yocto-docs rev: 46a524a0d6c9e2ed6428039a4454e384a6ef648d)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
I updated the section to include more detail on what the user
sees in the returned table.
(From yocto-docs rev: 671279548dde7f356888ba1ad1551b8bb1f70a65)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
I show the "devtool check-upgrade-status" command as part of
the "do_checkpkg" task section. I added a cross-reference
link to the "Checking on the Upgrade Status of a Recipe"
section in the ref-manual.
(From yocto-docs rev: 034054f431cc9f0e61c8e156eed66bdf6e1198a1)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The section "Upgrading Recipes" covers several ways of
upgrading a recipe. I added some stuff in the introductory
area that mentions the fact a user might want to check
the upgrade status of a recipe first.
(From yocto-docs rev: bd2fd8d0a91b64455c282ebf4974e40874274f46)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
I added a new section showcasing the "devtool check-upgrade-status"
command.
Also, added a note in the "Upgrading a Recipe" section" to tell
the reader they might want to check on a recipe's upgrade
status before upgrading a recipe.
(From yocto-docs rev: 974f63fe8836fccd51f12b511c23644235cbd4b4)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Specifically missing was the "devtool check-upgrade-status"
command.
(From yocto-docs rev: beaa40e8d1037f237ab588ec6ce256b8c3a18543)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This class was removed for 2.7 and I took the reference section
out of the class area.
(From yocto-docs rev: daf786bebb0286d2ced46c97957d80a94c921c56)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The migration section noted this class as becoming obsolete.
I removed it from the class reference section.
(From yocto-docs rev: 63128888242493fb5671f11a2e5e470f2da6b22a)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ousted these since it was noted in the migration section that
the class had been removed. Modified all references to the
removed class as needed.
(From yocto-docs rev: 7000acc76da6b7ee850b799347ab9083663c7e6a)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixes [YOCTO #10098]
This task needed to cross-reference the "Fetchers" section,
which is in the BitBake User Manual.
(From yocto-docs rev: c8862f4c8cc43777acead0e93fc728f2e8be4641)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In the "Use devtool upgrade to Create a Version of the Recipe
that Supports a Newer Version of the Software" section, we
reference "fetchers" but do not link to the section on them
in the BB manual. I added a link.
(From yocto-docs rev: 633ac78daa2ab3bdde17ecb1a203f6addd6ee4af)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
IrDA support was removed in upstream kernel 4.17,
and irda-utils as well as the feature are now also removed.
(From yocto-docs rev: 5650b3c733310b4d478440449d427de417946c81)
Signed-off-by: Scott Rifenbark <srifenbark@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
One array test attempts to create an array that is far too
large to exist. Different exceptions are thrown for 32 and 64
bit machines, so we account for that when catching them.
(From OE-Core rev: 03e3721674fe7fc22911ec738524cef86c7a5357)
Signed-off-by: Joe Slater <joe.slater@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The Makefile used for bash-ptest can pick up the path to the uninative
loader through BUILD_LDFLAGS. This includes the full path to the
uninative loader, which is not reproducible. Replace it with /bin/false.
It doesn't appear as if these native programs are used in the test
suites and if there are likely to be other problems related to building
them using the BUILD_* flags.
(From OE-Core rev: 1208ff934a2bb6378aa8b219345110a0d56bf767)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
They are now added automatically by the ptest class.
(From OE-Core rev: 25cf1820122bb2b15057aafe1c9e04a733f81bcf)
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We keep seeing "cannot allocate memory" errors from rpm when signing packages
on the autobuilder. The following were tried:
* checking locked memory use (isn't hitting limits)
* Restricting RPM_GPG_SIGN_CHUNK to 1
* Limiting to 10 parallel do_package_write_rpm tasks
* Allowing unlimied memory overcommit
* Disabling rpm parallel compression
and the test still failed. Further invetigation showed that the --auto-expand-secmem
wasn't being passed to gpg-agent which meant the secmem couldn't be expanded hence the
errors when there was pressure on the agent.
The reason this happens is that some of the early gpg commands can start the agent
without the option and it sticks around in memory so a version with the correct
option may or may not get started.
We therefore add the option to all the key gpg calls.
(From OE-Core rev: c7e131a76e522503df55e211dd261829feacfa28)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This fixes the openssh tests in minimal images since they use options
not present in the busybox versions of the commands.
[YOCTO #13295]
(From OE-Core rev: 4059d8eedc5cf6f46a834997b7120150fcec4c0e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
allow wic to list and search for kickstart file in .wks.in extension.
basename show by wic list images to fully exclude extension.
(From OE-Core rev: 2c0a292a790ad069648e37b1b29fcea656fcf3e4)
Signed-off-by: Chee Yang Lee <chee.yang.lee@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This was added 9 years ago as a workaround for a problem with
gcc 4.5 on mips.
Building webkitgtk works for me without it for qemumips.
Debian also builds webkitgtk for 32/64 bit big/little endian mips
without using this workaround.
(From OE-Core rev: 7af322a995a9385f7f452c2988188de98db300c2)
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
After the commit ad522ea6a6 ("runqemu: Let qemuparams override default
settings"), the order of the two "-serial" parameters when running the
qemu have been switched. The effect of this is that the logging thread
will use ttyS1 (of course can't capture the kernel boot message anymore),
and the test command will run on the ttyS0. So the output of the test
command may be mangled by the kernel message (such as call trace), and
let the test command produce a fake timeout error message. We can't fix
it by just adjusting the order of the threadport and serverport, since
it will break some machines such as qemuarm64 which use the virtio
serial. So using the tcpserial to setup both the threadport and
serverport.
[YOCTO Bug 13309]
(From OE-Core rev: 9f2005dee41b1ef5a0d1f7b69bcd6c8352dac016)
Signed-off-by: Kevin Hao <kexin.hao@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In some cases(such as the oeqa's qemurunner), we need to setup multi
serial devices via the '-serial 127.0.0.1:xx" and the order of them
is significant. The mixing use of "tcpserial" and "-serial 127.0.0.1:xx"
cause ambiguous issues and we can't fix it by only adjusting the order
of them. So add the support to pass multi ports to the tcpserial
parameter, this will make sure that the order of setting up the serial
is really what we want.
[YOCTO Bug 13309]
(From OE-Core rev: 766c3b56e5071b5a5a64e88df6d3abe5232dd958)
Signed-off-by: Kevin Hao <kexin.hao@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This introduces a variable, PKGDATA_VARS, that contains the names of
the variables that are to be output in the runtime pkgdata files.
(From OE-Core rev: 43e55bfa040425cf93d94ac626a31f6fd00a7a74)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ship some obex files in the appropriate obex package. This fixes boot
error:
[FAILED] Failed to start Bluetooth OBEX service.
that was caused by the obex.service being shipped in the main package,
rather than the -obex (that includes obexd).
(From OE-Core rev: bc9f5f4c107ea34171aad3245a49b25b671d8679)
Signed-off-by: Diego Rondini <diego.rondini@kynetics.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Several files that are part of libvte (e.g. src/widget.cc)
are licensed LGPLv3+.
(From OE-Core rev: 4e3b013daaa07934f1fa81f1b26fe40fea1e3435)
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
