Richard Purdie [Fri, 9 May 2014 12:29:13 +0000 (13:29 +0100)]
classes/lib/oe: Fix cross/crosssdk references
With the renaming of the cross packages, its no longer possible to use
endswith("-cross") and similar to detect cross packages. Replace these
references with other techniques.
This resolves certain build from sstate failures which were due to the
system believing cross packages were target packages and therefore
dependency handling was altered.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Saul Wold [Fri, 9 May 2014 00:39:22 +0000 (17:39 -0700)]
update-rc.d: Handle multilib case for initscripts
When building multilib, we also have to add the multlib prefix otherwise
we get a WARNING:
WARNING: Multilib QA Issue: lib32-dbus package lib32-dbus - suspicious values 'initscripts-functions' in RDEPENDS
Anders Roxell [Thu, 24 Apr 2014 22:09:05 +0000 (17:09 -0500)]
rt-tests: bump version 0.87 => 0.89
cyclictest was updated to correct some problems with CPU affinity selection
and display of help and/or execution details.
These patches were accepted by the maintainer of the cyclictest utility.
The changes were reflected in an updated release of rt-tests - v0.89
The cyclictest changes were successfully tested on both an ARM v7 and x86_64
host environment.
The OE recipe was updated locally to incorporate the newer rt-tests release
and the resulting builds were verified for the same ARM v7 target
and for qemux86.
Patch files obsoleted by the rt-tests 0.89 release were removed from the recipe
deleted: meta/recipes-rt/rt-tests/files/0001-rt-tests-Allow-for-user-specified-PYLIB.patch
deleted: meta/recipes-rt/rt-tests/files/0002-rt-tests-Break-out-install_hwlatdetect.patch
deleted: meta/recipes-rt/rt-tests/files/added-missing-dependencies.patch
Signed-off-by: Anders Roxell <anders.roxell@linaro.org> Signed-off-by: Gary S. Robertson <gary.robertson@linaro.org> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Matt Cowell [Thu, 17 Apr 2014 19:23:13 +0000 (14:23 -0500)]
chrpath: properly handle rootdir with '..' in path
When there is a '..' in the rootdir path, rootdir will not be a substring of
fpath. This causes an incorrect rpath of the difference between the workdir
and the sysroot to be computed, which is incorrect. Normalizing basedir
fixes this issue.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 7 May 2014 16:05:33 +0000 (17:05 +0100)]
libxcb: disable check to avoid floating dependencies
The "check" module is checked for in configure without any way to explicitly
enable or disable it, but it isn't a build dependency of libxcb. If it's found
libxslt is also checked for, which is a build dependency but will only be used
if check is also present.
As the libxcb unit test suite is minimal at present, forcibly disable the test
for check and remove the libxslt-native build dependency.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Chen Qi [Thu, 8 May 2014 08:31:52 +0000 (16:31 +0800)]
util-linux: fix parallel build issure for PATHFILES
When compiling util-linux, it's possible to encounter the following error.
/bin/sh: line 2:: misc-utils/uuidd.8.tmp: No such file or directory
This is because that the misc-utils directory doesn't exist when trying to
write to misc-utils/uuidd.8.tmp.
When generating misc-utils/uuidd.8 (or anything in PATHFILES), its directory
may not have been created yet. So we need to ensure the existence of the
directory to avoid the compilation error.
[YOCTO #6292]
Signed-off-by: Chen Qi <Qi.Chen@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changhyeok Bae [Thu, 8 May 2014 03:44:04 +0000 (12:44 +0900)]
mirrors.bbclass: Add mirror site for savannah
* The SRC_URI is not accessible.
So need to add mirror site referred by the original site.
* The problem is that
http://download.savannah.gnu.org/releases redirects to closest mirror
and few mirrors (e.g. .jp) weren't working correctly while
http://download-mirror.savannah.gnu.org/releases/ seems to be reliable.
* Add SAVANNAH_GNU_MIRROR and SAVANNAH_NONGNU_MIRROR variable in bitbake.conf.
* Change the SRC_URI using the new variable.
Signed-off-by: Changhyeok Bae <changhyeok.bae@lge.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 6 May 2014 12:50:55 +0000 (12:50 +0000)]
attr: Fix uclibc builds
attr needs libintl headers and libs. Add in the missing dependency and
ensure the linker flag gets passed in multilib builds by replacing the
PN == BPN check with a class-target override instead.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Maxin B. John [Wed, 7 May 2014 12:24:15 +0000 (14:24 +0200)]
libxml2: fix CVE-2014-0191
It was discovered that libxml2, a library providing support to read,
modify and write XML files, incorrectly performs entity substituton in
the doctype prolog, even if the application using libxml2 disabled any
entity substitution. A remote attacker could provide a
specially-crafted XML file that, when processed, would lead to the
exhaustion of CPU and memory resources or file descriptors.
Function buildhistory_list_installed_image fails with error "Argument
list too long". This patch uses a temporal file to pass the package list
to opkg-query-helper.py
ERROR: Function failed: buildhistory_list_installed_image
ERROR: Logfile of failure stored in: /var/lib/jenkins/jobs/qt5022-cesium/workspace/build/tmp/work/qt5022-poky-linux/qimage-dev/1.0-r0/temp/log.do_rootfs.16747
NOTE: recipe qimage-dev-1.0-r0: task do_rootfs: Failed
ERROR: Task 7 (/var/lib/jenkins/jobs/qt5022-cesium/workspace/repo/yocto/../qtec/meta-qt5022/recipes-core/images/qimage-dev.bb, do_rootfs) failed with exit code '1'
NOTE: Tasks Summary: Attempted 4999 tasks of which 30 didn't need to be rerun and 1 failed.
NOTE: Writing buildhistory
Auto packing the repository for optimum performance.
Summary: 1 task failed:
/var/lib/jenkins/jobs/qt5022-cesium/workspace/repo/yocto/../qtec/meta-qt5022/recipes-core/images/qimage-dev.bb, do_rootfs
Summary: There were 74 WARNING messages shown.
Summary: There were 2 ERROR messages shown, returning a non-zero exit code.
Petter Mabacker [Tue, 6 May 2014 14:03:12 +0000 (16:03 +0200)]
x-load: remove recipe
x-load has been superseded by u-boot SPL for all cortex based TI SoCs.
For legacy boards meta-ti layer (that contains a maintained recipe)
should be used instead.
Richard Purdie [Tue, 6 May 2014 13:49:50 +0000 (14:49 +0100)]
libiconv: Fix B != S with uclibc builds
Without this, uclibc builds fail with libtool version mismatches. The issue
is that we need to remove the files in ${S}, not ${B} which is now
the default after the B != S change.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Richard Purdie [Tue, 6 May 2014 13:48:37 +0000 (14:48 +0100)]
ptest: Work from PTEST_ENABLED, not DISTRO_FEATURES
Some classes of recipe disable ptest even though its in DISTRO_FEATURES
(e.g. nativesdk). We shouldn't attempt to build ptest packages when
its disabled. This replaces some DISTRO_FEATURE checks with PTEST_ENABLED
checks instead.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Chen Qi [Tue, 6 May 2014 07:53:12 +0000 (15:53 +0800)]
mmc-utils: fix compilation failure for mips64 target.
This patch fixes mmc-utils compilation failure for qemumips64.
Remove the 'include <asm-generic/int-ll64.h>' line from mmc.h,
because this file is automatically included if _MIPS_SZLONG
is not 64, otherwise, <asm-generic/int-l64.h> is included.
Expicitly including <asm-generic/int-ll64.h> will cause the
compilation failure for mips64 target.
Maxin B. John [Tue, 6 May 2014 00:53:34 +0000 (02:53 +0200)]
openssl: fix CVE-2014-0198
A null pointer dereference bug was discovered in do_ssl3_write().
An attacker could possibly use this to cause OpenSSL to crash, resulting
in a denial of service.
Tyler Hall [Mon, 5 May 2014 00:06:43 +0000 (20:06 -0400)]
python3: Revert python-config to distutils.sysconfig
The newer sysconfig module shares some code with distutils.sysconfig,
but the same modifications as in
12-distutils-prefix-is-inside-staging-area.patch
makes distutils.sysconfig affect the native runtime as well as cross
building. Use the old, patched implementation which returns paths in
the staging directory and for the target, as appropriate.
This change reverts this upstream patch
http://hg.python.org/cpython/diff/712970b019f7/Misc/python-config.in
Signed-off-by: Tyler Hall <tylerwhall@gmail.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Tyler Hall [Sun, 4 May 2014 22:37:50 +0000 (18:37 -0400)]
python3: Substitute correct python version in shebang
If python2 and python3 are both available, scripts that are subject to
this substitution can possibly run with the wrong python version.
python3-config is one such script.
Signed-off-by: Tyler Hall <tylerwhall@gmail.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Radek Dostal [Mon, 5 May 2014 07:38:22 +0000 (09:38 +0200)]
distutils.bbclass: only modify *.py file if it contains path to be removed
Currently sed command touches every single *.py file. This modifies the
timestamp of the file. All *.pyo files will be recompiled during the first
boot, because timestamp will not match. This should be only necessary if
sed command changes the file.
Signed-off-by: Radek Dostal <radek.dostal@streamunlimited.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Now that gcc-cross, gcc-crosssdk and others have expanded variables in
their names, the code that previous just applied to cross-canadian needs
expanding to cover the other cases. Improve the conditional and also
fix a bug where the multilib prefix wasn't being added into the generated
versions.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Khem Raj [Tue, 6 May 2014 07:32:50 +0000 (00:32 -0700)]
tcmode-default.inc: Adjust for new toolchain component changes
Recently we renames the cross compilers as well as spun out
libgcc-initial from gcc-initial out. We need to also express
it in default tcmode otherwise defining GCCVERSION and/or BINUVERSION
outside OE-Core does not cover all cases
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
- Updated copyright years for radeon firmware;
- Changed copyright terms (privilege of use) for qla2xxx firmware;
- Changed Broadcom firmware name for SDIO;
- Added TI wl18xx firmware package.
Bruce Ashfield [Mon, 28 Apr 2014 16:09:33 +0000 (12:09 -0400)]
linux-yocto-custom: provide defconfig guidance
As reported by Andrea, defconfigs created with savedefconfig should specify
their expansion mode as "--alldefconfig' for custom yocto recipes. To ensure
that this is documented, we can add a comment in the skeleton recipe.
Reported-by: Andrea Adami <andrea.adami@gmail.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Bruce Ashfield [Thu, 24 Apr 2014 18:09:39 +0000 (14:09 -0400)]
kernel-yocto: quote kconfig mode check
We allow inheriting recipes to control the kconfig mode used by merge_config.sh
via the KCONFIG_MODE variable. An error crept into the variable reference, and
since it is not quoted, the true condition always runs.
The result is that operations without an explicit kconfig mode cannot trigger
allnoconfig for defconfig builds, which can result in some options being
dropped from the final .config.
Quoting the reference allows it to evaluate properly.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
gcc 4.9 now generates slim objects when LTO is used
These slim objects only contain intermediate language
representation for LTO. Use -ffat-lto-objects to create files which
contain additionally the object code. When linking some .so file, the option
'-flto' is missed, so the error happens.
We add '-ffat-lto-objects' flag to make gcc 4.9 behaves the same as gcc 4.8.
Richard Purdie [Sun, 4 May 2014 16:36:10 +0000 (17:36 +0100)]
sstatetests: Use TUNE_ARCH instead of TARGET_ARCH
The scripts extact variable values from bitbake -e. Unfortunately
TARGET_ARCH is unset from that environment so we use TUNE_ARCH instead
which will have the value we need.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tudor Florea [Fri, 2 May 2014 22:21:23 +0000 (00:21 +0200)]
curl: Backport a fix for a build issue
mkhelp: generate code for --disable-manual as well
This allows configure --disable-manual to run and build without having
to regenerate the src/tool_hugehelp.c file which otherwise is necessary
since we ship tarballs with that file present.
Signed-off-by: Tudor Florea <tudor.florea@enea.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sat, 3 May 2014 10:14:47 +0000 (11:14 +0100)]
gcc-common: Ensure checksums don't change to match old behaviour
There is a fix about to go into bitbake to ensure that datastores
being accessed with a name other than "d" are correctly reflected
in checksums. This will cause this function to add in a number of
dependencies we don't want.
These do need to be properly unravelled in due course but would
only really affect multilib builds. For now therefore just exclude
the variables as per the old behaviour.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Max Eliaser [Thu, 1 May 2014 20:00:28 +0000 (13:00 -0700)]
Add texinfo.bbclass; recipes that use texinfo utils at build-time inherit it.
The class itself currently does nothing. The idea is to mark all recipes that
make use of the texinfo utilities. In the future, this class could be used to
suppress the generation/formatting of documentation for performance,
explicitly track dependencies on these utilities, and eliminate Yocto's
current dependency on the host system's texinfo utilities.
Signed-off-by: Max Eliaser <max.eliaser@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Thu, 1 May 2014 15:40:19 +0000 (16:40 +0100)]
gcc-common: Only apply fpu settings to target gcc
Within the OE build environment, we supply the correct fpu settings. These
only need to be spelt out for the on-target gcc.
Doing this means the checksums for the core compiler don't depend on the fpu
settings. We exclude the compiler tunes for similar reasons, it doesn't need
to influence the compiler build.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Thu, 1 May 2014 14:41:12 +0000 (15:41 +0100)]
cross: Update PKG_CONFIG variables after recent changes
The PKG_CONFIG definitions should now more closely match those from native.bbclass.
We can't use ${libdir} here since it contains the target arch prefix.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Otavio Salvador [Thu, 1 May 2014 19:38:04 +0000 (16:38 -0300)]
sanity: Bump BitBake minimal version to 1.23.0
The metadata started using the bb.utils.contains_any which has been
add in the BitBake 1.23.0 release, set this as the minimal version to
ensure parsing compatibility.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Wed, 30 Apr 2014 12:32:04 +0000 (13:32 +0100)]
oeqa: add proper handling for command errors where needed
For use outside of tests themselves, we want a better error than
AssertionError, so create one and allow us to request it when calling
runCmd(). This enables us to avoid tracebacks during master image
operations if the power control command fails.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Wed, 30 Apr 2014 12:32:03 +0000 (13:32 +0100)]
scripts/contrib/dialog-power-control: add a trivial power prompt script
If you want to do automated hardware testing but don't have a
controllable power strip this script can be useful so that you know when
you need to cycle the power.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Wed, 30 Apr 2014 12:32:02 +0000 (13:32 +0100)]
scripts/contrib/serdevtry: add script to handle transient serial terminals
When running automated tests (or just generally interacting with)
boards whose serial console devices are on the board itself and thus
disappear when powered down or practically disconnected, such as the
BeagleBone white, some terminal programs (e.g. picocom) will exit when
the device disappears and need to be restarted after the serial device
returns. This script handles this automatically for such terminal
programs.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Wed, 30 Apr 2014 12:32:01 +0000 (13:32 +0100)]
classes/testimage: if start fails, don't try to stop
If we couldn't start the target, it doesn't make sense to try and stop
it here since logically it shouldn't now be in any kind of "started"
state. (It's the start function's job to clean up after itself if it
fails - to that end, fix up the QemuTarget class so that it does.)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
