Kevin Hao [Thu, 13 Feb 2020 02:31:41 +0000 (10:31 +0800)]
wic: Honor APPEND
APPEND is used to add additional parameters to kernel by features, such
as read-only-rootfs. So we should honor this variable when we compose
the kernel parameter in wic. I know we also can resolve this kind of
issue by using the .wks.in template introduced by commit 42e870c5ed4b
("image_types.bbclass: support template .wks.in files for wic"), but the
APPEND is needed by all the wks, it would seems pretty ridiculous to me
to change all the .wks to .wks.in and then foist the APPEND into them.
So the APPEND is definitely deserved to export to the wic directly.
[Yocto #12809]
Signed-off-by: Kevin Hao <kexin.hao@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Robert Yang [Mon, 6 Jan 2020 10:30:32 +0000 (18:30 +0800)]
gnupg: Read GNUPG_BINDIR from environment variables for nativesdk
There is already a relocate.patch for native which is used for reading
GNUPG_BINDIR from environment variables, now also enable it for nativesdk.
Otherwise, command like the following one doesn't work for nativesdk:
$ gpg-connect-agent --homedir ../keys/ reloadagent /bye
gpg-connect-agent: no running gpg-agent - starting '/opt/path/to/sysroots/x86_64-wrlinuxsdk-linux/usr/bin/gpg-agent'
gpg-connect-agent: failed to start agent '/opt/path/sysroots/x86_64-wrlinuxsdk-linux/usr/bin/gpg-agent': No such file or directory
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Since we now have an example recipe on meta-skeleton to
build baremetal applications using OpenEmbedded, a user
produced SDK should be able to run such application.
Include nativesdk-qemu on TOOLCHAIN_HOST_TASK so its
built inside the newlib based SDK.
Signed-off-by: Alejandro Enedino Hernandez Samaniego <alejandro@enedino.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Since we now have an example recipe on meta-skeleton to
build baremetal applications using OpenEmbedded, a user
produced SDK should be able to run such application.
Include nativesdk-qemu on TOOLCHAIN_HOST_TASK so its
built inside the baremetal based SDK.
Signed-off-by: Alejandro Enedino Hernandez Samaniego <alejandro@enedino.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This error handling didn't work as expected since upon failure it would
inject bytestreams back into the code leading to tracebacks.
Instead, ignore the decode errors. Fixes:
Traceback (most recent call last):
File "/home/pokybuild/yocto-worker/a-full/build/scripts/resulttool", line 78, in <module>
sys.exit(main())
File "/home/pokybuild/yocto-worker/a-full/build/scripts/resulttool", line 72, in main
ret = args.func(args, logger)
File "/home/pokybuild/yocto-worker/a-full/build/scripts/lib/resulttool/store.py", line 70, in store
resultutils.save_resultsdata(results, tempdir, ptestlogs=True)
File "/home/pokybuild/yocto-worker/a-full/build/scripts/lib/resulttool/resultutils.py", line 178, in save_resultsdata
f.write(sectionlog)
TypeError: write() argument must be str, not bytes
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 14 Feb 2020 18:13:14 +0000 (18:13 +0000)]
populate_sdk_ext: We now require python3, not python
We no longer expect a "python" binary in PATH so update the eSDK's
expectations to match. This was the only failure on autobuilder test
systems with python missing.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Wang Mingyu [Wed, 12 Feb 2020 18:35:06 +0000 (10:35 -0800)]
ruby: upgrade 2.6.5 -> 2.7.0
0003-configure.ac-check-finite-isinf-isnan-as-macros-firs.patch
Removed since it is included in 2.7.0.
refresh the following patch:
0002-Obey-LDFLAGS-for-the-link-of-libruby.patch
add the following patch to solve build error:
0001-Modify-shebang-of-libexec-y2racc-and-libexec-racc2y.patch
[error message]
ERROR: ruby-2.7.0-r0 do_package_qa: QA Issue:
/usr/lib/ruby/gems/2.7.0/gems/racc-1.4.16/bin/y2racc contained in
package ruby requires /usr/local/bin/ruby, but no providers found in
RDEPENDS_ruby? [file-rdeps]
format of file COPYING is updated.
LEGAL notice information is updated.
Signed-off-by: Wang Mingyu <wangmy@cn.fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Nathan Rossi [Fri, 14 Feb 2020 05:59:43 +0000 (05:59 +0000)]
glibc-testsuite: Remove the do_install task
This task is not intended to be run as this recipe does not produce any
packages or install any sysroot content. Additionally running this task
behaves differently from a normal glibc build, and can result in issues.
[YOCTO #13737]
Signed-off-by: Nathan Rossi <nathan@nathanrossi.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 14 Feb 2020 12:50:18 +0000 (12:50 +0000)]
staging: Handle races between binaries and their libs
There is a long standing issue where a binary could be installed into the
sysroot before its library dependencies. We've always argued nothing should
use the binary until it has been installed by a dependency but there are issues
around binaries which conflict with the host system, for example patch,
python3, gzip and more.
With the recent patch changes we've seen issues like:
ERROR: gdb-cross-canadian-powerpc-8.3.1-r0 do_patch: Command Error: 'quilt --quiltrc /home/pokybuild/yocto-worker/qemuppc/build/build/tmp/work/x86_64-nativesdk-pokysdk-linux/gdb-cross-canadian-powerpc/8.3.1-r0/recipe-sysroot-native/etc/quiltrc push' exited with 0 Output:
Applying patch 0009-Change-order-of-CFLAGS.patch
patch: /lib64/libattr.so.1: version `ATTR_1.3' not found (required by patch)
Patch 0009-Change-order-of-CFLAGS.patch does not apply (enforce with -f)
which is a symptom of this issue (libattr-native is a dependency of patch-native).
There are other ways to fix this such as disabling libattr in patch, installing
patch to a subdirectory and requiring PATH manipulation and so on.
We can simply fix the staging code to handle /bin/ after everything else so
do that and avoid all these other complications.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jeremy Puhlman [Mon, 10 Feb 2020 21:29:26 +0000 (13:29 -0800)]
binutils: additional patch should be added with space.
If an append adds a patch or file with SRC_URI += , the spaceless
append causes the file listings to concatinate and cause failures
in the nativesdk-binutils parse/build.
Signed-off-by: Jeremy Puhlman <jpuhlman@mvista.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Alex Kiernan [Wed, 12 Feb 2020 11:33:36 +0000 (11:33 +0000)]
kernel-fitimage: fix devicetree reproducibility
Ensure that the order of dtb/dtbo files in the generated fitimage is
reproducible.
Fixes: 71bfa9838cbf ("kernel-fitimage: Handle overlays in EXTERNAL_KERNEL_DEVICETREE") Signed-off-by: Alex Kiernan <alex.kiernan@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
He Zhe [Wed, 12 Feb 2020 07:35:57 +0000 (15:35 +0800)]
linux-yocto-dev: Add paravirt_kvm support for qemux86-64
This feature includes paravirtualized KVM guest support, including
KVMCLOCK for enhancing clock accuracy of guest OS. With it we can prevent
the following error.
"clocksource: timekeeping watchdog on CPU3: Marking clocksource 'tsc' as
unstable because the skew is too large"
Signed-off-by: He Zhe <zhe.he@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yi Zhao [Wed, 12 Feb 2020 06:24:40 +0000 (14:24 +0800)]
python3: install _tkinter.*.so to python3-tkinter package
When enable PACKAGECONFIG[tk], we should install _tkinter.*.so to
python3-tkinter package rather than python3-misc package.
Fixes:
ERROR: python3-3.8.1-r0 do_package_qa: QA Issue:
/usr/lib/python3.8/lib-dynload/_tkinter.cpython-38-x86_64-linux-gnu.so
contained in package python3-misc requires libtk8.6.so()(64bit), but no
providers found in RDEPENDS_python3-misc? [file-rdeps]
Signed-off-by: Yi Zhao <yi.zhao@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Wed, 12 Feb 2020 03:14:35 +0000 (21:14 -0600)]
oeqa: reproducible: Run diffoscope on saved output
If there are differing packages and they are being saved for review,
automatically run diffoscope on them and include the output in the saved
output. The output is currently done in HTML format since these are
typically published on a webpage by the autobuilder.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 14 Feb 2020 08:25:18 +0000 (08:25 +0000)]
diffoscope: Exclude from musl builds
Various dependencies don't build for musl so exclude this recipe too, at
least for now until the issues are resolved to avoid build failures in
world builds.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is the official release for Wayland 1.18. The main new features in
this release are:
- Add support for the Meson build system (autotools is still supported
but will be removed in a future release)
- Add API to tag proxy objects to allow applications and toolkits to
share the same Wayland connection
- Track wayland-server timers in user-space to prevent creating too
many FDs
- Add wl_global_remove, a new function to mitigate race conditions with
globals
https://lists.freedesktop.org/archives/wayland-devel/2020-February/041207.html
2 upstreamed patches are dropped.
Signed-off-by: Denys Dmytriyenko <denys@ti.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Khem Raj [Thu, 13 Feb 2020 08:21:59 +0000 (00:21 -0800)]
binutils: Bail out if gold is used on 32/64 bit RISC-V
gold is not supporting risc-v yet and, this will cause subltle errors
during build e.g. autoconf not finding certain features like visibility
and it could take a while to unwind the problem. Its better to error out
early
Remove dwp and ld.gold for rv32 as well
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Khem Raj [Thu, 13 Feb 2020 08:21:58 +0000 (00:21 -0800)]
crosssdk: Set nativesdk specific distro features
Currently, normal distro features e.g. ld-is-gold is impacting
crosssdk recipes, which actually should not be the case, since
that feature is essentially intended for target packages and not
nativesdk packages
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Khem Raj [Thu, 13 Feb 2020 08:21:51 +0000 (00:21 -0800)]
oeqa: Use --disable-maintainer-mode configure option
since the versions of autotools might differ on target and build host, plus difference in
timestamps for configure and system can result in reconfigure lets avoid
that by disabling maintainer mode
Avoids
error: newly created file is older than distributed files!
[YOCTO #13779]
Suggested-by: Andre McCurdy <armccurdy@gmail.com> Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Since its a 32bit kernel 4010000000 address is truncated to 10000000 and ends up
in conflicts with VIRT_PCIE_MMIO, which ranges from 0x10000000 to 0x3efeffff
This is happening because the linux-yocto kernel is not compiled with
LPAE support, however, virt machine for qemuarm assumes that by default
Should LPAE be enabled by default in kernel config is a separate
question
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yeoh Ee Peng [Fri, 31 Jan 2020 05:47:42 +0000 (13:47 +0800)]
scripts/lib/resulttool/report: Enable report selected test case result
Enable reporting selected test case result given the user provided
the selected test case id. If both test result id and test case id
were provided, report the selected test case result from the
selected test result id.
Signed-off-by: Yeoh Ee Peng <ee.peng.yeoh@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The python3native.bbclass set PYTHON_{LIBRARY, INCLUDE_DIR} pointing to
python target build, a patch was add to use target _sysconfigdata too [1]
but is not available without python3 dependency.
Fixes waffle build when trying to find _sysconfigdata module,
...
| Program python3 found: YES (/home/anibal.limon/linaro/oe-rpb-master/build-410c/tmp-rpb-glibc/work/aarch64-linaro-linux/waffle/1.6.0-r0/recipe-sysroot-native/usr/bin/python3-native/python3)
|
| meson.build:36:0: ERROR: <ExternalProgram 'python3' -> ['/home/anibal.limon/linaro/oe-rpb-master/build-410c/tmp-rpb-glibc/work/aarch64-linaro-linux/waffle/1.6.0-r0/recipe-sysroot-native/usr/bin/python3-native/python3']> is not a valid python or it is missing setuptools
|
| A full log can be found at /home/anibal.limon/linaro/oe-rpb-master/build-410c/tmp-rpb-glibc/work/aarch64-linaro-linux/waffle/1.6.0-r0/build/meson-logs/meson-log.txt
| WARNING: exit code 1 from a shell command.
...
Mark Hatle [Fri, 7 Feb 2020 17:43:37 +0000 (11:43 -0600)]
populate_sdk_base: if the SDKIMAGE_FEATURES changes, refresh the SDK
Since the features are processed by a python fragment, we need to explicitly
list the variables that should affect the resulting hash, and thus sstate
re-use.
Signed-off-by: Mark Hatle <mark.hatle@kernel.crashing.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Alex Kiernan [Fri, 31 Jan 2020 16:24:42 +0000 (16:24 +0000)]
kernel-fitimage: Handle overlays in EXTERNAL_KERNEL_DEVICETREE
When using EXTERNAL_KERNEL_DEVICETREE, collect DTB overlays too (*.dtbo)
as well as iterating down into sub-directories so using the behaviour
for naming which matches KERNEL_DEVICETREE.
Signed-off-by: Alex Kiernan <alex.kiernan@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Kevin Hao [Sat, 8 Feb 2020 12:36:42 +0000 (20:36 +0800)]
xserver-nodm-init: Fix the start failure for non-root user
In order to start the xserver, a non-root user should have the
cap_sys_admin capability to set the drm master. We try to get
the cap_sys_admin capability by setting it in both the thread
and file inheritable set. The side effect of this is that we
would have to add the "pam" to the distro features if we want
use the xserver-nodm-init for a non-root user.
[Yocto #11526]
Signed-off-by: Kevin Hao <kexin.hao@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
testimage: Extend runtime testing infrastructure to allow unconventional booting processes to be tested
The current runtime infrastructure contains hardcoded values which Ill refer to
as patterns, these patterns are either searched through or sent via the serial
terminal to communicate between HOST and TARGET.
These patterns are required since they allow us to check when a device has
finished booting, to log in, and to check whether a command sent from our tests
has returned, this way we are able to check both the status of the commands that
were sent along with its output.
The testing process goes somewhat as follows:
1. Launch QEMU and start booting.
2. Check when the device has booted by looking for the pattern login:.
3. Log in as the root user (default for our images).
4. Check that we were able to log in succesfully.
5. Start running the runtime test cases defined by TEST_SUITES.
6. One of such test cases could send a command to the QEMU target.
7. Check whether that command returned.
8. Check its output and status, return whether the test case passed or failed.
This patch allows this set of patterns to be defined instead of being hardcoded,
but it also automatically sets the defaults that we have been using in the past
if they have not been manually defined, for this reason, the patch is less
invasive and should not affect in any way how tests are currently being run.
Cases that can be enabled with this patch:
- A customized image that does not use the root user (or maybe we want to check
what happens if we dont use the root user).
- An image where the PS1 env variable has been modified, and the prompt pattern
wouldnt match the default.
- Baremetal applications, which do not follow the conventional way of booting
Linux and would probably not show a prompt for a user to log in, same applies
for testing bootloaders.
- poky-tiny: Using DISTRO=poky-tiny and an image such as the core-image-tiny
from meta-intel, which boots directly to RAM, and does not show a log in prompt
since it does not contain a conventional init process.
The code itself contains comments that should be self explanatory but here is an
example on how these patterns can be defined in a hypothetical case where we
want to run test cases as the webserver user instead:
The variable TESTIMAGE_BOOT_PATTERNS defines which patterns to override when
used to communicate with the target when booting, anyone familiar with the
PACKAGECONFIG syntax should have no trouble setting these.
Other patterns would still be set up as default, e.g.
search_reached_prompt would still be login:
The accepted flags for TESTIMAGE_BOOT_PATTERNS are the following:
search_reached_prompt, send_login_user, search_login_succeeded,
search_cmd_finished.
They are prefixed with either search/send, to differentiate if the pattern is
meant to be sent or searched to/from the target terminal.
A working example of this code that falls under the baremetal case mentioned
above along with a test case is present on the meta-freertos layer, which tests
an RTOS image built with OpenEmbedded and automatically runs a test case on it
after booting such image:
As usual, INHERIT += "testimage" needs to be present on local.conf
$ bitbake freertos-demo -c testimage
RESULTS:
RESULTS - freertos_echo.FreeRTOSTest.test_freertos_echo: PASSED (2.00s)
SUMMARY:
freertos-demo () - Ran 1 test in 2.006s
freertos-demo - OK - All required tests passed (successes=1, skipped=0,
failures=0, errors=0)
testimage: Allow testing on QEMU machines with a single serial port
commmit 6cde67d0a84 enables the use of qemurunner on machines that
only have a single serial port, but still sets the default value
as serial_ports=2 if not provided.
The testimage class does not call qemurunner with a serial_ports
argument, hence always defaulting to two.
Pass the serial_ports argument from the testimage class to allow
tests to run on QEMU machines with a single serial port.
Signed-off-by: Alejandro Hernandez Samaniego <alejandro@enedino.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Alex Kiernan [Thu, 6 Feb 2020 16:48:13 +0000 (16:48 +0000)]
systemd: upgrade 243.4 -> 244.1
Drop 0002-use-lnr-wrapper-instead-of-looking-for-relative-opti.patch a
variant on which was introduced in 2c1047310970 ("Upgrade to systemd
211+") in 2014. The oldest supported build machine is now CentOS 7 which
has --relative support in ln, so there no longer appears to be any need
for this change.
Rebase/refresh musl patches:
- replace missing.h with specific missing_... header
- fix additional WRITE_STRING_FILE_DISABLE_BUFFER
- fix more places that need netinet/if_ether.h
Signed-off-by: Alex Kiernan <alex.kiernan@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changqing Li [Fri, 7 Feb 2020 06:07:44 +0000 (14:07 +0800)]
python3: Do not hardcode "lib" for distutils
Get the sys.lib from python3 itself and do not use
hardcoded value of 'lib' for distutils.
Solve the error below that occurs when run "python3 setup.py
install"
on lib64 multilib platform:
[Errno 2] No such file or directory:
'/usr/lib/python3.7/site-packages/test-easy-install-1828.write-test'
Signed-off-by: Li Zhou <li.zhou@windriver.com> Signed-off-by: Changqing Li <changqing.li@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changqing Li [Fri, 7 Feb 2020 06:07:43 +0000 (14:07 +0800)]
python3: fix the installation path of libpython3.7m.a for multilib
Fix the installation path of libpython3.7m.a on mulitlib lib64
platform to lib64 instead of lib
Signed-off-by: Li Zhou <li.zhou@windriver.com> Signed-off-by: Changqing Li <changqing.li@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jaewon Lee [Thu, 6 Feb 2020 21:45:11 +0000 (13:45 -0800)]
sstate.bbclass: fix issue while handling long sstate filenames
When moving to python3, divison using '/' now returns float instead of
an integer. In upstream commit b8025e972081b70960ffcbcbe43a7118041556a1
sstate filenames longer than the limit are changed to just include
necessary info + 3 fields just for information. The space left over
after the necessary info is divided into 3 for each of the fields.
Using '//' instead to do the division to solve the following error
message:
avail = (254 - len(hash + "_" + taskname + extension) -
len(components[0]) - len(components[1]) - len(components[5]) -
len(components[6]) - 7) / 3
> components[2] = components[2][:avail]
components[3] = components[3][:avail]
TypeError: slice indices must be integers or None or have an __index__
method
Signed-off-by: Jaewon Lee <jaewon.lee@xilinx.com> Signed-off-by: Mark Hatle <mark.hatle@xilinx.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mark Hatle [Fri, 7 Feb 2020 20:20:08 +0000 (14:20 -0600)]
package.bbclass: Support stripping and debug copy of static libraries
By default, we won't copy and strip static libraries. However, this
functionality can be useful in some cases where people are doing
development on the target, and don't generally want the larger debug
capable static libraries. To enable the new functionality set:
PACKAGE_DEBUG_STATIC_SPLIT = '1'
Add a new function splitstaticdebuginfo. Thus function will copy the
unmodified static library into the specific debug directory location.
By keeping an unmodified version, it is possible for a user trying
to debug something to use -L /usr/lib/.debug-static and their existing
build commands to switch from stripped to full debug versions.
The PACKAGE_DEBUG_SPLIT_STYLE will select between two different
approaches, /usr/lib/debug-static or <path>/.debug-static.
Additionally you can now choose to strip static libraries to conserve
space. If either 'PACKAGE_DEBUG_STATIC_SPLIT' or 'PACKAGE_STRIP_STATIC'
is set to 1, the static library will be stripped. (This is not on by
default, as it could make diagnosing static library usage difficult in
some cases.)
Add to insane.bbclass a skip to the staticdev warning for the specific
-dbg package versions.
Signed-off-by: Mark Hatle <mark.hatle@kernel.crashing.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mark Hatle [Fri, 7 Feb 2020 20:20:07 +0000 (14:20 -0600)]
package.bbclass: Allow INHIBIT_PACKAGE_STRIP_FILES to skip .ko and static libs
Change the order of the skip processing to happen before any .ko and static
library processing. This will allow these types of files to be individually
skipped if necessary.
Signed-off-by: Mark Hatle <mark.hatle@kernel.crashing.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mark Hatle [Fri, 7 Feb 2020 20:20:06 +0000 (14:20 -0600)]
package.bbclass: Fix debug source processing for static libraries
Format of the sources list is the [ (file, [source, ...]), ... ] before
this change, the static libraries were processed but the items were
included incorrectly causing no sources for static libraries to be
included.
Signed-off-by: Mark Hatle <mark.hatle@kernel.crashing.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
chkconfig is one of the options for virtual/update-alternatives,
however opkg-utils have been used as the default for a very long time,
while chkconfig isn't anymore tested in any way, and is stuck
at a very old version due to newer versions requiring selinux.
[YOCTO #11264]
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Trevor Gamblin [Fri, 7 Feb 2020 13:46:52 +0000 (08:46 -0500)]
qemurunner.py: add try/except for pid handling race
In some instances, attempts to remove the qemu pidfile within the
stop() method fail despite the os.path.exists() call immediately
before implying that the file is present. Add a try/except block
to log a warning if this occurs, rather than failing outright,
since the process simply appears to be exiting at an inconvenient
time.
[YOCTO #13675]
Signed-off-by: Trevor Gamblin <trevor.gamblin@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bruce Ashfield [Fri, 7 Feb 2020 14:52:04 +0000 (09:52 -0500)]
kern-tools: fix merge_config when LD contains parameters
To ensure that the kernel linker is used when allno/mod/yes config
merge_config steps were executed, the call to make was tweaked to
explicitly pass LD.
But since the variable wasn't quoted, any parameters to LD (like
the sysroot) were mistakenly passed to make, and hence could trigger
an error on some architectures.
We also tweak the logging to hightlight errors like this in the
future and avoid losing it in the noise of merge configs sometimes
overly verbose output.
Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Gavin Li [Thu, 6 Feb 2020 17:35:52 +0000 (12:35 -0500)]
kernel-yocto: fix defconfig detection in find_sccs()
The current code would cause a file like "config.bin" to added to the
config sources list. I am sure the intention was to add any files with
defconfig in its name and not the other way around.
Signed-off-by: Gavin Li <gavinli@thegavinli.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Martin Jansa [Thu, 6 Feb 2020 17:35:51 +0000 (12:35 -0500)]
kernel-yocto.bbclass: do_kernel_configme: don't use +errexit, merge_config_build.log and fail when /.config wasn't created
* for whatever reason, instead of silently continuing to build default kernel config
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Martin Jansa [Thu, 6 Feb 2020 17:35:50 +0000 (12:35 -0500)]
kernel-yocto.bbclass: set KERNEL_LD also for merge_config.sh
* when ld-is-gold is used, merge_config.sh silently fails and doesn't generate anything
useful in .config (and also include directory isn't created):
tmp-glibc-bfd-qemux86-do_kernel_configme/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/linux-qemux86-standard-build:
total 164K
drwxr-xr-x 4 martin martin 4.0K Feb 5 02:27 .
drwxrwxr-x 7 martin martin 4.0K Feb 5 02:27 ..
-rw-rw-r-- 1 martin martin 133K Feb 5 02:28 .config
-rw-rw-r-- 1 martin martin 39 Feb 5 02:27 .gitignore
drwxr-xr-x 4 martin martin 4.0K Feb 5 02:27 include
-rw-rw-r-- 1 martin martin 201 Feb 5 02:27 Makefile
drwxrwxr-x 4 martin martin 4.0K Feb 5 02:27 scripts
lrwxrwxrwx 1 martin martin 61 Feb 5 02:27 source -> /OE/build/oe-core/tmp-glibc/work-shared/qemux86/kernel-source
tmp-glibc-gold-qemux86-do_kernel_configme/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/linux-qemux86-standard-build:
total 28K
drwxr-xr-x 3 martin martin 4.0K Feb 5 02:26 .
drwxrwxr-x 7 martin martin 4.0K Feb 5 02:26 ..
-rw-rw-r-- 1 martin martin 74 Feb 5 02:26 .config
-rw-rw-r-- 1 martin martin 39 Feb 5 02:26 .gitignore
-rw-rw-r-- 1 martin martin 201 Feb 5 02:26 Makefile
drwxrwxr-x 4 martin martin 4.0K Feb 5 02:26 scripts
lrwxrwxrwx 1 martin martin 61 Feb 5 02:26 source -> /OE/build/oe-core/tmp-glibc/work-shared/qemux86/kernel-source
* and because the failure is silent, it will happily start building
kernel with default defconfig, which even builds OK for qemux86-64
but on qemux86 I've noticed this issue, because incorrectly
configured kernel build fails with:
$ tail -n 20 tmp-glibc-gold-qemux86/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/temp/log.do_compile
LD /OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/linux-qemux86-standard-build/tools/objtool/libsubcmd-in.o
AR /OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/linux-qemux86-standard-build/tools/objtool/libsubcmd.a
HOSTCC scripts/mod/mk_elfconfig
CC scripts/mod/devicetable-offsets.s
CC scripts/mod/empty.o
cc1: error: code model 'kernel' not supported in the 32 bit mode
cc1: sorry, unimplemented: 64-bit mode not compiled in
make[2]: *** [/OE/build/oe-core/tmp-glibc/work-shared/qemux86/kernel-source/scripts/Makefile.build:99: scripts/mod/devicetable-offsets.s] Error 1
make[2]: *** Waiting for unfinished jobs....
cc1: error: code model 'kernel' not supported in the 32 bit mode
cc1: sorry, unimplemented: 64-bit mode not compiled in
make[2]: *** [/OE/build/oe-core/tmp-glibc/work-shared/qemux86/kernel-source/scripts/Makefile.build:266: scripts/mod/empty.o] Error 1
make[1]: *** [/OE/build/oe-core/tmp-glibc/work-shared/qemux86/kernel-source/Makefile:1111: prepare0] Error 2
make[1]: *** Waiting for unfinished jobs....
LD /OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/linux-qemux86-standard-build/tools/objtool/objtool-in.o
/OE/build/oe-core/tmp-glibc/work-shared/qemux86/kernel-source/tools/objtool
LINK /OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/linux-qemux86-standard-build/tools/objtool/objtool
make: *** [/OE/build/oe-core/tmp-glibc/work-shared/qemux86/kernel-source/Makefile:179: sub-make] Error 2
WARNING: exit code 1 from a shell command.
* the issue happens in log.do_kernel_configme, but the log is completely useless:
cat tmp-glibc-gold-qemux86-do_kernel_configme/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/temp/log.do_kernel_configme
DEBUG: Executing python function extend_recipe_sysroot
NOTE: Direct dependencies are ['virtual:native:/OE/build/oe-core/openembedded-core/meta/recipes-devtools/bison/bison_3.5.bb:do_populate_sysroot', '/OE/build/oe-core/openembedded-core/meta/recipes-kernel/kern-tools/kern-tools-native_git.bb:do_populate_sysroot', 'virtual:native:/OE/build/oe-core/openembedded-core/meta/recipes-extended/bc/bc_1.07.1.bb:do_populate_sysroot', '/OE/build/oe-core/openembedded-core/meta/recipes-devtools/binutils/binutils-cross_2.33.bb:do_populate_sysroot', '/OE/build/oe-core/openembedded-core/meta/recipes-devtools/quilt/quilt-native_0.66.bb:do_populate_sysroot', '/OE/build/oe-core/openembedded-core/meta/recipes-devtools/gcc/gcc-cross_9.2.bb:do_populate_sysroot']
NOTE: Installed into sysroot: ['bison-native', 'bc-native', 'binutils-cross-i686', 'gcc-cross-i686', 'autoconf-native', 'xz-native', 'flex-native', 'gnu-config-native', 'automake-native', 'libtool-native', 'texinfo-dummy-native', 'gettext-minimal-native', 'readline-native', 'zlib-native', 'gmp-native', 'libmpc-native', 'mpfr-native', 'linux-libc-headers', 'm4-native', 'ncurses-native', 'pkgconfig-native']
NOTE: Skipping as already exists in sysroot: ['kern-tools-native', 'quilt-native']
DEBUG: sed -e 's:^[^/]*/:/OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/recipe-sysroot-native/:g' /OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/bison-native/fixmepath /OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/gcc-cross-i686/fixmepath /OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/autoconf-native/fixmepath /OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/gnu-config-native/fixmepath /OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/automake-native/fixmepath /OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/libtool-native/fixmepath /OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/gmp-native/fixmepath /OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/ncurses-native/fixmepath /OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/pkgconfig-native/fixmepath | xargs sed -i -e 's:FIXMESTAGINGDIRTARGET:/OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/recipe-sysroot:g; s:FIXMESTAGINGDIRHOST:/OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/recipe-sysroot-native:g' -e 's:FIXME_PSEUDO_SYSROOT:/OE/build/oe-core/tmp-glibc/sysroots-components/x86_64/pseudo-native:g' -e 's:FIXME_HOSTTOOLS_DIR:/OE/build/oe-core/tmp-glibc/hosttools:g' -e 's:FIXME_PKGDATA_DIR:/OE/build/oe-core/tmp-glibc/pkgdata/qemux86:g' -e 's:FIXME_PSEUDO_LOCALSTATEDIR:/OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/pseudo/:g' -e 's:FIXME_LOGFIFO:/OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/temp/fifo.10476:g'
DEBUG: Python function extend_recipe_sysroot finished
DEBUG: Executing shell function do_kernel_configme
DEBUG: Shell function do_kernel_configme finished
]because merge_config.sh is redirected to work-shared/qemux86/kernel-source/.kernel-meta/cfg/merge_config_build.log
and bbfatal_log called in do_kernel_configme only when merge_config.sh fails which it should in this case
but doesn't.
The merge_config_build.log shows 5.4 specific error:
scripts/Kconfig.include:39: gold linker 'i686-oe-linux-ld' not supported
but then it happily continues reporting what isn't applied in not created
.config file
--- tmp-glibc-bfd-qemux86-do_kernel_configme/work-shared/qemux86/kernel-source/.kernel-meta/cfg/merge_config_build.log 2020-02-05 02:28:01.455520207 +0100
+++ tmp-glibc-gold-qemux86-do_kernel_configme/work-shared/qemux86/kernel-source/.kernel-meta/cfg/merge_config_build.log 2020-02-05 02:26:30.656470909 +0100
@@ -171,586 +171,5113 @@
HOSTCC scripts/kconfig/symbol.o
HOSTLD scripts/kconfig/conf
scripts/kconfig/conf --alldefconfig Kconfig
-./.tmp.config.jDjIEYi9Yq:2208:warning: unexpected data: #
-./.tmp.config.jDjIEYi9Yq:2209:warning: unexpected data: # Generic Driver Options
-./.tmp.config.jDjIEYi9Yq:2210:warning: unexpected data: #
-./.tmp.config.jDjIEYi9Yq:3102:warning: symbol value 'm' invalid for SAMPLE_SECCOMP
-./.tmp.config.jDjIEYi9Yq:3119:warning: symbol value 'm' invalid for NF_CT_PROTO_GRE
-./.tmp.config.jDjIEYi9Yq:3120:warning: symbol value 'm' invalid for NF_CT_PROTO_SCTP
-./.tmp.config.jDjIEYi9Yq:3121:warning: symbol value 'm' invalid for NF_CT_PROTO_UDPLITE
-#
-# configuration written to .config
-#
+scripts/Kconfig.include:39: gold linker 'i686-oe-linux-ld' not supported
+make[2]: *** [/OE/build/oe-core/tmp-glibc/work-shared/qemux86/kernel-source/scripts/kconfig/Makefile:73: alldefconfig] Error 1
+make[1]: *** [/OE/build/oe-core/tmp-glibc/work-shared/qemux86/kernel-source/Makefile:567: alldefconfig] Error 2
make[1]: Leaving directory '/OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/linux-qemux86-standard-build'
+make: *** [Makefile:179: sub-make] Error 2
+grep: /OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/linux-qemux86-standard-build/.config: No such file or directory
+Value requested for CONFIG_LOCALVERSION not in final .config
+Requested value: CONFIG_LOCALVERSION=""
+Actual value:
+
+grep: /OE/build/oe-core/tmp-glibc/work/qemux86-oe-linux/linux-yocto/5.4.15-r0/linux-qemux86-standard-build/.config: No such file or directory
+Value requested for CONFIG_LOCALVERSION_AUTO not in final .config
+Requested value: # CONFIG_LOCALVERSION_AUTO is not set
+Actual value:
* I don't know why merge_config.sh uses separate more difficult to find merge_config_build.log
instead of leaving the output to end in log.do_kernel_configme, I'll send it as separate
commit so that it can be discussed there.
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Barker [Thu, 6 Feb 2020 17:35:49 +0000 (12:35 -0500)]
kernel-yocto: Only override CONFIG_LOCALVERSION if LINUX_VERSION_EXTENSION is set
CONFIG_LOCALVERSION may already be set in a defconfig or config fragment
and this should not be unconditionally overridden.
Signed-off-by: Paul Barker <pbarker@konsulko.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Barker [Thu, 6 Feb 2020 17:35:48 +0000 (12:35 -0500)]
kernel-yocto: Drop setting of unused variables in do_kernel_metadata
The machine_branch and machine_srcrev variables were set but not used in
do_kernel_metadata.
Signed-off-by: Paul Barker <pbarker@konsulko.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Barker [Thu, 6 Feb 2020 17:35:47 +0000 (12:35 -0500)]
kernel-yocto: Move defaults and tasks from linux-yocto.inc into bbclass
This allows the kernel-yocto bbclass to be inherited in a recipe without
needing to include linux-yocto.inc. The bbclass should stand on its own
and linux-yocto.inc does a few things which may not be desired in other
kernel recipes (such as modifying KERNEL_FEATURES).
The LINUX_VERSION_EXTENSION default is not moved as other kernel recipes
may not want this setting in place.
Signed-off-by: Paul Barker <pbarker@konsulko.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Martin Jansa [Thu, 6 Feb 2020 17:35:46 +0000 (12:35 -0500)]
kern-tools-native: use more common S value and oe_runmake
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bruce Ashfield [Thu, 6 Feb 2020 17:35:45 +0000 (12:35 -0500)]
kern-tools: integrate merge_config fix for gold linker
Bumping the SRCREV to pickup the following fix:
Author: Martin Jansa <Martin.Jansa@gmail.com>
Date: Wed Feb 5 03:26:57 2020 +0100
merge_config.sh: pass LD variable from shell environment to make
* since 5.4 kernel Kconfig will fail immediately when it detects
that LD points to gold linker:
scripts/Kconfig.include:39: gold linker 'i686-oe-linux-ld' not supported
* in OE we already pass bfd linker in KERNEL_LD variable to merge_config.sh
but we need to pass it also into the make call here
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 7 Feb 2020 15:09:48 +0000 (15:09 +0000)]
ncurses: Fix reproducibility issue
The build was deciding whether to rename manpages based upon the presence of
/etc/debian_version. Be explicit about the configuration instead and
ensure determinism.
[YOCTO #13781]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
