Age | Commit message (Collapse) | Author |
|
The tests are run from a makefile so this dependency is needed.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Without this we see test failures due to the sudo binary being missing.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Bash's ptest needs glibc-utils (for locale), some extra locales for various tests
it uses options busybox doesn't support for some tools, hence coreutils and also runs
perl for some tests.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Adding these fixes the acl and attr ptests to work within minimal images.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
There are some missing dependencies for the util-linux-ptest package
that causes inconsistencies in the package tests run in different images.
The kernel module in RRECOMMENDS is not build at this time, it needs
more testing and check if the configuration change can be part of the
yocto-kernel-cache repository.
Signed-off-by: Mariano López <just.another.mariano@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Fixes [YOCTO #11028]
When dealing with node modules which use "node-gyp" [1] to build native
addons to node. Some temporary build files stay in the image: object
files, static library files, dependency files, ...
This commit does not keep only the required files, but remove the files
which can leads to QA issues (staticdev with static library files).
[1]: https://github.com/nodejs/node-gyp
Signed-off-by: Jean-Marie LEMETAYER <jean-marie.lemetayer@savoirfairelinux.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Fixes [YOCTO #13349]
When dealing with node modules which have declared "bin" files [1], npm
will create a link in '/usr/bin' with a relative link to
'../lib/node_modules/<module bin file>'.
The commits e9270af4296ce2af292059617a717e42fc17425c and
2713d9bcc39c712ef34003ce8424416441be558e explicitely use
'/usr/lib/node/' as install directory, but does not care about the "bin"
symbolic linked files.
In order to keep valid links, and to keep it as simple as possible, the
path '/usr/lib/node_modules/' is used as install directory for npm. And
a symbolic link is created to have a valid '/usr/lib/node/' path, needed
for node.
[1]: https://docs.npmjs.com/files/package.json#bin
Signed-off-by: Jean-Marie LEMETAYER <jean-marie.lemetayer@savoirfairelinux.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Fixes [YOCTO #12534]
When using npm packages with exotic names, like "JSONSelect" (with
uppercase) or "@angular/cli" (with at sign and slash), there are three
different names:
- the recipe name ("jsonselect" or "angular-cli")
- the npm module name ("JSONSelect" or "@angular/cli")
- the npm pack name ("JSONSelect" or "angular-cli")
The commit fa9c077068a2acea04389fa2b44eb2e93548fce2 allow to have
different recipe name and npm module name by setting the NPMPN variable.
This commit allows to have yet another npm pack name. The pack filename
is now dynamically retrieved from the 'npm pack' command.
Signed-off-by: Jean-Marie LEMETAYER <jean-marie.lemetayer@savoirfairelinux.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
The thread tests need libgcc for pthread_cancel to work.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Avoid:
libgcc_s.so.1 must be installed for pthread_cancel to work
when running the ptest without libgcc.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
This solves ptest runtime errors where make was missing causing the ptests
to fail.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[Fixup for warrior context]
[Dropped ptest fixes for pkg w/o ptests in warrior]
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Many different ptests are breaking as they assume that ${PN}-ptest
depends on ${PN}. It doesn't currently but should. If we fix this, many
different ptests start passing when they previously failed.
It does depend on fixing an issue in the dbus-test recipe which is done
in the preceeding patch (mentioned in case this gets backported).
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Avoid the warning:
WARNING: Nothing RPROVIDES 'nativesdk-rng-tools' (but virtual:nativesdk:/home/pokybuild/yocto-worker/build-appliance/build/meta/recipes-connectivity/openssh/openssh_7.9p1.bb RDEPENDS on or otherwise requires it)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Drop backported patches
License-update: copyright years
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Source: tar.git
MR: 97928
Type: Security Fix
Disposition: Backport from http://git.savannah.gnu.org/cgit/tar.git/commit/?id=cb07844454d8cc9fb21f53ace75975f91185a120
ChangeID: 7aee4c0daf8ce813242fe7b872583560a32bc4e3
Description:
Affects tar < 1.32
fixes CVE-2019-9923
Signed-off-by: Armin Kuster <akuster@mvista.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
The only change is a regression fix:
- Fix dbclient regression in 2019.77. After exiting the terminal would be left
in a bad state. Reported by Ryan Woodsmall
Signed-off-by: Andrej Valek <andrej.valek@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
This reverts commit c3a325b5c2d9315629d014e5ebba552fe045171c.
This seems to be causing:
WARNING: acpica-20180508-r0 do_package: acpica: alternative target (/usr/bin/acpidump or /usr/bin/acpidump.acpica) does not exist, skipping...
WARNING: acpica-20180508-r0 do_package: acpica: NOT adding alternative provide /usr/bin/acpidump: /usr/bin/acpidump.acpica does not exist
WARNING: acpica-20180508-r0 do_package: acpica: alt_link == alt_target: /usr/bin/acpidump == /usr/bin/acpidump
because the 20180508 version in warrior unlike the 20190405 in master doesn't install acpidump binary.
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
do_bundle_initramfs does not have to depend on
${INITRAMFS_IMAGE}:do_image_complete if INITRAMFS_IMAGE_BUNDLE is not
set.
Signed-off-by: Ming Liu <liu.ming50@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Adding the valgrind debug symbol information caused the genericx86-64 image to
overflow the 4GB boundary. Tweak the sizes to avoid autobuilder failures yet
leave enough space all the tests still run successfully.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
* before trying to copy unistd.h into it
* older kernels don't have uapi in tools/include and do_configure fails
with:
DEBUG: Executing shell function do_configure
cp: cannot create regular file '.../perf/1.0-r9/perf-1.0/tools/include/uapi/asm-generic/unistd.h':
No such file or directory
WARNING: exit code 1 from a shell command.
* tools/include/uapi was added in kernel 4.8 with
https://github.com/torvalds/linux/commit/c4b6014e8bb0c8d47fe5c71ebc604f31091e5d3f
tools: Add copy of perf_event.h to tools/include/linux/
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
* gnutls from meta-gplv2 isn't new enough for this version of glib-networking:
meson.build:74:0: ERROR: Invalid version of dependency, need 'gnutls' ['>= 3.4.6'] found '3.3.30'
* disabling gnutls PACKAGECONFIG isn't an option either:
meson.build:131:2: ERROR: Problem encountered: No TLS backends enabled. Please enable at least one TLS backend
* add an option to replace gnutls with openssl
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
volume name should refer to --label in .wks.
Replace the hardcoded volume name with label.
set "ESP" as default name when no lable specified.
Signed-off-by: Chee Yang Lee <chee.yang.lee@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
As runqemu with launch_cmd means directly run the command, don't need set
rootfs or env vars.
Since commit [a847dd7202 runqemu: Let qemuparams override default settings]
applied in oe-core, if launch_cmd contains "qemuparams='***'", it does not
work, which is overridden by latter qemuparams="-serial tcp:127.0.0.1" in
QemuRunner.launch();
So we set qemuparams as a parameter in runqemu, the fix makes it work
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
run-postinsts runs a given set of scripts during the first boot of the
device, when one of these scripts prints something to stdout (isnt
daemonized correctly), since stdout is not available at that time,
the script execution immediately returns with an error (exit_group()),
this error causes the script to terminate all threads within the process,
causing undesired behavior since the script might still had to execute
some other code.
Replace eval built-in with (), since () executes in a subshell,
even if one of the scripts exits, all threads of that process will only
be within that session, this ensures other scripts meant to be run are
still run afterwards.
[YOCTO #13266]
Signed-off-by: Alejandro Enedino Hernandez Samaniego <alejandr@xilinx.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
The commands only work with with bash. If /bin/sh is
dash like in Debian, the command execution fails with
errors like:
Standard Output: /bin/sh: 5: export: --sysroot: bad variable name
and all SDK tests fail.
Signed-off-by: Mikko Rapeli <mikko.rapeli@bmw.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Since openssl 1.1.1 and openssh which uses it, sshd
startup is delayed. The delays range from few seconds
to minutes and even to hours. The delays are visible
in host keys generation and when sshd process is started
in response to incoming TCP connection but is failing
to provide SSH version string and clients or tests time out.
In all cases traces show that sshd is waiting for getentropy()
system call to return from Linux kernel, which returns only
after kernel side random number pool is initialized. The pool
is initialized via various entropy source which may be
missing on embedded development boards or via rngd from
rng-tools package from userspace. HW random number generation
and kernel support help but rngd is till needed to feed that data
back to the Linux kernel.
Example from an NXP imx8 board shows that kernel random number pool
initialization can take over 400 seconds without rngd,
and with rngd it is initialized at around 4 seconds after boot.
The completion of initialization is visible in kernel dmesg with line
"random: crng init done".
More details are available from:
* https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=912087
* https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=897572
* https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=43838a23a05fbd13e47d750d3dfd77001536dd33
* http://www.man7.org/linux/man-pages/man2/getrandom.2.html
Signed-off-by: Mikko Rapeli <mikko.rapeli@bmw.de>
Cc: Mark Hatle <mark.hatle@windriver.com>
Cc: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Cc: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Set NO_RECOMMENDATIONS with weak assignment then it could be set a
default value somewhere else.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Currently setupClass errors were not being mapped back to the failing tests
and they were hence being marked as UNKNOWN and the test statistics were
inaccurate.
This is because whilst the errors were being encoded into the test results
stream, the decoder doesn't cope with an error outside a testStart event.
We patch in an addError handler to the outsideTest parser so that this
does get handled in a way similar to the non-concurrent case.
It would be nice if we didn't have to do this but there doesn't seem
to be any other way to fix this other than forking subunit.
We also make a minor change so another of our changes can cope with
tests without a start time.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
The string format for subunit setupClass/setupModule failures is slightly
different, tweak the regex to correctly handle both cases.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
The third field in the /etc/shadow file (sp_lstchg) contains the date of
the last password change expressed as the number of days since Jan 1,
1970.
Backport the upstream changes to honour SOURCE_DATE_EPOCH for build
reproducibility.
Signed-off-by: Alex Kiernan <alex.kiernan@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
We continued to see encoding problems with ssh commands run in oeqa. After much
research the conclusion was we should use ignore the errors since some occasional bad
locale encoding is better than the unicode decoding issues we were seeing which crashed
large parts of tests.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Paul Barker <paul@betafive.co.uk>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Instead of showing:
RESULTS - ptest.PtestRunnerTest.test_ptestrunner - Testcase 1600: UNKNOWN (32.30s)
map unexpectedSuccesses to PASSED and improve the way they're displayed. We
expect/allow ptest runner to fail but if it passes we should handle it correctly.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
These IDs refer to testopia which we're no longer using. We would now use the test
names to definitively reference tests and the IDs can be dropped, along with their
supporting code.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[Backported to keep in-sync with future qa changes]
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Use HTTPS instead of HTTP whenever one is available.
Signed-off-by: Oleksandr Kravchuk <open.source@oleksandr-kravchuk.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
acpidump is both provided by acpica and pmtools, so use
update-alternatives to fix conflicts:
...
|Error: Transaction check error:
| file /usr/bin/acpidump conflicts between attempted installs of
pmtools-20130209+git0+3ebe0e54c5-r0.i586 and acpica-20190405-r0.i586
...
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
flac uses both 'flac' and 'libflac' as cve product.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Without this, the user has no idea what went wrong.
Signed-off-by: Max Kellermann <max.kellermann@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
This code really needs to be rewritten to not split potential
multibyte characters, for now work around it to avoid exceptions like:
File "/home/pokybuild/yocto-worker/qa-extras2/build/meta/lib/oeqa/core/target/ssh.py", line 211, in run
data = reader.read(1024, 4096)
File "/usr/lib64/python3.6/codecs.py", line 503, in read
newchars, decodedbytes = self.decode(data, self.errors)
UnicodeDecodeError: 'utf-8' codec can't decode byte 0x92 in position 0: invalid start byte
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
When using u-boot-nodtb, the symlink didn't install correctly to the
${DEPLOYDIR}. This commit fixes this bug.
Signed-off-by: Ying-Chun Liu (PaulLiu) <paulliu@debian.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
When UBOOT_DTB_BINARY is empty and because the code now changes
directory into ${B}, the test for the existence becomes `[ -f ]` which
succeeds and subsequently the install fails.
Reorder the code so it's clear that UBOOT_DTB_BINARY empty is an
expected configuration and then quote UBOOT_DTB_BINARY everywhere so
no one trips over this again.
Fixes: bacb59079eb6 ("uboot-sign: add support for different u-boot configurations")
Signed-off-by: Alex Kiernan <alex.kiernan@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
This is done by concatenating the DTB with the kernels public key
to all built u-boot binaries. Furthermore the installation of all
the binaries is required.
Signed-off-by: Andreas Obergschwandtner <andreas.obergschwandtner@skidata.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Richard Leitner <richard.leitner@skidata.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
there is small issue with ptest packaging in cases where
OPKGLIBDIR is set to /usr/lib.
Then all ptest files get packaged in libopkg instead of opkg-ptest and correct QA error is triggered:
ERROR: QA Issue: /usr/lib/opkg/ptest/tests/opkgcl.py contained in package libopkg requires /usr/bin/python3, but no providers found in RDEPENDS_libopkg? [file-rdeps]
# $FILES_libopkg
# set /jenkins/mjansa/build-webos-master/oe-core/meta/recipes-devtools/opkg/opkg_0.4.0.bb:62
# "${libdir}/*.so.* ${OPKGLIBDIR}/opkg/"
FILES_libopkg="/usr/lib/*.so.* /usr/lib/opkg/"
# $FILES_opkg-ptest [2 operations]
# set /jenkins/mjansa/build-webos-master/oe-core/meta/classes/ptest.bbclass:9
# "${PTEST_PATH}"
# rename from FILES_${PN}-ptest data.py:117 [expandKeys]
# "${PTEST_PATH}"
# pre-expansion value:
# "${PTEST_PATH}"
FILES_opkg-ptest="/usr/lib/opkg/ptest"
# $PACKAGES [4 operations]
# set /jenkins/mjansa/build-webos-master/oe-core/meta/conf/bitbake.conf:292
# "${PN}-src ${PN}-dbg ${PN}-staticdev ${PN}-dev ${PN}-doc ${PN}-locale ${PACKAGE_BEFORE_PN} ${PN}"
# set /jenkins/mjansa/build-webos-master/oe-core/meta/conf/documentation.conf:314
# [doc] "The list of packages to be created from the recipe."
# prepend /jenkins/mjansa/build-webos-master/oe-core/meta/classes/ptest.bbclass:20
# "${@bb.utils.contains('PTEST_ENABLED', '1', '${PN}-ptest', '', d)}"
# prepend /jenkins/mjansa/build-webos-master/oe-core/meta/recipes-devtools/opkg/opkg_0.4.0.bb:60
# "libopkg"
# pre-expansion value:
# "libopkg ${@bb.utils.contains('PTEST_ENABLED', '1', '${PN}-ptest', '', d)} ${PN}-src ${PN}-dbg ${PN}-staticdev ${PN}-dev ${PN}-doc ${PN}-locale ${PACKAGE_BEFORE_PN} ${PN}"
PACKAGES="libopkg opkg-ptest opkg-src opkg-dbg opkg-staticdev opkg-dev opkg-doc opkg-locale opkg"
The easiest fix should be to reorder PACKAGES (the _prepends) so that ${PN}-ptest is prepended later -> ends before libopkg).
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Adds an option to dump all the ptest logs to individual files in a
specified directory. If multiple test runs are present, the
'--prepend-run' argument will create separate directories for each test
run under the target directory and put the logs there to prevent each
test run from clobbering the others.
[YOCTO #13331]
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Whenever cml1 do_configure is used with a defconfig, oldconfig waits for
input. This silently fails on recent kconfig projects with:
"Error in reading or end of file."
We cannot use a more up to date kconfig target such as olddefconfig,
because busybox does not support it.
Signed-off-by: Stefan Müller-Klieser <s.mueller-klieser@phytec.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
The python distutils generate a python wrapper script for each package,
containing shebang lines pointing to the python executable.
In our case, this is a fully-qualified path to python-native in the
recipe-sysroot-native.
Ubuntu 18.04 restricts the useful length of the shebang line to 125
characters, and Ubuntu 16.04 restricts it to 77. In both cases, the
staged python script fails to run due to the length of the path to
the python-native executable.
Replace the shebang line with nativepython or nativepython3 as appropriate.
The nativepython symlink is installed by the python-native recipe:
#!/usr/bin/env nativepython
We were already doing this for on-target distutils components.
This change applies the sed-line to -native distutils components as well.
In this way, -native clients of these components can invoke the wrapper scripts
directly, without themselves needing to inherit pythonnative.
This works around a known setuptools issue:
https://github.com/pypa/setuptools/issues/494
Even once this issue has been resolved upstream,
we will still need to replace `python` with `nativepython`
Signed-off-by: Douglas Royds <douglas.royds@taitradio.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|