| Age | Commit message (Collapse) | Author |
|---|
|
As Chris Larson points out in the bug, "BUILD_ARCH changing doesn't change
native/cross sstate checksums, because its calculated in ${@}, so just its
unexpanded form is in the signature. But BUILD_ARCH ends up included in the
sstate filenames, so changes to BUILD_ARCH will result in rebuilds of
native/cross but not rebuilds of the target recipes which depend on
them, which is just what we want."
However it does mean we can't easily test 32 and 64 bit signatures. In order
to make the tests work, we need to add BUILD_ARCH to the HASHBASE whitelist.
BUILD_ARCH is used in the workdir paths and so on so changing it does still
rebuild when we need it to.
With improvements to function dependency tracking, a dependency on SSTATE_PKGARCH
was also introduced causing problems for 32/64 bit build signature equivalence.
Since this is reflected in the sstate filenames, we can safely whitelist this too.
[YOCTO #5970]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If -D_FORTIFY_SOURCE=2 is included in CFLAGS for debug builds,
many warnings will be generated and some packages will fail to
build. So, only conditionally include it.
Signed-off-by: Joe Slater <jslater@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Use i686 as TARGET_ARCH for 32bit core2 (and corei7 and atom) builds.
In most cases, i586 and i686 are equivalent values for TARGET_ARCH, however
one important exception is glibc. When configured for i686, glibc enables
optimised string functions (SSE, SSE2, etc), which are not used when
building for i586.
The benefits of i686 optimised string functions vary depending on the
application and the CPU, however in some cases the improvements are
significant. In one test, a 50% increase in FPS was seen when running the
'smashcat' benchmark [1] in a qtwebkit browser on an Intel Atom based SoC.
The gain seems to comes from a 3x improvement in memcpy performance when
copying graphics buffer lines (5120 bytes, or 1280 x 4 bytes/pixel), from
the CPU to GPU. Note that very large memcpy's (e.g. 32MB) on the same
machine show no particular performance increase between i586 and i686.
[1] http://www.smashcat.org/av/canvas_test/
Warning: The change in TARGET_ARCH means that _i586 architecture specific
over-rides will no longer take effect. Both oe-core and meta-oe have been
updated to replace _i586 over-rides with _x86, however other layers may
still need review and updating.
Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
* my previous thumb related commit:
commit 3e760031f91fb87c3e2f62b77a117eb41164f259
Author: Martin Jansa <martin.jansa@gmail.com>
Date: Wed Feb 18 15:40:35 2015 +0100
feature-arm-thumb.inc: respect ARM_INSTRUCTION_SET when adding thumb
suffix
unfortunately removed conditional on "thumb" in TUNE_FEATURES, when
setting ARMPKGSFX_THUMB
* in case we have MACHINE without "thumb" in TUNE_FEATURES and distro
setting ARM_INSTRUCTION_SET to "thumb" we end with:
ARM_INSTRUCTION_SET="thumb"
ARM_THUMB_OPT="thumb"
ARM_M_OPT="thumb"
# TUNE_CCARGS correctly not adding -mthumb
TUNE_CCARGS=" -march=armv7-a -mthumb-interwork -mfloat-abi=softfp -mfpu=neon"
# but ARMPKGSFX_THUMB and TUNE_PKGARCH including "t2":
ARMPKGSFX_THUMB="t2"
TUNE_PKGARCH="armv7at2-vfp-neon"
# causing following error:
Error, the PACKAGE_ARCHS variable does not contain TUNE_PKGARCH (armv7at2-vfp-neon).
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
This tune file is needed to enable a GAS option specific to this cpu family
in order to disable the usage of lock prefix instructions.
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Rather than just use d.getVar(X), use the more explict d.getVar(X, False)
since at some point in the future, having the default of expansion would
be nice. This is the first step towards that.
This patch was mostly made using the command:
sed -e 's:\(getVar([^,()]*\)\s*):\1, False):g' -i `grep -ril getVar *`
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The following over-rides were both defined twice:
SECURITY_CFLAGS_pn-grub-efi-x86-64-native
SECURITY_CFLAGS_pn-ltp
Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Add license file XFree86-1.0 which is needed by packages such as hwdata.
It is from:
http://www.x.org/archive/X11R7.5/doc/LICENSE.html#AEN28
And add 'XFree86-1.0' to SRC_DISTRIBUTE_LICENSES.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Changes due to new combined support for both vmdk and vdi files.
Signed-off-by: Juro Bystricky <juro.bystricky@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
* Increase OELAYOUT_ABI: this is required because the
xf86-video-modesetting package moves from its own recipe to
xserver-xorg and sanity check cannot handle that currently.
The upgrade will delete old xf86-video-modesetting files from
sysroots.
* Remove upstreamed xserver-xorg patches
* Remove xf86-video-modesetting recipe: the driver is now included in
xserver-xorg recipe, which now produces the xf86-video-modesetting
package. The package version changes from 0.9 to 1.17.1
* Update xserver-xorg license checksum: modesetting license
info (another MIT one) has been added to the file
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
|
|
Otherwise, if MACHINEOVERRIDES is expanded before SOC_FAMILY is set
(which may happen as MACHINEOVERRIDES is included in OVERRIDES) we can
see:
ExpansionError: Failure expanding variable MACHINEOVERRIDES, expression was
${@['', '${SOC_FAMILY}:']['${SOC_FAMILY}' != '']}p1022ds
which triggered exception SyntaxError: EOL while scanning string literal (MACHINEOVERRIDES, line 1)
To avoid this, give SOC_FAMILY a default empty value so it doesn't
get read as None.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
It was pointed out that people couldn't easily see who used this or
why so add some comments about that.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Use immediate expansion for BUILD_ARCH and BUILD_OS since there is no
point in repeatedly calling os.uname() throughout parsing. This is
worth around 2% of parsing time, small but measurable.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Instead of hard-coding the set of features that can be considered in
COMBINED_FEATURES, simply generate the intersection of DISTRO_FEATURES and
MACHINE_FEATURES.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
PRINC is no longer needed since we have the PR service instead
which people should be using. Drop the PRINC code.
This has been deprecated with a warning and then an error for quite some
time so we can remove it now.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
With poky-lsb (security flags enabled), python-numpy doesn't build
with pie flags.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If bluez4 is not explicitly passed as a
distro feature, bluez5 should be used as
the default, so backfill it.
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
There are several entries here which are not needed with the
modern license handling code:
gcc-source - moved to direct handling in base.bbclass
(due to version appended to the name)
libgcc - Listed as GPLv3 exception for its packages
libgcc-initial - Listed as GPLv3 exception
gcc-runtime - Indivisual packages listed as GPLv3 exception where appropriate
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This means you can have one gcc version for some gcc recipes
(e.g. crosssdk/nativesdk) and another gcc version for target code.
Also remove the preferred version entry from the default toolchains
list since the version issue is now handled automatically.
We also need to specifically handle gcc-source in the license handling
code since expanding ${PV} in the base class isn't possible. Since
gcc-source doesn't generate any packages directly this shouldn't be
an issue and whitelisting in this way is easiest (and matches the
rest of the toolchain handling).
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
It supports it, so lets enable it even though
we can compile lot of stuff without NLS but its
harder to get them all compiled without NLS
Change-Id: I49a06c05b004654dabbef980c4e6ad991d581341
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
There is no reason to build sed for the host, however now:
ERROR: Nothing RPROVIDES 'sed-native' (but virtual:native:/OE/sources/openembedded-core/meta/recipes-extended/groff/groff_1.22.2.bb RDEPENDS on or otherwise requires it)
Signed-off-by: Dmitry Eremin-Solenikov <dmitry_eremin@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The error messages look like this:
R_ARM_TLS_LE32 relocation not permitted in shared object
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Change-Id: I034a3bf7b9288df4579744f242e990f213193e85
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
* Remove the following patches since they are already in the new code:
binutils/fix-pr15815.patch
binutils/fix-pr16428.patch
binutils/fix-pr16476.patch
binutils/fix-pr2404.patch
binutils/replace_macros_with_static_inline.patch
binutils/binutils_CVE-2014-8484.patch
binutils/binutils_CVE-2014-8485.patch
binutils/binutils_CVE-2014-8501.patch
binutils/binutils_CVE-2014-8502.patch
binutils/binutils_CVE-2014-8502_1.patch
binutils/binutils_CVE-2014-8503.patch
binutils/binutils_CVE-2014-8504.patch
binutils/binutils_CVE-2014-8737.patch
* The file src-release is gone.
* Updated patches for the new code.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Conflicts:
meta/conf/distro/include/tcmode-default.inc
|
|
We update the toasterconf.json for the 1.8 "fido"
release.
A small update to the release help text in the
toasterconf.json file shipped with the
openembedded-core layer.
We now make explicit that Toaster will build with
the tip of the selected branch, and we add links
to the OpenEmbedded repository.
Signed-off-by: Belen Barros Pena <belen.barros.pena@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Sometimes we need to override the gdb from third party
tool SDKs, this helps out, moreover it also makes it consistent
with in OE-Core too to have multiple versions of gdb if we ever
needed to
Change-Id: Ibe1ae59175984bbc661c243764c81cd99fef54d1
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
The past few weeks ftp://ftp.gnu.org has been intermittently giving errors like this:
WARNING: Failed to fetch URL ftp://ftp.gnu.org/gnu/emacs/emacs-23.4.tar.gz;name=tarball, attempting MIRRORS if available
ERROR: Fetcher failure: Fetch command failed with exit code 4, output:
Cannot parse PASV response.
accept: Connection timed out
Cannot parse PASV response.
Error in server response, closing control connection.
Which is annoying because binutils lives there. Using http://ftp.gnu.org hasn't given any problems so far.
Signed-off-by: Koen Kooi <koen.kooi@linaro.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The error-report tooling command-line options were changed and anyone doing
automated builds and using error-reports needs to know what options to parse.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
These recipes both fail to build with "relocation R_X86_64_PC32 against
undefined hidden symbol `__init_array_start' can not be used when making a
shared object" when using PIE.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
other DISTRO values
If you create a build directory with poky, or set DISTRO to "poky" and
run bitbake, you'll get a bblayers.conf file with LCONF_VERSION set to
6. If you then set DISTRO to any other value where the new distro config
doesn't pull in poky's config, e.g. "nodistro", you would then get a
bblayers.conf version error which didn't immediately make sense. (The
layer versions have been out-of-step ever since meta-yocto-bsp was
split out of meta-yocto several years ago).
This is just painful and we'd rather users didn't have to deal with
it. Obviously it isn't an OE-Core problem per se, but a simple way to
resolve it for everyone is to bump OE-Core's version to 6 with an
automatic no-op upgrade. Also ensure that multiple upgrade functions
(such as the poky one) have a chance to execute by not breaking out of
the loop as we were before.
Fixes [YOCTO #6139].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
- Based on Paul Eggletons work to partially upgrade to Python 2.7.6
Modified:
default-versions.inc: switched to python 2.7.9
generate-manifest-2.7.py: fixed _sysconfigdata
python-2.7-manifest.inc: fixed _sysconfigdata
python.inc: Updated checksums and source, no LICENSE
change just updated some dates
python-native_2.7.3 -> python-native_2.7.9 and updated patches
python_2.7.3 -> python_2.7.9, and added ac_cv_file__dev_ptmx=no
ac_cv_file__dev_ptc=no in EXTRA_OECONF to solve python
issue #3754, only needed when cross compiling, also updated patches
use_sysroot_ncurses_instead_of_host.patch: New patch to use ncursesw
from sysroot instead of hosts, introduced by fix for python issue #15268
Rebased:
01-use-proper-tools-for-cross-build.patch
03-fix-tkinter-detection.patch
05-enable-ctypes-cross-build.patch
06-avoid_usr_lib_termcap_path_in_linking.patch
avoid_warning_about_tkinter.patch
builddir.patch
fix_for_using_different_libdir.patch
host_include_contamination.patch
multilib.patch
nohostlibs.patch
search_db_h_in_inc_dirs_and_avoid_warning.patch
Deleted (fixed on upstream):
06-ctypes-libffi-fix-configure.patch
CVE-2013-4073_py27.patch
gcc-4.8-fix-configure-Wformat.patch
json-flaw-fix.patch
posix_close.patch
pypirc-secure.patch
python-2.7.3-CVE-2012-2135.patch
python-2.7.3-CVE-2013-1752-smtplib-fix.patch
python-2.7.3-CVE-2014-1912.patch
python-2.7.3-CVE-2014-7185.patch
python-2.7.3-berkeley-db-5.3.patch
python-fix-build-error-with-Readline-6.3.patch
remove-BOM-insection-code.patch
remove_sqlite_rpath.patch
python2.7.3-nossl3.patch
[YOCTO #7059]
Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com>
|
|
There is no good reason not to use ext4 at this point, it has advantages
and few drawbacks. Therefore switch the qemu machines over (and the default
runqemu script options).
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Add a variable that allows us to map the OE-Core layer to the name that
represents it in the OpenEmbedded layer index. This will be used by
bitbake-layers layerindex-fetch so that it knows that for example
OE-Core is already fetched and included in the current configuration.
Note - it won't be required for us to go around and set this for every
layer - we can do the matching based on repo and subdirectory, but
OE-Core is somewhat special in that it does sometimes appear in
different places, for example in poky.
[YOCTO #5348]
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Add BBLAYERS_LAYERINDEX_URL variable that bitbake-layers can use to find layer index.
Add BBLAYERS_FETCH_DIR variable that bitbake-layers can use to specify fetch directory.
[YOCTO #5348]
Signed-off-by: Chong Lu <Chong.Lu@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Lets try to use musl version of libintl and see how far we get
Change-Id: I0259049eb6461ab9b579c7f3ad401fefede24621
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
* each DEFAULTTUNE with thumb enabled should list it's arm variants in
PACKAGE_EXTRA_ARCHS, otherwise packages which force arm ISA won't be
found in do_rootfs
* armv7athf-neon-vfpv4 was missing its own PACKAGE_ARCH and also the arm
variant
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
* this means that recipes with ARM_INSTRUCTION_SET explicitly changed
to arm will be built in feed without thumb suffix, the same does apply
for workdir, e.g. after "bitbake glib-2.0" you can see:
tmp-glibc/work/armv5e-oe-linux-gnueabi:
glib-2.0 glibc glibc-initial
tmp-glibc/work/armv5te-oe-linux-gnueabi:
acl db gdk-pixbuf kmod ....
and
tmp-glibc/deploy/ipk:
all armv5e armv5te qemuarm
* feed config should be ok, because all default DEFAULTTUNEs always
include "arm" variants of all supported PACKAGE_ARCHs
* for more details see
http://lists.openembedded.org/pipermail/openembedded-core/2014-April/091960.html
the toolchain path issues were resolved in 1.8
* add ARM_INSTRUCTION_SET = "arm" to glibc-collateral.inc and comment in
glibc.inc to fix glibc-locale and glibc-scripts build
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Updating the libc-headers now that the kernel has released.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Drop CVE backports and ppc/fpu detection patch which is not needed anymore
Forward port eglibc option groups patch
Default to using glibc 2.21
Additional patches needed to appease -Werror option
Change-Id: I1873097cec8387ea9e8186a255122938fc28c976
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
bitbake will currently 'selecting bluez4 to satisfy runtime
libasound-module-bluez due to PREFERRED_PROVIDER_bluez4 = bluez4'
which in the case of bluez5 isn't correct.
This slightly unusual construct avoids this.
Ultimately this is a bitbake issue that needs fixing in
a better way but this means we can merge the bluez5 changes
until bitbake gets fixed.
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If BlueZ5 is added to a build as a replacement for BlueZ4,
the provider for bluez-hcidump will be bluez5.
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Update the config section of the meta/conf/toasterconf.json file
to set the default project variables as in meta-yocto/conf/toasterconf.json
[YOCTO #7248]
Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
It turns out that changing WARN_QA and ERROR_QA results in do_configure's QA
postfunc re-executing, so changing a QA test results in a complete rebuild.
This is just too much and the lesser evil of needing to do a full rebuild to
verify changed QA flags is preferable to an enforced full rebuild.
This reverts commit daecfc3438122b5d146a59a5053e57006d55ccc4.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The name 'opkg-arch-config' is much more descriptive.
Signed-off-by: Paul Barker <paul@paulbarker.me.uk>
|
|
Changing WARN_QA and ERROR_QA should cause do_package_qa to re-execute, so
removing them from the sstate hashes is harmful.
They were added back when sanity testing was part of packaging and this was the
lesser evil, compared to changing sanity tests causing a re-package of
everything.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Disable PIE in expect as otherwise it tries to link the shared library as an
executable.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
In meta/conf/bitbake.conf, PKGDATA_DIR is default to:
PKGDATA_DIR = "${STAGING_DIR_HOST}/pkgdata"
But in meta/conf/multilib.conf, PKGDATA_DIR is set as:
PKGDATA_DIR = "${STAGING_DIR}/${MACHINE}/pkgdata"
When multilib enabled, linux-libc-headers cache will be machine
specific:
$ bitbake-diffsigs sstate-cache/1a/sstate:linux-libc-headers:ppce6500-poky-linux:3.17.7:r0:ppce6500:3:1a0c3934d91479fd7242a5b1d407d155_package.tgz.siginfo sstate-cache/28/sstate:linux-libc-headers:ppce6500-poky-linux:3.17.7:r0:ppce6500:3:28c918e8f9f4a4cfceb3a38b258f7501_package.tgz.siginfo
basehash changed from 8d3158bbddcee612fa30badd05f47b8e to 68ac258fc6c8e489f360fde3123a5894
Variable MACHINE value changed from 'b4420qds' to 'b4860qds'
Signed-off-by: Ting Liu <ting.liu@freescale.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
[YOCTO #7230]
In certain system configurations TRANSLATED_TARGET_ARCH will not
expand in the right order for gcc-cross-candian-mips64n32 to be
generated properly.
This will cause SDKs to fail to generate properly.
Changing the global definition of TRANSLATED_TARGET_ARCH always
expands the ABIEXTENSION, which causes the OVERRIDES to pick it up
as well. This effectively defines a new class of overrides for the 'n32'.
The side effect is that we need to duplicate some mips64 overrides, and
redefine others that were previously 'n32' or 'mips64' exclusive to have
the correct semantics.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
Richard Purdie
Joe Slater
Andre McCurdy
Martin Jansa
Marek Vasut
Saul Wold
Kai Kang
Juro Bystricky
Jussi Kukkonen
Ross Burton
Cristian Iorga
Khem Raj
Dmitry Eremin-Solenikov
Denys Dmytriyenko
Robert Yang
Belen Barros Pena
Koen Kooi
Paul Eggleton
Alejandro Hernandez
Chong Lu
Bruce Ashfield
Belen Barros Pena
Paul Barker
Ting Liu
Mark Hatle