Adding ability to use postinstalls intercepts in the nativesdk env, and
making sure the correlate between repo + SDK.
This to enable rootfs generation from a package repository using only a
package repository and the toolchain tarball.
See https://github.com/nysan/rootfs-sandbox for examples.
(From OE-Core rev: 5d5ef0f5620ec525491c53e8457e7c65c226bfb4)
Signed-off-by: David Nyström <david.nystrom@enea.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Use python-distribute-native in packagegroup-toolset-native
Also fixes warnings like
WARNING: The recipe python-setuptools is trying to install files into a
shared area when those files already exist. Those files and their
manifest location are:
/srv/ssd/sgw/builds/world/tmp/sysroots/qemux86-64/usr/lib/python2.7/site-packages/setuptools.pth
Matched in manifest-qemux86-64-python-distribute
/srv/ssd/sgw/builds/world/tmp/sysroots/qemux86-64/usr/lib/python2.7/site-packages/site.pyc
Matched in manifest-qemux86-64-python-distribute
/srv/ssd/sgw/builds/world/tmp/sysroots/qemux86-64/usr/lib/python2.7/site-packages/site.py
Matched in manifest-qemux86-64-python-distribute
Conflicts:
meta/recipes-devtools/python/python-setuptools_1.4.bb
(From OE-Core rev: ddb0c70a7d4d51a1f404c194a562325137222146)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
So it can get the correct python direcotries for python2
(From OE-Core rev: cfc780ec26474a61bd52a8f6a66c7280fa12a075)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
3.3.3 is latest bugfix release in 3.3 series
License text is changed to reflect year and prev releases
+++ ../Python-3.3.2/LICENSE 2013-05-15 09:32:54.000000000 -0700
@@ -74,7 +74,9 @@
3.2.1 3.2 2011 PSF yes
3.2.2 3.2.1 2011 PSF yes
3.2.3 3.2.2 2012 PSF yes
+ 3.2.4 3.2.3 2013 PSF yes
3.3.0 3.2 2012 PSF yes
+ 3.3.1 3.3.0 2013 PSF yes
Footnotes:
@@ -110,8 +112,8 @@
distribute, and otherwise use Python alone or in any derivative
version,
provided, however, that PSF's License Agreement and PSF's notice of
copyright,
i.e., "Copyright (c) 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008,
2009, 2010,
-2011, 2012 Python Software Foundation; All Rights Reserved" are
retained in Python
-alone or in any derivative version prepared by Licensee.
+2011, 2012, 2013 Python Software Foundation; All Rights Reserved" are
retained
+in Python alone or in any derivative version prepared by Licensee.
We use native python in disutils bbclass and this needs
to peek into target sysroot when building plugins in cross
environment. Otherwise anything that inherits distutils3.bbclass
will not build.
Fix host include contamination issue
Let compiler append sysroot to include path if it can
Fix the compiler invocation and linker flags when cross compiling
The details are in patch header. This should fix the QA errors about
host contamination
Add virtual/libintl to dep list
There is use of libintl.h but we dont
express the dependency
Add lzma dependency for python3-misc
Fixes
and from test-dependencies.sh run there is undeterministic lzma
dependency:
WARN: packages/armv5te-oe-linux-gnueabi/python3/python3-misc/latest lost
dependency on liblzma
Fix staging warnings due to duplicate install
WARNING: The recipe python3 is trying to install files into a shared
area when those files already exist. Those files and their manifest
location are:
builds/world/tmp/sysroots/qemux86-64/usr/include/python3.3m/pyconfig.h
builds/world/tmp/sysroots/qemux86-64/usr/lib/libpython3.3m.so.1.0
builds/world/tmp/sysroots/qemux86-64/usr/lib/libpython3.3m.so
builds/world/tmp/sysroots/qemux86-64/usr/lib/python3.3/config/Makefile
Let compiler search includes relative to sysroot
This makes it cross compilable and assumption on hardcoded
paths is fixed.
Errors like below would appear if ncursesw is installed on build host
cc1: warning: include location "/usr/include/ncursesw" is unsafe for
cross-compilation [-Wpoison-system-directories]
(From OE-Core rev: 185fcfc3286b355394ae89300f22b8cb7aaa43c1)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Fix linking not happening issue in modules
We need to tweak CONFIGURE_LDFLAGS for python3 instead of LDFLAGS
directly
Add second compile stage which uses original makefile
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In python3 this has to be taken care of, it defines a variable to
denote ABI currently its at 'm' and it uses this to construct
the directory names for installing the python headers and library
names in sysroot. e.g. it will be something like ../python3.3m/...
We need this information when we are using distutils to cross build
python extentions and want to know the locations of python headers
and libraries install locations
(From OE-Core rev: 9a3e4ac4a4eeed64ec80d400130dff3d26daf336)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In line with python2 add distutils and setuptools
classes for handing python3
Use python-distribute instead of python-setuptools in setuptools bbclass
Remove --single-version-externally-managed since its setuptools
specific and we dont use it anymore
Do build_ext as separate step during compile
Add DISTUTILS_BUILD_EXT_ARGS for modules to pass flags to build_ext step
in setup.py
Add build_ext as sepate step during compile and add
the cross sysrooted library and headers since we are cross
compiling
Use ${PYTHON_PN} in place of hardcoding python name
Remove the /etc/share if its empty
Since OE-Core times we now have machine specific sysroots
for targets unlike before when we used arch specific sysroots
so reflect that here
Use MACHINE for sysroot when not building for build host
Python's machinery replaces directories in sysroot path to match OE's
staging area sysroots. Earlier we use to have HOST_SYS represent sysroot
always but now we use MACHINE to represent target sysroots but HOST_SYS
to represent host sysroot. This patch caters to that difference
(From OE-Core rev: 8bb0206ed67228c88dd5bc2d8b36ce28f48b78f4)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is needed to accomodate python3 alongside
python2
(From OE-Core rev: ae931c2cf9e48e1fb74b4b727dbf668ea880023f)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This causes issues when postinstalls have ERROR keywords
its interpreted as error and image build is cancelled
(From OE-Core rev: f5bc3cfac9545c402b415695c4e0f98ad38fb2b0)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Like python2 mostly except
This will package collections into python3-core instead
of python3-misc fixes errors like
Traceback (most recent call last):
File "/usr/lib/python3.3/site.py", line 69, in
import os
File "/usr/lib/python3.3/os.py", line 659, in <module>
from collections.abc import MutableMapping
ImportError: No module named 'collections'
(From OE-Core rev: e314404876e62c05c7ea5f5e79b2b05c3ed9ab84)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* packagedata task was introduced in:
commit 6107ee294afde395e39d084c33e8e94013c625a9
Author: Richard Purdie <richard.purdie@linuxfoundation.org>
Date: Wed Jan 23 14:27:33 2013 +0000
Subject: Split do_packagedata task from do_package
* rm_by_stamps wasn't using do_packagedata or do_packagedata_setscene
stamp files to find which sstate archives to keep, so it was removing
all of them
(From OE-Core rev: 66ef2e62a7b5fe36c718f56a8ea9d7f6fd77c393)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* grep for AVAILTUNES isn't enough in cases where AVAILTUNE doesn't
match exactly with TUNE_PKGARCH, e.g. AVAILTUNE "cortexa8thf-neon"
and TUNE_PKGARCH "cortexa8t2hf-vfp-neon", instead of trying to find
dynamically every available TUNE_PKGARCH (we have _a lot_ of them
even with oe-core only), add parameter --extra-archs where user can
define extra architectures he supports in given build
* Don't replace '-' with '_' for extra-archs, it does apply to MACHINE
names and some AVAILTUNES, but e.g. cortexa8thf-neon shouldn't be
converted to cortexa8thf_neon
* Add empty architecture for populate_lic sstate archives
* Add ${build_arch}_${arch} combinations for toolchain recipes (e.g.
gcc-cross is using x86_64_i586
(From OE-Core rev: a27cc54fb2d0e59f3a800893c1848cb26a7c5335)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* it's good to see some the ratio of delted files until now it was
showing only when all or none files were to be removed
(From OE-Core rev: 54e6e25f1a369fa6c21ce0f9db3479b1a481825f)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* format of filenames for sstate archives was changed in:
commit 6f823a23c5f1d0ffa0a27db1c1bc1907de788505
Author: Richard Purdie <richard.purdie@linuxfoundation.org>
Date: Fri Dec 20 12:06:02 2013 +0000
Subject: sstate: Convert to use ':' as a filename sperator and use SSTATE_SWSPEC globally
* this one doesn't need special care for old sstate- names
they will be removed automatically as they don't match with
any checksum in rigth format from stamps directory
(From OE-Core rev: aa36f9c9b5abac58de899f98803d1c4375678044)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* format of filenames for sstate archives was changed in:
commit 6f823a23c5f1d0ffa0a27db1c1bc1907de788505
Author: Richard Purdie <richard.purdie@linuxfoundation.org>
Date: Fri Dec 20 12:06:02 2013 +0000
Subject: sstate: Convert to use ':' as a filename sperator and use SSTATE_SWSPEC globally
* remove_duplicated() wasn't able to find available architectures and
duplicate files since this change
* add extra step to remove old sstate archives starting with sstate-
(instead of sstate:)
(From OE-Core rev: ddb26341611c3dff41ea92a73d93ec01ae2865de)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Its possible something can delete $1 (since it may be empty) whilst cp-noerror
is starting. Add an exception to handle this issue since if this happens, we
shouldn't return an error.
(From OE-Core rev: 13061ed1e1f347589d6955d5cc50a50574b00218)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
For source URI, the use of ${PN}-${PV} leads to
build failures for multilib and x32.
${BP} is properly used instead.
Partial fix for [YOCTO #5781].
(From OE-Core rev: 2e0dbfaed17fb89381e3bf155a29cd606b698c7c)
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This solves the integration problem with busybox.
Previously, there was a patch in busybox to move 'watch' to /usr/bin.
Such patch is not accepted by upsteam and really not necessary as
our ALTERNATIVE system can easily solve such intergration problem.
(From OE-Core rev: 9e6e2aee21ff59687ddfd0e23fd0add9ee81d397)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
To correctly integrate with busybox in our system, we should add
'stat' to base_bindir_progs so that the 'stat' commands from busybox
and coreutils both register to /bin/stat.
Previously there was a patch in busybox to move 'stat' to /usr/bin.
But as we can easily solve this integration problem by modifying the
coreutils recipe, this patch has been removed. After all, maintaining
a patch that's not accepted by upsteam should not be our No.1 choice.
(From OE-Core rev: d98d6122bdfd84faaa37912ca66dabebc7eb9da6)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As busybox has been upgraded, rename this bbappend file to make it
match the current version of busybox.
(From OE-Core rev: 12e33bd0dc02720122eb8088b08a181b13b6ef0e)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Upgrade busybox to the stable release 1.22.1.
During this upgrade, 9 patches are removed. Reasons are detailed below.
The following 6 patches are removed as they have been merged.
meta/recipes-core/busybox/busybox/busybox-lineedit-initialize-delptr.patch
meta/recipes-core/busybox/busybox/busybox-list-suid-and-non-suid-app-configs.patch
meta/recipes-core/busybox/busybox/busybox-sed-fix-sed-clusternewline-testcase.patch
meta/recipes-core/busybox/busybox/busybox-sulogin-empty-root-password.patch
meta/recipes-core/busybox/busybox/find-get-rid-of-nested-functions.patch
meta/recipes-core/busybox/busybox/testsuite-du-du-k-works-fix-false-positive.patch
The following three patches are removed because they are mainly about moving
binaries from /bin to /usr/bin to make the update-alternative work correctly
at rootfs time. We can easily solve this problem by changing recipes. There's
no compelling reason why such patches are needed.
meta/recipes-core/busybox/busybox/run-parts.in.usr-bin.patch
meta/recipes-core/busybox/busybox/stat-usr-bin.patch
meta/recipes-core/busybox/busybox/watch.in.usr-bin.patch
(From OE-Core rev: fa282186d82e8fe3c590d4ea79e464116e5ceea2)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There would be some unexpected errors when the whole TMPDIR is located
on nfs, so add a test for it in sanity.bbclass.
Note:
The better way to get the filesystem id should be get f_fsid from struct
statvfs, but there is no f_fsid in os.stat() or os.statvfs(), so we use
'stat -f -c "%t"' here.
BTW., s/tmpdir/TMPDIR/ in the previous comment message to make them have
a uniform.
[YOCTO #5442]
(From OE-Core rev: ee4061b43522c4893b41c3be63d06be1ee7e3c70)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As sysvinit doesn't inherit update-rc.d, we need to add this dependency
manually.
(From OE-Core rev: 4ea9ca2bbacb20c3ed454b80e2c020c1073f3299)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Automatically add 'initscripts-functions' to packages when needed.
This extra dependency is added only when the package's init script
would be used at system start-up. In other words, in a systemd based
image, if the package has already got its service files, it will not
have this extra dependency.
(From OE-Core rev: de875cb8893fd8372dd77babaef54ab2d790693d)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Instead of manually adding initscripts to RDEPENDS of each package,
we should make it automatically handled by the update-rc.d.bbclass.
This solution would have the benefit of backward compatibility. In
other words, users need not modify their recipes.
This reverts commit 16080a3485.
(From OE-Core rev: f9f193219bd510160b6b09bae652a9dc8ea01e7b)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The mips64 uses long long for u64 in the kernel, but powerpc's asm/types.h
prevents 64-bit userland from seeing this definition, instead defaulting to u64
== long in userspace. Define __SANE_USERSPACE_TYPES__ to get int-ll64.h
included.
We had a similar fix on ppc64, use it for mips64 will fix the problem.
[YOCTO #5758]
(From OE-Core rev: 97bc0d1575b08830e3953933cbbc5732b1a82b31)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Split the class in two, one to handle the process and the
timeout based on output and one for the actual ssh/scp commands.
The ssh/scp methods now use the same run method.
It does the same thing as before but:
- it looks cleaner.
- adds support for using a different user than root
- optionally, raises an exception when exit code != 0
(that's useful for code outside of tests, where you wouldn't want
to check the return code every time as the tests do)
(From OE-Core rev: bb14a7598d3c0636dc249f719bde0d9d65b2694d)
Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
A issue was fixed in target recipe with commit: 95893404
[
quilt: added ac_cv_path_BASH to CACHED_CONFIGUREVARS
On Fedora 17, bash has moved to /usr/bin/bash and the configure process finds it
on the host machine there, this ensures that it is set correctly for the target.
[YOCTO #2363]
Signed-off-by: Saul Wold <sgw@linux.intel.com>
]
But it's also impacting native package in the case that the sstate is used
between Fedora 17(or later) and other hosts.
(From OE-Core rev: 34bd53c657f82a402723fcf2640b0511a68b6af5)
Signed-off-by: Ming Liu <ming.liu@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add ffsb-remove-hardcoded-configure.patch: get rid of the hardcoded configure call preventing cross-compilation
Add do_configure_prepend to explicitly force regeneration the configure file and pass configure options as appropriate.
(From OE-Core rev: ce48f4ebfeae0bca599c6e39958ec8dec11e09dd)
Signed-off-by: Fathi Boudra <fathi.boudra@linaro.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If a kernel is built without any external modules (aka no CONFIG_*=m),
then during a modules_install of the kernel an empty directory is
created at /lib/modules/${KERNEL_VERIONS}/kernel. This is behaviour of
the kernel infrastructure, the directory would normally be populated
with the modules that were built.
However because of the expectations of kernel-modules-split, no packages
are created when there are no modules and an empty directory lingers.
This raises QA issues as warning or errors (depending on the distro).
The following patch changes the kernel_do_install task to check if the
directory is empty and if so removes it.
(From OE-Core rev: 63d3070b7e1207164891b154a5b9017731e75872)
Signed-off-by: Nathan Rossi <nathan.rossi@xilinx.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
After syslinux was updated from 4.07 to 6.01, booting memtest86+ using
pxelinux would fail:
PXELINUX 6.01 2013-07-04 Copyright (C) 1994-2013 H. Peter Anvin et al
Loading memtest86... ok
Booting kernel failed: invalid argument
This backports the necessary upstream patches to allow memtest86+ to
boot using pxelinux again.
[YOCTO #5501]
(From OE-Core rev: ed9ccb8622b347173602be8b2126324d4fdf54d8)
Signed-off-by: Jonathan Liu <net147@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
A line got dropped from the original patch when porting to the latest
unfs3. The regression was introduced in commit 7d8075c64 (unfs3: Fix
dependencies and allow target builds).
This patch restores the missing line from the original which had been
working fine for over a year's time.
--- a/Config/Makefile.in
+++ b/Config/Makefile.in
@@ -12,7 +12,9 @@ all: lib.a
lib.a: $(OBJS)
$(AR) crs lib.a $(OBJS)
-y.tab.h y.tab.c: exports.y
+y.tab.h: y.tab.c
+
+y.tab.c: exports.y
$(YACC) -d exports.y
(From OE-Core rev: 9edbd9d872c128038fd7b56fac713256fd69c6c5)
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Shared work directories work by assuming bitbake will not run
more than one task with a specific stamp name. Recent runqueue optimisations
accidentally broke this meaning there could be races. This fixes the code.
(Bitbake rev: b1628b1a260ddf43fc9985535b1ddcfcebbb1e5b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
GCC 4.8.0, 4.8.1 and 4.8.2 can generate broken epilogues for the
ABI used by the kernel. Apply the patch that is included for GCC
4.8.3 from http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58854.
The issue was found on Yocto/Dora and the patch should be backported
to this branch. A kernel built with Dora's GCC 4.8.1 misbehaved on:
while true;
do
(for i in `seq 1 100`;
do
echo "Log message... $RANDOM";
done) | logger;
done
busybox's syslogd would from time to read a huge negative value and
then exit, strace would get stuck waiting on a syscall. After this
patch it appears to work better.
(From OE-Core rev: 3004eb3b7ee5fd8dfe9c4e5749b4e125d0bd4b59)
Signed-off-by: Holger Hans Peter Freyther <holger@moiji-mobile.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We're seeing similar failures that we saw from libtool-native and libtool-cross
where /bin/sh changed from bash to dash on different machines after sstate
reusage. This patch fixes nativesdk to avoid the same issue.
(From OE-Core rev: 7994b2870dac8fd5f6db6d47043378534b644515)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
