It is a busybox upstream known bug. When the busybox sed sub-command 'n'
hit the files EOF, it print an extra character that have been printed, but
the GNU sed would not print it.
In busybox source code ../editors/sed.c
------------------------------------------------------------------------
case 'n':
if (!G.be_quiet)
sed_puts(pattern_space, last_gets_char);
if (next_line) {
free(pattern_space);
pattern_space = next_line;
last_gets_char = next_gets_char;
next_line = get_next_line(&next_gets_char, &last_puts_char, last_gets_char);
substituted = 0;
linenum++;
break;
}
/* fall through */
/* Quit. End of script, end of input. */
case 'q':
/* Exit the outer while loop */
free(next_line);
next_line = NULL;
goto discard_commands;
------------------------------------------------------------------------
when read at the end of the file, the 'next_line' is null, it would go
"case 'q'" and goto discard_commands, the discard_commands would print
the old pattern space which have been printed.
So in order to comply with GNU sed, in case 'n', when the next_line is null
I add "else" at the end of the second "if": "goto again;" and send it to
the busybox upstream, the busybox maintainer adopt it and make a little
changes to the patch, we can see it at:
kernel.bbclass: include signing keys when copying files required for module builds
The absence of signing_key.* in $kerneldir made signing of
out-of-tree kernel modules fail (silently). Add copying of these
files during the shared_workdir task.
Signed-off-by: Mattias Waldo <mattias.waldo@saabgroup.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Wed, 5 Aug 2015 14:48:00 +0000 (15:48 +0100)]
scripts: ensure tinfoil is shut down correctly
We should always shut down tinfoil when we're finished with it, either
by explicitly calling the shutdown() method or by using it as a
context manager ("with ...").
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Juro Bystricky [Mon, 29 Aug 2016 22:45:36 +0000 (15:45 -0700)]
gcc-runtime.inc: add CPP support for mips64-n32 tune
This patch fixes the problem where the CPP compiler cannot find include files.
The compiler is configured to look for the files in places that do not exist.
When querying the CPP for search paths, we observe messages such as these:
Ed Bartosh [Wed, 31 Aug 2016 11:02:51 +0000 (14:02 +0300)]
sstate.bbclass: skip packaging if SSTATE_SKIP_CREATION is set
SSTATE_SKIP_CREATION variable will be used to skip creation of
sstate .tgz files. It makes sense for image creation tasks as
tarring images and keeping them in sstate would consume a lot of
disk space.
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mike Looijmans [Wed, 31 Aug 2016 13:22:51 +0000 (15:22 +0200)]
initscripts: Start devpts at 06 instead of 38
For example bootlogd needs devpts to be running, but bootlogd starts
at 07. Starting bootlogd early makes perfect sense, so the best option
here is to move devpts up to 06 to prevent this error message at boot:
cannot allocate pseudo tty: No such file or directory
Systems that have CONFIG_LEGACY_PTYS in the kernel will not see this
message. Since it is called "LEGACY" for a reason, fixing this in
userspace appears to be the better option here.
The devpts script does not need anything except a mounted "/dev" which
has been arranged in "S02sysfs.sh" already.
Signed-off-by: Mike Looijmans <mike.looijmans@topic.nl> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Wed, 31 Aug 2016 01:48:06 +0000 (13:48 +1200)]
scripts/contrib: update scripts for changes to internal API
The multiconfig changes altered some of the functions being called here,
so update the calls. Make use of the new Tinfoil.parse_recipe_file()
function to make parsing easier.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jackie Huang [Tue, 30 Aug 2016 09:03:22 +0000 (17:03 +0800)]
meta-ide-support: inherit nopackages
The recipe is to generate an environment script in
do_populate_ide_support for using an IDE and it
doesn't generate packages at all, so inherit nopackages
Signed-off-by: Jackie Huang <jackie.huang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Nicolas Dechesne [Mon, 29 Aug 2016 09:40:35 +0000 (10:40 +0100)]
gstreamer1.0-plugins-bad: add packageconfig for egl
In commit 9c3a94aea1d (gstreamer1.0-plugins-bad: Move EGL requirement for
Wayland), --enable-egl was explicitely added to the wayland packageconfig. While
this is correct that enabling wayland requires egl, it should be possible to
enable egl without wayland, even when using X11. For example, glimagesink can be
used for GPU based color conversion using EGL/GLES.
As such, let's make egl and wayland two separate PACKAGECONFIG flags.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Mon, 29 Aug 2016 08:40:38 +0000 (20:40 +1200)]
lib/oe/patch: commit with a dummy user/email when PATCHTOOL=git
When using PATCHTOOL = "git", the user of the system is not really the
committer - it's the build system itself. Thus, specify "dummy" values
for username and email instead of using the user's configured values.
Various parts of the devtool code that need to make commits have also
been updated to use the same logic.
This allows PATCHTOOL = "git" and devtool to be used on systems where
git user.name / user.email has not been set (on versions of git where
it doesn't default a value under this circumstance).
If you want to return to the old behaviour where the externally
configured user name / email are used, set the following in your
local.conf:
Paul Eggleton [Mon, 29 Aug 2016 07:27:59 +0000 (19:27 +1200)]
oe-selftest: devtool: fix test after recent change
OE-Core commit d3057cba0b01484712fcee3c52373c143608a436 fixed handling
of wildcard bbappends, which means that this test's expectations about
the bbappend file name are no longer met. devtool finish is meant to use
wildcard bbappends so fix the test accordingly.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
glibc-scripts: add RDEPENDS on libsotruss package required by sotruss script
It solves below error observed on qemux86 target:
root@qemux86:~# sotruss ./hello
ERROR: ld.so: object '/usr/$LIB/audit/sotruss-lib.so' cannot be loaded as audit
interface: cannot open shared object file; ignored.
Hello World
root@qemux86:~#
With this change, we get:
root@qemux86:~# sotruss ./hello
hello -> libc.so.6 :*__libc_start_main(0x8048300, 0x1,
0xbfc86274)
hello -> libc.so.6 :*puts(0x804851c, 0xb74af000, 0x0)
Hello World
root@qemux86:~#
Signed-off-by: Jagadeesh Krishnanjanappa <jkrishnanjanappa@mvista.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
oe.path: preserve xattr in copytree() and copyhardlinktree()
Pass appropriate options to tar invocations in copytree() and
copyhardlinktree() to ensure that any extended attributes on the files
are preserved during the copy.
We have to drop the use cpio in "Copy-pass" mode in copyhardlinktree()
because cpio doesn't support extended attributes on files. Instead we
revert back to using cp with different patterns depending on whether
or not the directory contains dot files.
Signed-off-by: Joshua Lock <joshua.g.lock@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
One motivation for the use of cpio in oe.path.copytree() was to
ensure that files with spaces in their names were copied. Add a new
unittest module to test the OE module with a test case for copytree
with a spaces in a filename.
Signed-off-by: Joshua Lock <joshua.g.lock@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mingli Yu [Wed, 31 Aug 2016 02:10:25 +0000 (10:10 +0800)]
ltp: remove useless script STPfailure_report.pl
* Remove useless script STPfailure_report.pl to
avoid confusing about this script fails to run
as it lacks dependency on some perl module such
as LWP::Simple
- The script STPfailure_report.pl previously is
added as a tool to analyze failures from LTP
runs on the OSDL's Scaleable Test Platform (STP) as below:
Added tool for analyzing failures from LTP runs on
the OSDL's Scaleable Test Platform (STP)
- And the script STPfailure_report.pl mainly accesses
http://khack.osdl.org to retrieve ltp test results
run on OSDL's Scaleable Test Platform (STP) and prints
the reports, and now the website http://khack.osdl.org
not accessible, so the script is useless and drop it
and not ship it on target system
Signed-off-by: Mingli Yu <Mingli.Yu@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Mon, 29 Aug 2016 19:48:28 +0000 (22:48 +0300)]
oeqa.buildperf: convert buildstats into json format
Instead of archiving buildstats in raw text file format convert all
buildstats into one json-formatted file. Some redundant information,
i.e. 'Event:', 'utime:', 'stime:', 'cutime:' and 'cstime:' fields, are
dropped.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Mon, 29 Aug 2016 19:48:26 +0000 (22:48 +0300)]
oeqa.buildperf: don't use Gnu time
Use Python standard library functionality instead of the time utility
for measuring elapsed (wall clock) time of commands. The time.* log
files are also ditched. However, the same detailed resource usage data,
previously found in time.* logs is now provided in results.json file.
This data is collected through the resource module of Python.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Mon, 29 Aug 2016 19:48:25 +0000 (22:48 +0300)]
oeqa.buildperf: rename buildstats directories
Change directory name from 'buildstats-<test_name>' to just
'buildstats'. However, this patch adds the possibility to label
buildstats directory name with a postfix which makes it possible to save
multiple buildstats per test, for example.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Mon, 29 Aug 2016 19:48:22 +0000 (22:48 +0300)]
oe-build-perf-test: rename log file and implement --log-file
Rename the (main) log file of the oe-build-perf-test script from
'output.log' to 'oe-build-perf-test.log'. Also, add a new command line
option --log-file which makes it possible to use an alternative log file
name/path, if needed. Note that the file name/path is relative to the
output directory.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Fri, 26 Aug 2016 07:33:11 +0000 (10:33 +0300)]
oe-build-perf-test: update globalres and git even if tests failed
Write globalres log file and commit results to Git even if some tests
failed. Now that tests do not depend on each other there should be no
risk of bogus results caused by test failures.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Fri, 26 Aug 2016 07:33:10 +0000 (10:33 +0300)]
oeqa.buildperf: treat failed measurements as errors
Now failed measurements correctly cause a test failure (recorded as an
error). There should be no need to continue the test if one step fails,
especially now that the tests don't depend on each other.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Fri, 26 Aug 2016 07:33:09 +0000 (10:33 +0300)]
oeqa.buildperf: make tests independent
Add test set-up functionality so that the individual tests do not depend
on each other. This should make sure that a failure in one test does not
affect the results of another test. The patch also makes it reasonable
to run only a subset of the tests by using the --run-tests option.
The increase in total execution time of the full suite - caused by the
additional set-up steps - is insignificant because normally no
additional tasks need to be run. The previous test has already done all
set-up work.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 31 Aug 2016 10:31:29 +0000 (11:31 +0100)]
lttng-ust: Update 2.7.1 -> 2.8.1
Drop aarch64_be patch which is now upstream.
Update doc patch to apply to latest version.
Disable man generation in configure options to match docs patch (for now).
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Martin Jansa [Tue, 30 Aug 2016 14:19:58 +0000 (16:19 +0200)]
base, autotools: Append PACKAGECONFIG_CONFARGS to EXTRA_OECONF only in autotools.bbclass
* recipes which don't inherit autotools or cmake bbclass and want to
use the configure options from PACKAGECONFIG need to handle
PACKAGECONFIG_CONFARGS themselves.
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Note, the libnl-genl2 is introduced by REPLACES_${PN}-genl = "libnl-genl2".
So that we don't need set libnl-genl-3-200 in the RREPLACES and
RCONFLICTS, otherwise it would cause do_rootfs errors when install both
libnl-genl.rpm and lib32-libnl-genl.rpm:
Computing transaction...error: Can't install libnl-genl-3-200-1:3.2.28-r0.0@core2_64: conflicted package libnl-genl-3-200-1:3.2.28-r0.0@lib32_x86 is locked
We didn't meet this error before was because there was no libnl-genl.rpm,
but libnl-3-genl.rpm, and it doesn't provide libnl-genl-3-200 by default.
Remove libnl-genl-3-200 from RREPLACES and RCONFLICTS will fix the problem.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
André Draszik [Fri, 26 Aug 2016 10:31:50 +0000 (11:31 +0100)]
libnl: fix packaging mistakes
- *.la files belong into -dev packages
- the genl-ctrl-list command line utility should go to into the CLI
package, so as to prevent the libnl-genl library package from
pulling in all of the command line utilities (as genl-ctrl-list
is linked against libnl-cli-3.so.200)
Signed-off-by: André Draszik <git@andred.net> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mikko Ylinen [Tue, 30 Aug 2016 05:38:45 +0000 (08:38 +0300)]
image_types: check COMPRESS_DEPENDS for backwards compatibility
To complete the transition/renaming to chained image type CONVERSION
while maintaining bacwards compatibility to COMPRESS(ION), make sure also
COMPRESS_DEPENDS is checked. Without this, the dependencies for legacy
COMPRESSIONTYPES do not get built.
Signed-off-by: Mikko Ylinen <mikko.ylinen@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
init-install: Fixes the install script failing when not finding any mmcblk devices
The init-install.sh and init-install-efi.sh scripts perform a check
to see which devices are available on a booted system for installation.
Recently, the way we check for these devices changed on 993bfb,
greping for devices found on /sys/block/, this change caused the installer
to fail (at least) when not finding any mmcblk devices, due to the fact
that we call sh -e to execute this script, so any command (grep)
or pipeline exiting with a non-zero status causes the whole script to exit
This patch throws in a harmless true exit status at the end of the pipeline(s)
of the grep commands to avoid the installer script from exiting, fixing the issue.
[YOCTO #10189]
Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 24 Aug 2016 21:50:57 +0000 (22:50 +0100)]
linux-firmware: set a preferred provider for brcmfmac-sdio.bin
This recipe packages six alternatives to brcmfmac-sdio.bin but as they all have
equal priority there is no determinism on what provider will be used if they are
all installed.
Arbitrarily select 4339 to be the highest priority.
Signed-off-by: Ross Burton <ross.burton@intel.com>
