On 14 October 2016 at 19:38, Jussi Kukkonen
wrote:
> Let me know if do_configure_prepend() is not an appropriate place for
> a hack like this.
>
Anywhere is reasonable I guess, do_configure_prepend or do_install_append
will work just as well.
Ross
--
On 2016年10月13日 22:59, Burton, Ross wrote:
On 13 October 2016 at 15:40, Kang Kai > wrote:
gcc checks tsan and lsan support in source file
gcc/libsanitizer/configure.tgt, it seems only support x86_64. Then
tsan and lsan related
Actually, we just hit another failure mode running with these two patches.
Again, we're currently running on the jethro branch.
g_module_open() failed for
/poky/build/tmp/sysroots/x86_64-linux/usr/lib/gdk-pixbuf-2.0/2.10.0/loaders/libpixbufloader-jpeg.so:
libjpeg.so.9: cannot open shared object
On 10/13/2016 06:39 PM, Burton, Ross wrote:
On 13 October 2016 at 10:45, Robert Yang > wrote:
This is because "copyhardlinktree(deploy_arch_dir, arch_channel)" does:
"cp -afl deploy_arch_dir/* arch_channel", while the
On Fri, Oct 14, 2016 at 9:35 AM, Jussi Kukkonen
wrote:
>
>
> On 14 October 2016 at 06:22, Carol Zhu wrote:
>>
>> Hi Ross,
>>
>>
>>
>> So could we try integrating them to master branch and proving them
>> firstly?
>>
>> If they are all ok,
On Fri, Oct 14, 2016 at 6:54 AM, Khem Raj wrote:
> On Fri, Oct 14, 2016 at 9:35 AM, Jussi Kukkonen
> wrote:
>>
>>
>> On 14 October 2016 at 06:22, Carol Zhu wrote:
>>>
>>> Hi Ross,
>>>
>>>
>>>
>>> So could we try integrating
Fixed:
MACHINE = "qemumips64"
DEFAULTTUNE = "mips64-o32"
$ bitbake linux-yocto
ERROR: linux-yocto-4.8+gitAUTOINC+03bf3dd731_674818dad5-r0 do_package_qa: QA
Issue: Bit size did not match (32 to 64) linux-yocto on
Fixed:
MACHINE = "qemumips64"
DEFAULTTUNE = "mips64-o32"
$ bitbake linux-yocto
ERROR: OE-core's config sanity checker detected a potential misconfiguration.
Either fix the cause of this error or at your own risk disable the checker
(see sanity.conf).
Following is the list of potential
On 10/14/2016 05:58 PM, Robert Yang wrote:
On 10/13/2016 06:39 PM, Burton, Ross wrote:
On 13 October 2016 at 10:45, Robert Yang > wrote:
This is because "copyhardlinktree(deploy_arch_dir, arch_channel)" does:
"cp -afl
The following changes since commit d976454406adea2edd738096227ad53d600783ad:
source-highlight: use with-boost-libdir (2016-10-12 20:17:35 -0700)
are available in the git repository at:
git://git.openembedded.org/openembedded-core-contrib rbt/mips
On 14 October 2016 at 06:22, Carol Zhu wrote:
> Hi Ross,
>
>
>
> So could we try integrating them to master branch and proving them
> firstly?
>
> If they are all ok, then we can try next.
>
> They won’t block the RC1 of 2.2.
>
That would be the way to go,
This patch fixes an extramely rare race condition in creation of rpmdb
temporary directory. The "rpmdb-more-verbose-error-logging" patch is
still left in place, just for the case.
[YOCTO #9416]
Signed-off-by: Markus Lehtonen
---
On Fri, Oct 14, 2016 at 11:02 AM, Kang Kai wrote:
> On 2016年10月13日 22:59, Burton, Ross wrote:
>
>
> On 13 October 2016 at 15:40, Kang Kai wrote:
>>
>> gcc checks tsan and lsan support in source file
>> gcc/libsanitizer/configure.tgt, it seems only
Pascal,
On 10/14/2016 05:26 AM, Pascal Bach wrote:
Only relevant for krogoth since version 2.24+ (master, morty) is not affected.
Signed-off-by: Pascal Bach
---
meta/recipes-core/glibc/glibc/CVE-2016-1234.patch | 427 ++
This error is visible when using clang but not when using gcc
this has been reported and fixed upstream.
llvm bug https://llvm.org/bugs/show_bug.cgi?id=29017
binutils bug https://sourceware.org/bugzilla/show_bug.cgi?id=20648
Signed-off-by: Khem Raj
---
Only relevant for krogoth since version 2.24+ (master, morty) is not affected.
Signed-off-by: Pascal Bach
---
meta/recipes-core/glibc/glibc/CVE-2016-1234.patch | 427 ++
meta/recipes-core/glibc/glibc/CVE-2016-3075.patch | 37 ++
On 14 October 2016 at 14:21, Amarnath Valluri
wrote:
> meta-oe/recipes-benchmark/glmark2/glmark2_git.bb | 3 +++
>
This should go to openembedded-devel, not -core.
Ross
--
___
Openembedded-core mailing list
On Fri, Oct 14, 2016 at 3:21 PM, Amarnath Valluri
wrote:
> The failure was:
> | ../src/image-reader.cpp: In member function 'bool PNGReader::init(const
> string&)':
> | ../src/image-reader.cpp:123:16: error: 'unique_ptr' in namespace 'std' does
> not name a template
Hi Ross,
These gstreamer patches are bug fixing and all backported from higher
version, so the risk is low.
Is that ok to integrate them into morty?
B.R.
Carol
From: Burton, Ross [mailto:ross.bur...@intel.com]
Sent: Tuesday, October 11, 2016 8:26 PM
To: Carol Zhu
It's possible - albeit unlikely - that gdk-pixbuf isn't present in the sysroot
when a recipe inheriting this class is and the sysroot is finalised.
One example would be if the sstate archive has librsvg but not gdk-pixbuf:
librsvg will be extracted from the sstate but gdk-pixbuf will be built to
The eSDK generation assumes that DL_DIR is downloads/ under the build directory,
and puts files such as a freshly buily uninative tarball in there expecting
bitbake will find it later.
Whilst ${TOPDIR}/downloads/ is in fact the default value for DL_DIR in
bitbake.conf, and any instances of DL_DIR
From: Leonardo Sandoval
The poky-tiny distro points to linux-yocto-tiny instead of linux-yocto,
so setting the correct name depending on the distro.
[YOCTO #8525]
Signed-off-by: Leonardo Sandoval
---
From: Leonardo Sandoval
There are two bbtests that just makes sense on certain distros:
test_bitbake_g cannot be done on poky-tiny and test_non_gplv3
just makes sense on poky and poky-tiny. Skip these when
necessary under the latter conditions.
[YOCTO
From: Leonardo Sandoval
The test test_read_only_image builds core-image-sato which
is not buildable in poky-tiny so skip it if this is the case.
[YOCTO #8525]
Signed-off-by: Leonardo Sandoval
---
From: Leonardo Sandoval
core-image-clutter|weston can only be built on poky and poky-lsb, so skip
these tests in case other distro is defined. In the other hand, poky-tiny
does not support ext4 filesystem types, so skip test_bmap which includes
these
From: Leonardo Sandoval
The test defines several bitbake targets, including libx11, this one in
particular cannot be provided with a poky-tiny distro so skip it if this is
case.
[YOCTO #8525]
Signed-off-by: Leonardo Sandoval
This will allow for more flexibility and overrides in BSP
layers.
Signed-off-by: Saul Wold
---
meta/classes/image_types.bbclass | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/meta/classes/image_types.bbclass b/meta/classes/image_types.bbclass
index
From: Leonardo Sandoval
Originally, the test defined inside oeqa/selftest were defined for the 'poky'
distro. However, there are more supported distros, including nodistro, poky-lsb
and poky-tiny. These series takes into consideration the distribution,
From: Leonardo Sandoval
Distro poky-tiny cannot build the core-image-full-cmdline image, so skip the
test if this is the case.
[YOCTO #8525]
Signed-off-by: Leonardo Sandoval
---
From: Leonardo Sandoval
The poky-tiny distro does not supported the intended images to be created with
wic, so skp these tests.
[YOCTO #8525]
Signed-off-by: Leonardo Sandoval
---
From: Leonardo Sandoval
When distro is poky-tiny, there is no recipe providing libx11, thus skip
relevant
tests.
[YOCTO #8525]
Signed-off-by: Leonardo Sandoval
---
meta/lib/oeqa/selftest/devtool.py | 7
From: Leonardo Sandoval
Some tests builds images that poky-tiny cannot build so skip those. Also,
bitbake world fails if distro does not contain opengl, so skip those
tests if opengl is not present. One other change was made into the sstate
module:
The failure was:
| ../src/image-reader.cpp: In member function 'bool PNGReader::init(const
string&)':
| ../src/image-reader.cpp:123:16: error: 'unique_ptr' in namespace 'std' does
not name a template type
This change enables c++11 features, as code useing them.
Signed-off-by: Amarnath Valluri
The XML W3C conformance test suite contains thousands of xml files all
marked executable. We dutifully try to strip all these files of debug
info in do_package.
"chmod -x" improves build time by ~40 seconds.
Signed-off-by: Jussi Kukkonen
---
Let me know if
Modify packaging so that generated tests are available in two forms:
* piglit-generated-tests contains the tests as they are now (1.5GB
when installed)
* piglit-generated-tests-compressed contains a tar.gz with the same
files (45 MB when installed)
Add wrapper script that decompresses the
Piglit was removed from testapps in 9e227ecb3a because it made
sdk images massive. Add piglit back now that the installed size is
no longer humongous: it still needs 1.5 GB free space at runtime but
the image sizes stay reasonable.
Signed-off-by: Jussi Kukkonen
---
Piglit was removed because the current version was making sdk
images too large. This patchset reduces the installed size by 1.5GB
and adds piglit back to core-tools-testapps.
Running piglit still requires the same amount of space so in practice
this is useful for installed images only.
Q: Do I
37 matches
Mail list logo