If we don't do this, the SDK target sysroot is named generically even
when it contains package architecture specific optimisations.
(From OE-Core rev: 0616557a8c29b42bae0ffd5fd665a046810047e4)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This variable should be split with \n sequences and these need to be
specified literally in the string. A corrected version of the example
given in the original commit (OE-core rev
75e3875341ddc8940e9ee2ccbbb2ec18194a68e6):
SANITY_TESTED_DISTROS = " \
Ubuntu 11.04 \n \
Fedora release 14 (Laughlin) \n \
"
(From OE-Core rev: cfc72d5796b6f83a01e06f3a1f044869db2d5d18)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Otherwise the class doesn't work if ${bindir} is set to a different value;
likewise for /var vs ${localstatedir}.
(From OE-Core rev: 21371df16917cd82642b39763793783d61ee5516)
Signed-off-by: Phil Blundell <philb@gnu.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Commit 35fa8dc5f7 changed the gcc recipes to use
baselib for the compiler location. This is fine as long as baselib happens to
match the platform multilib definition which is enabled at the time.
This patch fixes things so that gcc will honour whatever ${base_libdir} is
set to re-allowing suitable customisation of the system layout.
[YOCTO #1362]
(From OE-Core rev: bc5f293b151b9ba0d6660814d88ee5041efce318)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
For a given system we only want one kernel to be built. This change makes
the main kernel recipe provide all of the provides of the various enabled
multilibs hence allowing it to fulfil all the appropriate dependencies.
To make this work a global multilib class file needed to be created.
This patch also enables this multi provider functionality for "allarch"
packages.
[YOCTO #1361]
(From OE-Core rev: 2fd257f6c610624f05c8dd3fe1486364af04696f)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
libsdl is required by sato image, so extend it for multilib.
(From OE-Core rev: 88abab9ba9632e87f3a081915a69cbc5c36b4d3c)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Thinking of the senario that, if we already built out a 64bit image
along with the full toolchain bootstrapped, then we need to build some
32bit libraries, which needs lib32 versions of gcc and eglibc. These
toolchain recipes will bootstrap again in the same sysroot, resulting
that lib32-gcc-cross-initial will find some macros owned by eglibc have
already been defined and thus it includes non-existed headers that
provided by later lib32-eglibc.
The solution for the above issue is to use different sysroot for
multilib recipes, here we add ${MLPREFIX} in front of the machine
specific sysroot directory name.
[YOCTO #1372]
(From OE-Core rev: a1508ad1aec2d2f9ee040aa217c33193cd5bd871)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
To get the MULTILIB_PACKAGE_ARCHS, we need to get the corresponding
DEFAULTTUNE value. This fixes the multilib arch directory missing issue
in solvedb-ml_archs.conf.
(From OE-Core rev: 54306ff373e13696637b547fa1514e0ef8633248)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
hal has runtime dependency on kernel, but not build time. Remove it from
"DEPENDS" list.
Also fix a wrong PACKAGE_ARCH setting when building multilib lib32-hal,
because ":=" will be extended immediately which is not the right value.
Using TUNE_PKGARCH instead.
(From OE-Core rev: 74646a2b2d7d452dfe95b08940389a686e8addcb)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We rename readprofile to readprofile.util-linux so we need to use that binary
name in the FILES entry for the readprofile package.
(From OE-Core rev: 55168655ec95e8eff70f90a462ed0a8d87d8d87f)
Signed-off-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There are some links become invalid after rename, so that failed dependencies
detected when install rpm package. This commit update links to resolve it.
[YOCTO #1158] got fixed.
(From OE-Core rev: 48ec20d1331eb665d9fc1a06bdb1ea79e4513159)
Signed-off-by: Zhai Edwin <edwin.zhai@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
dropbear will check "/etc/pam.d/sshd" which comes from package "openssh" \
When enabling pam supporting. But if we only install dropbear \
package without package "openssh", then "dropbear" will not \
find a configuration file.
The changes are as follow for fixing this bug:
- Change the path to find configuration file (/etc/pam.d/sshd --> /etc/pam.d/dropbear)
- Add a configuration file "/etc/pam.d/dropbear"
(From OE-Core rev: 48dcb8fc7b669b27160dde33079f40551853702b)
Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[YOCTO #999]
qt4e-demo-image needs qt-demo-init when starting qtdemoE.
qt-demo-init was pulled from Openembeded.
(From OE-Core rev: e21e8b502ab2f982836cf1f7a30e33bff1bd5b7b)
Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Latest connman-0.75 change the WiFi security type, causing the
incompatibility of connman-gnome. Fix connman-gnome accordingly to show
the correct security informtion.
[YOCTO #1343]
(From OE-Core rev: 38589c37923ecfe34f1dbd5f12d89d55dfb11e2a)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In the event the xserver fails to start the dbus-wait has to time out and
this causes many users a lot of confusion. If we wish to reinstate this,
we need to find a safer mechanism to do it where X failing to start cancels
the timeout (sends a dbus event at that point?) The comments are left in the
file as an example in case some user does wish to enable is.
(From OE-Core rev: 0471b17b061e57231387ef90c95fc0c34fc0e66b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[YOCTO #1398]
Firstly we should start checking if MACHINE is set. It if isn't lets
make this a sanity warning since its very hard to successfully build
without setting one (and anyone wanting to do that can disable the
sanity checks easily enough anyway).
Some of the checks depend on a MACHINE being specified. This change
moves those checks to a separate function so they only run if MACHINE
has been set correctly.
Both these issues combine to ensure the user sees a sane message
and avoids the nasty tracebacks in the bug report referenced above.
(From OE-Core rev: 02aceca132f9e259cdc5283c4bfe84e6a55df54d)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[YOCTO #1388]
This change is needed to correctly add the dependencies for the do_package
task which bitbake is unable to automatically detect itself.
(From OE-Core rev: 0614b9aa62a46f81d334ca4230080cc707347f3c)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently in the native case we have a path that can easily exceed the
interpretor limit so use "env nativeperl" in that case.
This patch also fixes up the target version's interpretor path but
to do this we need to bypass the configure checks.
(From OE-Core rev: b9fa8fd4f09b68cbcc12b557451a58a1f218d7f3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This reverts 19fb07bf337e1d724798e2eb4479c35fc45b1941 and restores
the behaviour of the code to the way it was prior to 561d875404.
See http://lists.linuxtogo.org/pipermail/openembedded-core/2011-August/007407.html and subsequent messages.
(From OE-Core rev: e5810439cc394d8ebfc264b05e1fbfad19e8fcd3)
Signed-off-by: Phil Blundell <philb@gnu.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There are no locales resources in an lsb image because the locale resources were splitted from eglibc-package.inc \
and putted into a new file "eglibc-locale.inc". So the above variables lost their function in eglibc-package.inc and
caused no locale resources in an lsb image. The purpose of moving the above variables from eglibc-package.inc to \
eglibc-locale.inc is to make the above variables to take effect.
eglibc-binaries include packages "eglibc-binary-*"
eglibc-localedatas include packages "eglibc-localedata-*"
eglibc-gconvs include packages "eglibc-gconv-*"
eglibc-charmaps include packages "eglibc-charmap-*"
(From OE-Core rev: 0903a2c62a58f94b7d424162e8c852b502d3840a)
Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* because 7.3 archive is not fetchable anymore
(From OE-Core rev: c27690a04acf3870f6e0133cd05ba93d6c66126e)
Signed-off-by: Sebastian Krzyszkowiak <dos@dosowisko.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently it uses ${TARGET_SYS}-. This is inconsistent, as the recipe
and kernel bbclass both use the prefix. While there aren't many cases
where the two differ, it is harmless to ensure that we are behaving
consistently.
(From OE-Core rev: 8aec52f4fd1ad3e4148e2ad32700a4378e69dcd3)
Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
- Don't use a variable reference in the IMAGE_DEPENDS
- Inherit kernel-arch to get UBOOT_ARCH
- Don't include the .bz2 variants, since the base types aren't in
oe-core
- Add the new types to IMAGE_TYPES
- Inherit image_types, to be certain we load after it
(From OE-Core rev: f67789b83599b86be052b3f2d686791cbf24f540)
Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
binutils is now at version 2.21.1a.
(From OE-Core rev: 92808484b1dfa55afa1af3a365898aab6d45f746)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2.21.1 has disappeared upstream. There is no ChangeLog information for
2.21.1a, however diff reports only the addition of some cpu/opc files in
cgen/ subdirectory; the source code is otherwise identical.
(From OE-Core rev: 4ecf379aba11983b66be4fbe26e65aa28740ddfd)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Use KERNEL_OUTPUT variable to find the generated kernel image rather than
duplicating the existing path. This also means it can be overridden simply.
(From OE-Core rev: de5b502330ded38c0efe2c4e30967ef12dbad72a)
Signed-off-by: Mike Crowe <mac@mcrowe.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This allows pango to build on DISTROs which don't enable x11.
Also adjust the SECTION since pango isn't really an X library.
(From OE-Core rev: ac92a573dd38270a28967c76472435cfd7f9bdfe)
Signed-off-by: Phil Blundell <philb@gnu.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Package files disappearing during index generation don't cause a fatal
error. Make package file disappearing during filelist generation also
a non-fatal error.
(From OE-Core rev: f55c96b6d53ecff865145fa22f286c89c7336515)
Signed-off-by: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* gconv is directory so test -f was always false as reported by Phil Blundell in
http://lists.linuxtogo.org/pipermail/openembedded-core/2011-August/008973.html
* -e used also for localedef which is file, just to be more consistent
(From OE-Core rev: feaf1bab56138cc1bc998a4201a60202a5b8cd28)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Initscripts are missing in target image in multilib case. This commit
adds the renaming logic for the related variables in multilib.bbclass.
This fixes the no response of mouse/keyboard in target system due to
the missing of udev startup script.
(From OE-Core rev: 477fc6e2fc034c68a250005461774bc8ecf91a52)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add "pkg_postinst" and "pkg_postrm" as renaming elements, which fixes
missing post install/rm scripts in target image.
(From OE-Core rev: a5c9f8af79c1632d2f3de5dd9627a7ca24fa0b76)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In the orignal logic, the renaming will not work for "FILES_" if defined
variables as:
PACKAGES = "${PN}"
FILES_abc = "/usr/include/abc.h"
It is because ${PN} is "lib64-abc" so it will not be contained in
pkgrename.
This commit enumerates all element in PACKAGES, getting the original
packages and multilib packages, then doing renaming for "FILES_",
"RDEPENDS_", etc. This fixes a lot of missing files and incorrect
dependencies.
(From OE-Core rev: ea7c196b4097d64b4f42faccaf075334c297ba20)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Use BPN instead of PN in FILES path to support multilib.
(From OE-Core rev: cb247225a96d57e67ce32c5674bb13b703e30a7c)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Use BPN instead of PN in FILES paths.
This fixes the login issue by /etc/passwd missing in multilib images.
(From OE-Core rev: ec62502988f0d79a69013513f2246980e0d8c01c)
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[YOCTO #999]
Copy metadata of qt4e-demo-image from OE for easy testing qt4embeded on embedded platform.
This image was tested on qemuarm and can run program "qtdemoE".
(From OE-Core rev: 612ab0aaf3e693b90bb18234e2517bfc87ee3f7c)
Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixes [YOCTO #1392]
Updating the SRCREVs to pickup:
[
mips/rt: convert cascade interrupt non threaded
The preempt_rt kernel forces all irq interrupts to be threaded,
but special interrupts can be excluded from this conversion.
The cascade interrupt should be part of these exceptions.
In this case, irq2 is initialized before "kthreadd" task, which
converts irq interrupt to threaded.
If this irq is threaded, the kernel calls "try_to_wake_up" function
to wake up "kthreadd" task, but at that moment, "kthreadd" task
has no been initialize and try_to_wake_up wakes up a NULL task.
(From OE-Core rev: 7ceb78a42f7d9b6aa4b984d04ac13f8dc23a9095)
Signed-off-by: Liming Wang <liming.wang@windriver.com>
]
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie
Paul Eggleton
Phil Blundell
Dongxiao Xu
Koen Kooi
Saul Wold
Joshua Lock
Jessica Zhang
Zhai Edwin
Xiaofeng Yan
Sebastian Krzyszkowiak
Chris Larson
Mike Crowe
Dmitry Eremin-Solenikov
Martin Jansa
Bruce Ashfield