Qing He [Thu, 17 Jun 2010 09:23:20 +0000 (17:23 +0800)]
syslinux: upgrade to versin 3.86
from version 3.36
partly based on the recipe from openembedded project
changes:
- add cross-build.patch to use poky toolchain
- mangle LDFLAGS so $LD invocation works
- use precompiled version of $BTARGET for target (OE)
- stick on ${sbindir} instead of OE's ${base_sbindir}
Qing He [Fri, 11 Jun 2010 05:02:35 +0000 (13:02 +0800)]
zlib: upgrade to version 1.2.5
from 1.2.3
update homepage and license
changes:
- remove 1.2.3.3 patch
- breakdown autotools.patch to configure.ac and Makefile.am, update
src in Makefile.am and use do_configure_prepend instead
- remove README from LIC_FILES_CHKSUM because it duplicates zlib.h
and also causes checksum to change for every release
- remove visibility.patch: the purpose of the patch was to change
"<type> ZEXPORT" to "ZEXPORT <type>", however:
1. the use of ZEXPORT and ZEXTERN have been changed a lot in
zlib, there is no changelog to indicate which version was
the patch based on and its rationale
2. zlib used to set default visibility to `hidden'
(-fvisibility=hidden), and override external symbols with
`default' visibility, it had caused a lot of problems on
different platforms. But starting from 1.2.5, default
visibility is now `default', `hidden' visibility override
has effect only for the correct GCC versions, most visibility
related issues should have already been fixed
3. as for now, ZEXPORT is used in Win/BeOS only, and can expand
to WINAPI or __declspec(dllexport), both of which should be
put behind return type, so there is no need to patch
In conclusion, the patch is not applicable for the current release,
so it's removed
Dongxiao Xu [Sat, 26 Jun 2010 06:51:07 +0000 (14:51 +0800)]
gstreamer: Upgraded to version 0.10.29
Rebase check_fix.patch
Remove po-makefile-fix.patch since it is not being used
Remove do_configure_prepend since ${S}/common/m4/lib-link.m4 is not exist in the latest version
Change the local file directory name to ${PN}-${PV}
Dongxiao Xu [Sat, 26 Jun 2010 09:43:55 +0000 (17:43 +0800)]
lame: Upgraded to 3.98.4
Removed Makefile-lm.patch since the code is not contained in latest version
Rebased no-gtk1.patch and commented for it
Change the local file directory name to ${PN}-${PV}
Richard Purdie [Wed, 30 Jun 2010 08:48:59 +0000 (09:48 +0100)]
classes/package_ipk|_deb.bbclass: Make sure versions correctly make it into output packages
The version information part of package dependency strings was not ending up
in the output packages correctly. This patch fixes this and ensures the version
information isn't lost.
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
Kevin Tian [Fri, 25 Jun 2010 09:48:52 +0000 (17:48 +0800)]
iproute: upgrade to 2.6.34
(borrow from OE 2.6.34 recipe)
[Patches-poky]
REMOVE _iproute2-2.6.15_no_strip.diff_: not required now
REMOVE _new-flex-fix.patch_: not required now
[Patches-OE]
TAKE _configure-cross.patch_: don't invoke gcc directly
[Recipes]
- Add metadata
- fix SRC_URI for new version
- no {S} now
- use alternative to enable 'ip' instead of using busybox
- move shared documents to right destination
- not take from OE:
* no need for specific do_configure
Kevin Tian [Wed, 23 Jun 2010 07:25:39 +0000 (15:25 +0800)]
blktool: upgrade to 4-6
Unlike the big jump in version (4 -> 4-6), blktool is actually quite stable since
version 4 (2005). All the changes in 4-6 are about typo fix and debian package
control enhancement.
Kevin Tian [Tue, 15 Jun 2010 22:37:17 +0000 (06:37 +0800)]
pax-utils: upgrade to 0.2.1
- update metadata. license checksum is skipped, as pax-utils is depended by do_package task itself.
I.e. the 1st do_package will be run after pax-utils-native is built. So far checksum verification
is included in do_package task
- no need to override {S} since {S} uses {BPN}
- no need to define "_GNU_SOURCE" which has been included by source
tar ball. Also remove do_compile override accordingly
- remove recipe internal md5sum, which is not used by Poky
Jeff Dike [Tue, 29 Jun 2010 12:33:24 +0000 (13:33 +0100)]
bitbake: create cooker object after cleaning the environment
Previously, the cooker object was created before the environment was
cleaned, saving everything that was in the environment and dumping
into the run scripts.
The patch ensures that the cooker gets a cleaned environment when it's
created.
Signed-off-by: Jeff Dike <jdike@linux.intel.com> Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
Jeff Dike [Tue, 29 Jun 2010 12:32:04 +0000 (13:32 +0100)]
bitbake: Move the logger initialization from cooker to utils
In order to move the environment cleaning, which wants to log a
message, before cooker creation, the logging facility initialization
needs to happen earlier. So, it is now in init_logger in utils.py and
called from bitbake before the creation of the data object.
It also seems more correct to initialize a global facility like this
from a more global context than the creation of an object, of which
there could theoretically be many.
Signed-off-by: Jeff Dike <jdike@linux.intel.com> Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
Kevin Tian [Mon, 21 Jun 2010 08:08:57 +0000 (16:08 +0800)]
qemux86/xorg.conf: no DefaultDepth for VMware SVGA driver
VMware SVGA driver needs to have same depth between the host and the guest. Or put in
other word, the depth read by the guest is the value read from host. The guest is not
allowed to change virtual depth to other value. With DefaultDepth option xorg.conf,
vmware driver rejects to work with suggestion "Please do not specify a depth on the
command line or via the config file".
Kevin Tian [Tue, 29 Jun 2010 00:54:03 +0000 (08:54 +0800)]
qemu: fix VMware VGA depth calculation error
VMware SVGA presents to the guest with the depth of the host surface it renders
to, and rejects to work if the two sides are mismatched. One problem is that
current VMware VGA may calculate a wrong host depth, and then memcpy from virtual
framebuffer to host surface may trigger segmentation fault. For example, when
launching Qemu in a VNC connection, VMware SVGA thinks depth as '32', however the
actual depth of VNC is '16'. The fault also happens when the host depth is not
32 bit.
Qemu <4b5db3749c5fdba93e1ac0e8748c9a9a1064319f> tempts to fix a similar issue, by
changing from hard-coded 24bit depth to instead query the surface allocator
(e.g. sdl). However it doesn't really work, because the point where query
is invoked is earlier than the point where sdl is initialized. At query time,
qemu uses a default surface allocator which, again, provides another hard-coded
depth value - 32bit. So it happens to make VMware SVGA working on some hosts,
but still fails in others.
To solve this issue, this commit introduces a postcall interface to display
surface, which is walked after surface allocators are actually initialized.
At that point it's then safe to query host depth and present to the guest.
Richard Purdie [Mon, 28 Jun 2010 23:35:02 +0000 (00:35 +0100)]
gcc: Add patch to allow disabling of libstdc++ linkage and hence fix gcc-runtime which was having broken configure tests due to the linker failures and assuming maths primitives were not in libm
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
Joshua Lock [Tue, 22 Jun 2010 12:20:08 +0000 (13:20 +0100)]
pkgconfig: add patch to disable legacy scripts such as glib-config
On an F13 host with glib-config installed pkgconfig-native can get into a
horrible state with recursive calls between pkg-config and glib-config.
The patch adds a configure time option to disable legacy script support in
pkgconfig and makes use of the option for Poky.
Joshua Lock [Fri, 25 Jun 2010 12:11:39 +0000 (13:11 +0100)]
cross-canadian: ensure package dependencies are generated correctly
cross-canadian packages need to look for their SOLIBS in the nativesdk
sysroot so that dependencies are correctly picked up and meta-toolchains are
correctly built.
