Yu Ke [Tue, 22 Feb 2011 12:41:19 +0000 (20:41 +0800)]
xserver-nodm-init: add rootless-x support
most rootless X work are already done in the kernel, xserver and
graphics driver, this patches add the the remaining userspace setting:
- create /etc/X11/Xusername to set rootless X user
- add rootless X user to group video, tty to access /dev/tty[0-4]
and /dev/dri/card0
- grant rootless X user access right to /dev/input/*, /var/log
Yu Ke [Tue, 22 Feb 2011 12:36:45 +0000 (20:36 +0800)]
xf86-video-intel: add config option to support rootless X
- add global config option ROOTLESS_X to control if enable
rootless X for the machine. ROOTLESS_X requires graphics
driver supporting KMS (kernel mode setting), so far, only
atom_pc support this. so enable ROOTLESS_X for atom_pc machine
- add config options for xf86-video-intel to support rootless X
Part of the fix for Bug 628 is to add a note that bitbake requires
python 2.6. I added this note before the example bitbake command
that builds an image. I also added a linked reference to the Poky
Reference Manual and mentioned the FAQ appendix. There will be more
information about the python requirement in the FAQ.
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
The mechanism to trigger these options is in the form of an
optional kernel feature that is only appended for qemux86
and qemux86-64, but is contained within the kernel tree.
This allows several things:
- the options to be available/shared for all boards
- the options to be in tree
- to not add the options to every board, which unecessarily
bloats the default configuration.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Scott Garman [Wed, 23 Feb 2011 02:32:45 +0000 (18:32 -0800)]
sstate.bbclass: add SSTATEPOSTINSTFUNCS
When doing builds using sstate cache, there was no way to run
the equivalent of a pkg_postinst function. This is needed by
the SGML-related documentation recipes to properly update the
catalog files when new DTDs and stylesheets are installed.
SSTATEPOSTINSTFUNCS is a new variable you can set to function(s)
in your recipe to run after install is completed from sstate.
Thanks to Richard Purdie for suggesting this solution.
Signed-off-by: Scott Garman <scott.a.garman@intel.com>
If expanding a variable triggers an exception the caller currently has no
way to supress the error message or otherwise handle the siutation. An
example of where this is a problem is "bitbake -e" showing tracebacks and
errors for variables like SRCPV in OE/Poky.
Secondly in a chained expansion fails, log mesages are recorded for
every step of the expansion, not just the innermost error which is
where the real failure occured.
To fix this we introduce a new exception ExpansionError which callers
can handle as appropriate.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 23 Feb 2011 11:48:01 +0000 (11:48 +0000)]
bitbake/data_smart: Improve the way lazyassignment works
Currently, if a variable has been set with ??= and the code looks it up
before the data finalisation phase, no value is found. This is causes
serious problems for anonymous python functions which manipulate data, or
for the fetcher revision handling code where revisions can be set with
??=.
There is also a significant performance implication for processing lazy
assignment in finalise.
Moving the check for a default value into getVarFlag addresses both
the timing issue and the performace. This change gives a 7% real time
performance improvement to parsing the Poky metadata. The cost of the
check at this point is minimal since we have all the data flags available.
This should also fix Yocto bug 752.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 23 Feb 2011 11:09:07 +0000 (11:09 +0000)]
bitbake/cooker: Fix parsing failure zombie problem
When parsing if a SystemExit event is triggered, it causes the parsing thread to
exit and the main process hangs waiting for it to finish indefintely. Add code to
catch BaseExceptions and raise these with the main process gracefully instead
of just hanging indefinitely with zombie processes.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Zhai Edwin [Tue, 22 Feb 2011 01:58:11 +0000 (09:58 +0800)]
qemu: enable audio support
On qemux86, export ac97 & es1370 emulated device to guest, and enable host oss&alsa
driver. So end user can get sound from qemux86 guest if the sound card
driver installed.
Dexuan Cui [Wed, 16 Feb 2011 09:16:39 +0000 (17:16 +0800)]
mklibs-native: fix the build on Ubuntu 9.04 by defining STT_GNU_IFUNC
mklibs-native needs elf.h (that is supplied by libc6-dev).
since STAGING_INCDIR_NATIVE doesn't contain elf.h (we don't have eglibc-native),
so the host's /usr/include/elf.h is used.
Unluckily, the libc6-dev in Ubuntu 9.04 is so old that the elf.h doesn't
define STT_GNU_IFUNC, so we have to define it ourselves.
Richard Purdie [Tue, 22 Feb 2011 10:53:15 +0000 (10:53 +0000)]
package_deb.bbclass: Place the whole task under fakeroot context to fix build failures
If we don't do this, the ipk/rpm backends can create temporary files and then
when the deb package creates new files, those inodes can be reused and permission
confusion results.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Darren Hart [Fri, 18 Feb 2011 23:38:12 +0000 (15:38 -0800)]
linux.inc: remove the obsolete linux.inc kernel recipe include
linux.inc was used by older kernel recipes which have now been removed from the
core meta data. I references machines now only defined in meta-extras. The
configure prepend mangles the kernel .config in non-intuitive ways and the
install peroforms some odd boot image manipulation that is not used nor required
by supported machines. The required functionality, such as defconfig setup, has
already been moved to the kernel bbclass. Remove linux.inc to avoid confusion
for new kernel recipe authors and clean up the kernel meta data.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Darren Hart [Fri, 18 Feb 2011 23:32:38 +0000 (15:32 -0800)]
linux-2.6.33.2: remove linux-2.6.33.2 recipe
The one machine listed for this kernel recipe does not have a machine config.
This recipe uses some older mechanism which are being phased out. Remove it to
avoid confusion and clean up the kernel recipes metadata.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Liping Ke [Mon, 21 Feb 2011 21:36:15 +0000 (05:36 +0800)]
ADT: Fix check_result script cond comparison bug
When meeting errors, the return number can't be directly compared with
-1. Actually, it might be represented as 255. The correct way is to
compared it with 0. If the result is non-zero number, we meet error.
This patch is for fixing [BUGID #742]
Richard Purdie [Mon, 21 Feb 2011 16:29:43 +0000 (16:29 +0000)]
tar-replacement-native: Add a target to replace the default tar
tar < 1.24 has symlink issues where extracting a tar archive containing a symlink
to a directory where that symlink already exists will cause the symlink to be
dereferenced. If that target doesn't exist tar can fail with a permissions error.
Since we need to be able to do this for packages containing symlinks like
xorg-minimal-fonts and eglibc, we have to ensure a tar 1.25 is available early
in the build process.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Scott Rifenbark [Thu, 17 Feb 2011 14:07:06 +0000 (08:07 -0600)]
documentation/kernel-manual/kernel-how-to.xml: Changes to section 3.4.4 and new 3.4.5
I made some updates to the 3.4.4 section according to Bruce Ashfield's
feedback. I also added a new section 3.4.5 (Creating a BSP Based on an
Existing Similar BSP Without a Local Kernel Repository).
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Scott Rifenbark [Tue, 15 Feb 2011 22:07:45 +0000 (16:07 -0600)]
documentation/kernel-manual/kernel-how-to.xml: fix to step 3 and 4 of section 3.4.4
Added text after the sample bbappend file stating that the user needs
to update the KSRC assignment statement and also remove its comment.
Also they have to remove the comment from the SRC_URI line.
Changed the bitbake command in step 4 to use the linux-yocto-stable
file rather than the linux-yocto file.
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Scott Rifenbark [Tue, 15 Feb 2011 21:54:25 +0000 (15:54 -0600)]
documentation/kernel-manual/kernel-how-to.xml: Fix to step 3 of section 3.4.4
Changed 'linux-yocto_git.bbappend' to 'linux-yocto-stable_git.bbappend'
in step 3 just before the code example. This makes the code consistent
with the rest of the example supporting the 2.6.34 kernel.
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Scott Rifenbark [Tue, 15 Feb 2011 21:49:33 +0000 (15:49 -0600)]
documentation/kernel-manual/kernel-how-to.xml: Fix to step 1 of section 3.4.4
Step one states that the linux-yocto_git.bbappend file is left. This
file is associated with distro 2.6.37 kernel and the rest of the
example is associated with the 2.6.34 kernel. So, I changed the
step to indicate that the directory is left with a linux-yocto-stable_git.bbappend
file.
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Richard Purdie [Fri, 18 Feb 2011 15:32:57 +0000 (15:32 +0000)]
conf/machine: Drop older machines with no recent updates
These are all moving to meta-extras. Ideally in the future machines
such as these will be maintained to topic specific layers as we move
to a more layer oriented model. If this causes a problem for anyone
please discuss it on the mailing list.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Nitin A Kamble [Thu, 17 Feb 2011 00:39:21 +0000 (16:39 -0800)]
perl-native: fix parallel build
perl native recipe was failing on 40-way system with the parallel build turned
on. With this patch the parallel build on 40 way build system is not failing.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Passed 31 clean/build cycles on the 40-way.
Scott Garman [Wed, 16 Feb 2011 00:36:12 +0000 (16:36 -0800)]
opensp: improve configuration and install steps
* For the -native case we need to point two configure
options to the SGML catalogs in the native sysroot
* Some packages (e.g, iputils) use different names for
the openSP utilities, so create these names as
symlinks after do_install
Signed-off-by: Scott Garman <scott.a.garman@intel.com>
Lianhao Lu [Tue, 15 Feb 2011 11:13:19 +0000 (19:13 +0800)]
linux-libc-headers-yocto: Added RPROVIDES.
[BUGID #714] Added RPROVIDES_${PN}-dev and RPROVIDES_${PN}-dbg for
linux-libc-headers-yocto to provides "linux-libc-headers-dev" and
"linux-libc-headers-dbg" respectively.
This resolved the dependency issue of libc6-dev depending upon
linux-libc-headers-dev. Package linux-libc-headers-yocto-dev will be
installed as linux-libc-headers-dev.
Mark Hatle [Wed, 16 Feb 2011 04:56:13 +0000 (22:56 -0600)]
package_rpm: Fix solverdb generation
The RPM solverdb was potentially being generated multiple times.
Fix this by ensuring we only process each directory once.
Also correct an issue where the solution did not necessarily follow
the preferred architecture ordering, reverse the default Poky ordering
so that preferred is listed first.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Mark Hatle [Wed, 16 Feb 2011 04:43:42 +0000 (22:43 -0600)]
rpm: Avoid fsync when writing files
By default RPM strives for the safest possible file creation and
enforces fsync on each file. This changes that default by backporting
the fsync disabled from the latest CVS snapshot.
This will cut a few minutes off the rootfs generation using RPM.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Mark Hatle [Wed, 16 Feb 2011 04:20:49 +0000 (22:20 -0600)]
rpm: Add the ability to use the platform file during install
Add a new rpm macro, rpmrc_platform_path to specify an alternative platform
file. This is required to allow the dep resolver to identify compatible
packages.
Also workaround a minor problem with the --showrc command in RPM. A bug
has been reported upstream on this.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Bruce Ashfield [Tue, 15 Feb 2011 18:56:08 +0000 (13:56 -0500)]
linux-yocto: add jasperforest BSP and fix lttng NMI bug
During the introduction of the jasperforest BSP, a lttng
issues was picked up with NMI tracing. The upstream (and
our solution) is to revert that commit.
This change updates the meta repo for jasperforest and pushes
out the nmi revert to all BSPs.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Darren Hart [Mon, 14 Feb 2011 22:52:47 +0000 (14:52 -0800)]
README.hardware: update the beagleboard serial port documentation
Update the README.hardware beagleboard description to account for the use
of the OMAP_SERIAL in recent kernel recipes. Add a note documenting how to
use the old 8250 based serial devices for older kernels.
Correct a type.
Correct the boot script name from user.scr to boot.scr.
Signed-off-by: Darren Hart <dvhart@linux.intel.com> CC: Gary Thomas <gary@mlbassoc.com>
Bruce Ashfield [Mon, 14 Feb 2011 19:07:38 +0000 (14:07 -0500)]
linux-libc-headers-yocto: disable device tree generation
Fixes [BUGID #719]
The common routines for handling a git based yocto kernel
are included from the lib headers recipe to checkout the
appropriate branch of the kernel for header generation.
linux-yocto.inc includes device tree installation rules
which typically apply to a kernel user of the include file,
but do not apply to a simple header generation. The fix is
to override the DTB variables in this recipe, which disable
the device tree installation rules.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
