Richard Purdie [Fri, 19 Aug 2016 10:03:00 +0000 (11:03 +0100)]
parselogs: Ignore uvesafb timeouts
We're periodically seeing uvesafb timeouts on the autobuilder. Whitelist these
errors as there is little it seems we can do about them and we therefore
choose to ignore them rather than fail the builds.
[YOCTO #8245]
There is a better solution proposed in the bug with a -1 timeout however
this avoids failed builds until such times as that is implemented.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ed Bartosh [Fri, 19 Aug 2016 14:35:53 +0000 (17:35 +0300)]
image_types: use COMPRESSIONTYPES variable for backward compatibility
Recent renaming of COMPRESSIONTYPES variable can break recipes that
still use it. Including value of COMPRESSIONTYPES variable into
CONVERSIONTYPES should prevent this.
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Wed, 3 Aug 2016 13:32:15 +0000 (16:32 +0300)]
license: simple verification of LICENSE_<pkg> values
LICENSE should be a superset of all LICENSE_<pkg> values. That is,
LICENSE should contain all licenses and LICENSE_<pkg> can be used to
"filter" this on a per-package basis. LICENSE_<pkg> shouldn't contain
anything that isn't specified in LICENSE.
This patch implements simple checking of LICENSE_<pkg> values. It does
do not do advanced parsing/matching of license expressions, but,
checks that all licenses mentioned in LICENSE_<pkg> are also specified in
LICENSE. A warning is printed if problems are found.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Markus Lehtonen [Wed, 3 Aug 2016 13:32:14 +0000 (16:32 +0300)]
license.bbclass: do not process LICENSE_pn variables
The loop iterating over LICENSE_pn variables has never worked. In
addition, the LICENSE variable is supposed to contain all licenses
defined in LICENSE_pn variables. Thus, it is simpler just to use LICENSE
as the data we get is essentially the same.
[YOCTO #9499]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ed Bartosh [Wed, 24 Aug 2016 10:16:15 +0000 (13:16 +0300)]
combo-layer: python3: fix UnicodeDecodeError
check_patch function opens patch file in text mode. This causes
python3 to throw exception when calling readline():
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xa7 in position
NNNN: invalid start byte
Opening file in binary mode and using binary type instead of strings
should fix this.
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Thomas Witt [Mon, 22 Aug 2016 14:27:11 +0000 (16:27 +0200)]
cmake.bbclass: call cmake with a relative path
CMake wants a relative path for CMAKE_INSTALL_*DIR, an absolute path
breaks cross-compilation. This fact is documented in the following
ticket: https://cmake.org/Bug/view.php?id=14367
$sysconfdir and $localstatedir are not relative to $prefix, so they are
still set as absolute paths. With his change ${PROJECT}Targets.cmake
that are generated by cmakes "export" function will contain relative
paths instead of absolute ones.
Signed-off-by: Thomas Witt <Thomas.Witt@bmw.de> Signed-off-by: Clemens Lang <clemens.lang@bmw-carit.de> Signed-off-by: Ross Burton <ross.burton@intel.com>
Robert Yang [Mon, 22 Aug 2016 07:52:46 +0000 (00:52 -0700)]
python-3.5-manifest.inc: the signal module RDEPENDS on enum
Fixed:
$ python3
>>> import signal
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/path/to/sdk/sysroots/x86_64-pokysdk-linux/usr/lib/python3.5/signal.py", line 4, in <module>
from enum import IntEnum as _IntEnum
ImportError: No module named 'enum'
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Maxin B. John [Tue, 16 Aug 2016 11:12:03 +0000 (14:12 +0300)]
useradd_base: avoid unintended expansion for useradd parameters
Now, useradd dollar sign requires three prepending backslash characters to
avoid unintended expansion. It used to be just one prepending backslash
character before Krogoth. Restore that behaviour.
[YOCTO #10062]
Signed-off-by: Niko Mauno <niko.mauno@vaisala.com> Signed-off-by: Maxin B. John <maxin.john@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
André Draszik [Thu, 18 Aug 2016 16:15:00 +0000 (17:15 +0100)]
gettext_0.16.1: use musl gettext implementation
gettext uses internal symbols to detect whether the
implementation is compatible with GNU gettext. However,
these symbols don't are not part of the public API, they
are specific to glibc.
While musl implements the GNU gettext *API* version 1 and 2
http://www.openwall.com/lists/musl/2015/04/16/3
it doesn't implement glibc internals. This means that
gettext fails to detect musl's working implementation.
More recent versions of gettext have changed the way
GNU gettext compatibility is done
https://lists.gnu.org/archive/html/bug-gettext/2016-04/msg00000.html
http://git.savannah.gnu.org/cgit/gettext.git/commit/gettext-runtime/m4/gettext.m4?id=b67399b40bc5bf3165b09e6a095ec941d4b30a97
and while we could backport the corresponding patch to
gettext.m4, we avoid doing that so as to avoid any
potential GPL-v3 issues.
So instead we force ./configure to assume that the gettext
implementation of the c-library (musl) is compatible.
As a side-effect, this also reduces image sizes as the
internal gettext implementation isn't built anymore, and
it's otherwise packaged into the main gettext package
which blows up the image as the main gettext package
contains a lot of things.
Similarly, libintl.h isn't generated anymore, as the one
from musl is OK.
Signed-off-by: André Draszik <git@andred.net> Signed-off-by: Ross Burton <ross.burton@intel.com>
Hongxu Jia [Wed, 17 Aug 2016 08:57:30 +0000 (04:57 -0400)]
grub: fix load module all_video failed
While using oe-core toolchain to strip grub module 'all_video.mod',
it stripped symbol table:
--------------
root@localhost:~# objdump -t all_video.mod
all_video.mod: file format elf64-x86-64
SYMBOL TABLE:
no symbols
--------------
It caused grub to load module all_video failed.
(This module will be loaded by defalut which configed in grub.cfg)
--------------
grub> insmod all_video
error: no symbol table.
--------------
Tweak strip option to keep symbol .module_license could workaround
the issue.
--------------
root@localhost:~# objdump -t all_video.mod
2. Change intl.right correspond to the unicode3.sub's output
In sub-test unicode3.sub of intl.tests have this:
printf %q "$payload"
The payload variable was assigned by ASCII characters, when using
'%q' format strings, it means print the associated argument shell-quoted.
When the strings contain the non-alpha && non-digit && non-punctuation &&
non-ISO 646 character(7-bit), it would output like this: " $'...', ANSI-C
style quoted string. We can check the bash source code at:
Ming Liu [Wed, 17 Aug 2016 11:26:09 +0000 (13:26 +0200)]
bootchart2: fixes a BOOTLOG_DEST typo
A flaw was observed in bootchartd that BUILDLOG_DEST should actually be
BOOTLOG_DEST, this seems to be a typo or mix-up which has been fixed in
upstream.
Cherry pick the fix since bootchart2 0.14.8 is still the newest release
so far.
Signed-off-by: Ming Liu <peter.x.liu@external.atlascopco.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ovidiu Vancea [Fri, 12 Aug 2016 15:04:12 +0000 (18:04 +0300)]
initscripts: Check for logrotate in dmesg.sh
Autodetect previously hardcoded logrotate location because it can be
installed in multiple places like /usr/bin/logrotate which is very
common besides /usr/sbin
Mark Hatle [Sat, 20 Aug 2016 01:44:54 +0000 (20:44 -0500)]
glibc: Fix scope resolution in glibc to be breadth first.
The ELF specification indicates symbol resolution should be breadth first, not
depth first.
The dl-deps.c: dl_build_locale_scope function is processing in a depth first
mode. This is causes certain symbols to be incorrectly reported when
LD_TRACE_PRELINKING=1 is enabled.
See glibc BZ #20488 for more information.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bruce Ashfield [Sat, 20 Aug 2016 04:30:35 +0000 (00:30 -0400)]
linux-yocto/4.4: fix configuration warnings
Integrating the following commits to address configuration warnings for
intel-corei7-64 and intel-core2-32:
features: Fix dependencies and =m vs =y discrepancies for corei7
intel-core2-32.cfg: Explicitly disable CONFIG_64BIT
Signed-off-by: California Sullivan <california.l.sullivan@intel.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bruce Ashfield [Sat, 20 Aug 2016 04:30:34 +0000 (00:30 -0400)]
linux-yocto/4.1/4.4: -stable updates and configuration changes
Updating the 4.4 kernel to v4.4.18 and the 4.1 kernel to v4.1.30.
We also tweak the configuration with the following commits to remove
warnings being generated from the 4.4 kernel (due to options being
dropped from the final .config):
features: Create mfd-intel-lpss feature and use where appropriate
features/iio: Set IIO_BUFFER_CB to =m instead of =y
features: Add 6lowpan feature and add it where necessary
Tested on qemux86, qemuppc, qemumips and qemuarm.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Robert Yang [Fri, 19 Aug 2016 01:29:16 +0000 (18:29 -0700)]
kbd: remove PARALLEL_MAKEINST = ""
There isn't anything wrong when looked into its Makefile, I guess that
it had been fixed during ugprade, and I've applied this patch locally
for more than 2 months, there isn't anything wrong.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Awais Belal [Fri, 19 Aug 2016 11:08:13 +0000 (16:08 +0500)]
asciidoc-native: add dependency on docbook-xml-dtd4-native
During the compilation phase asciidoc runs a2x for validation
of some xmls which in turn invokes xmllint with --nonet
parameter that requires DTDs to be available locally in order
to succeed otherwise the do_compile fails.
We now add a direct dependency on docbook-xml-dtd4 so the
DTDs are always available locally.
Signed-off-by: Awais Belal <awais_belal@mentor.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Andrej Valek [Fri, 19 Aug 2016 10:51:06 +0000 (12:51 +0200)]
openssl: fix add missing dependencies building for test directory
Regarding the last commit about missing dependencies, another issue
was found. The problem was found, while ptest has been built with some
set extra settings. It means, when ptest is going to be built,
it is necessary to rebuild dependencies for test directory too.
Signed-off-by: Andrej Valek <andrej.valek@siemens.com> Signed-off-by: Pascal Bach <pascal.bach@siemens.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
André Draszik [Fri, 19 Aug 2016 10:49:02 +0000 (11:49 +0100)]
libffi: fix a typo (mips)
While code elsewhere checks for
MIPS_INSTRUCTION_SET == mips16e in order to decide how
to compile, hence the typo doesn't affect behaviour, the
intention was to set it to 'mips', as is done everywhere
else. Fixing the typo also helps to avoid confusion.
Signed-off-by: André Draszik <git@andred.net> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ed Bartosh [Fri, 19 Aug 2016 14:46:39 +0000 (17:46 +0300)]
syslinux.bbclass: ensure creation of output directory
build_syslinux_cfg function creates syslinux configuration file.
The code assumes that the output directory exists, which is not
always the case. For example rm_work task removes rootfs directory
structure and causes build_syslinux_cfg to fail with this error:
Unable to open ../<image>-<version>/syslinux_vm.cfg
Made build_syslinux_cfg depend on output directory to ensure that
directory is created before running the function.
[YOCTO #10159]
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 19 Aug 2016 15:28:17 +0000 (16:28 +0100)]
busybox: Add parallel make fix
We're seeing regular parallel make failures in applet headers in busybox.
This adds a patch to try and avoid the issue, building upon a fix already
backported from upstream. The patch has been sent to upstream.
[YOCTO #10116]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 19 Aug 2016 15:27:43 +0000 (16:27 +0100)]
sanity.bbclass: Ensure we expand BUILD_PREFIX
This likely used to work when we expanded python functions and broke when
we stopped. Since it defaults to "", it never caused an issue but
is incorrect usage so fix it.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mark Hatle [Thu, 18 Aug 2016 17:04:16 +0000 (12:04 -0500)]
local.conf.sample: Disable ARM and PPC due to prelink test case failures
Internal prelink test cases reloc8 and reloc9 are failing on both ARM
and PPC systems. Disable them by removing the prelink from the
IMAGE_CLASSES setting.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mark Hatle [Thu, 18 Aug 2016 17:04:15 +0000 (12:04 -0500)]
prelink: Move to latest version of prelink
* Uprev rtld emulation to glibc-2.23
* Fix compilation warnings
* Add additional debug scopes
* Change rtld build_local_scope to be breadth-first
* Fix LD_PRELOAD emulation
* Change function reordering to work with latest binutils
Signed-off-by: Mark Hatle <mark.hatle@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Thu, 18 Aug 2016 13:46:32 +0000 (14:46 +0100)]
libunwind: Fix build race conflict with gcc and musl
Building libunwind, then gcc-runtime causes build failures. This is hard
to fix since gcc-runtime wants the internal gcc unwind.h header but libunwind
wants to provide this. There are differences in include behaviour between gcc
and glibc which are by design.
This patch hacks around the issue by looking for a define used during gcc-runtime's
build and skipping to the internal header in that case. The patch is only enabled
on musl and is the best workaround I could come up with to unblock failing builds
on our autobuilder.
[YOCTO #10129]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 16 Aug 2016 17:00:13 +0000 (18:00 +0100)]
devtool/recipetool/meta: Adapt to bitbake API changes for multi-configuration builds
Unfortunately to implenent multiconfig support in bitbake some APIs
had to change. This updates code in OE to match the changes in bitbake.
Its mostly periperhal changes around devtool/recipetool
[Will need a bitbake version requirement bump which I'll make when merging]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
parselogs.py: Add failed to setup card detect gpio error on x86
This error has occurred on the MinnowBoard Max and Turbot since its
inception. It supposedly indicates a non-working SD card reader, but
ours works fine. Whitelist the error.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
parselogs.py: Ignore Skylake graphics firmware load errors on genericx86-64
These errors can't be fixed without adding the firmware to the initramfs
and building it into the kernel, which we don't want to do for
genericx86-64. Since graphics still work acceptably without the firmware
blobs, just ignore the errors for that MACHINE.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Patrick Ohly [Fri, 29 Jul 2016 14:58:52 +0000 (17:58 +0300)]
image.bbclass: rename COMPRESS(ION) to CONVERSION
With the enhanced functionality, the term "compression" is no longer
accurate, because the mechanism also gets used for conversion
operations that do not actually compress data.
It is possible to remove this naming problem in a backward-compatible
manner by including COMPRESSIONTYPES in CONVERSIONTYPES and checking for
the old COMPRESS_CMD/DEPENDS as fallbacks.
[YOCTO #9346]
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com> Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
grub-editenv edits the env block at runtime on a booted system. Other
tools can depend on it to configure a live system, for ex. to set next
boot mode upon reboot. By splitting grub-editenv, tools don't have to
depend on the entire grub package (grub-editenv just edits one file).
Signed-off-by: Alejandro del Castillo <alejandro.delcastillo@ni.com> Signed-off-by: Ioan-Adrian Ratiu <adrian.ratiu@ni.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
André Draszik [Tue, 16 Aug 2016 10:38:34 +0000 (11:38 +0100)]
connman: add missing space in _append
We do that everywhere else, and otherwise anybody
extending SRC_URI through bbappend must know to
add a space at the end, which is an unusual
requirement.
Signed-off-by: André Draszik <git@andred.net> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bruce Ashfield [Mon, 15 Aug 2016 18:26:59 +0000 (14:26 -0400)]
kernel-yocto: streamline patch, configuration and audit phases
We've been running with a set of kern-tools that were designed to work
with build systems that knew nothing about git, trees, commits, etc.
As such, there's been a set of shims/wrappers in place to work with
within bitbake/oe-core. These were the *me scripts: createme, updateme,
patchme and configme.
With this commit, we strip that legacy code and use the tools directly.
This means less complexity, fewer corner cases .. and no surprises
when the tools are arunning. As another benefit, the tools consume
much less time during a typical build and have no noticeable impact
on the overall build time.
Existing .scc files, features, and processing are not impacted as
these tools are compatible with existing feature descriptions and
kerne configuration fragments.
The audit of kernel configuration fragments is now detached
from the linux-yocto build structure and process. This means that
they can eventually be tweaked to offer kernel audit to any type of
kernel build and configuration process.
Additionally, the kernel symbol audit phase can now resolve symbol
dependencies and offer guidance when a symbol is missing:
WARNING: linux-yocto-4.4.15+gitAUTOINC+b030d96c7b_f5e2c49d58-r0 do_kernel_configcheck: [kernel config]: specified values did not make it into the kernel's final configuration:
Bruce Ashfield [Mon, 15 Aug 2016 18:26:58 +0000 (14:26 -0400)]
linux-yocto/4.4: -rt update patch meta-data to remove ()
The existing kernel patching scripts don't like () in patch names, since they
are detected as function calls. Although the scripts will be updated to avoid
this error, it is worthwhile fixing the patch names in the meantime.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bruce Ashfield [Mon, 15 Aug 2016 18:26:57 +0000 (14:26 -0400)]
linux-yocto/4.1: config updates
Integrating the following configuration changes:
features: usb-net: provide more coverage on USB network devices
features: broxton: enable iTCO watchdog support
features: broxton: enable iSMT support
features: broxton: enable LPC bridge function for Intel ICH and SCH
Signed-off-by: Bruce Ashfield <bruce@zedd.org> Signed-off-by: Ross Burton <ross.burton@intel.com>
There was an issue with a netfilter backport in 4.1.28-stable. To
address it, we backport the -stable fix:
netfilter: x_tables: fix stable backport
Stable-4.1 backport of mainline commit 364723410175 ("netfilter:
x_tables: validate targets of jumps") doesn't handle correctly the fact
that 4.1 kernel is missing commit 482cfc318559 ("netfilter: xtables:
avoid percpu ruleset duplication") so that t->entries is still a per-cpu
array in find_jump_target().
Use the same fix as e.g. stable-3.14 backport.
Fixes: 8163327a3a92 ("netfilter: x_tables: validate targets of jumps") Signed-off-by: Michal Kubecek <mkubecek@suse.cz> Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Richard Purdie [Wed, 17 Aug 2016 11:53:16 +0000 (12:53 +0100)]
gobject-introspection: Ensure prelink config file exists to avoid build failures
gobject-introspection relies upon prelink-rtld. In order to function correctly,
we generate an ld.so.conf file which is generated before users of prelink-rtld
are called.
There is currently a race in gobject-introspection since the configuration file
may not have been created. This adds in code to ensure that regardless of codepath
(new build, existing build, from sstate), we trigger the creation of the configuration
file and avoid build failures.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Wed, 29 Jun 2016 17:11:28 +0000 (20:11 +0300)]
oe-build-perf-test: suppress logger output when tests are being run
Prevent logger from writing to stderr when the tests are being run by
the TestRunner. During this time the logger output is only written to
the log file. This way the console output from the script is cleaner and
not mixed with possible logger records.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
