<feed xmlns='http://www.w3.org/2005/Atom'>
<title>linux/meta-java.git/recipes-core, branch thud</title>
<subtitle>[no description]</subtitle>
<id>https://git.enea.com/cgit/linux/meta-java.git/atom?h=thud</id>
<link rel='self' href='https://git.enea.com/cgit/linux/meta-java.git/atom?h=thud'/>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/'/>
<updated>2019-12-27T10:35:59+00:00</updated>
<entry>
<title>Revert "base-files: Define JAVA_HOME by default"</title>
<updated>2019-12-27T10:35:59+00:00</updated>
<author>
<name>Fang Jia</name>
<email>fang.jia@windriver.com</email>
</author>
<published>2019-02-21T03:23:07+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=2eb86a8feffc98471f5fc3f9c20dba41844b74a5'/>
<id>urn:sha1:2eb86a8feffc98471f5fc3f9c20dba41844b74a5</id>
<content type='text'>
This reverts commit ada652564f09756a81f54495cbc2091125239625.

When enable multilib on a 64-bit system, probably both base-file and
lib32-base-files are installed.
Then both packages want to modify the file "profile", there will be
an error like:
"
Error: Transaction check error:
file /etc/profile conflicts between attempted installs of
lib32-base-files-3.0.14-r148.gpb_virt and
base-files-3.0.14-r148.gpb_virt
"

The task can be added by adding a file to ${sysconfdir}/profile.d/.

Signed-off-by: Fang Jia &lt;fang.jia@windriver.com&gt;
Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
<entry>
<title>openjdk-8: disable stringop-overflow compiler errors</title>
<updated>2019-12-27T10:35:26+00:00</updated>
<author>
<name>Richard Leitner</name>
<email>richard.leitner@skidata.com</email>
</author>
<published>2019-06-25T11:08:43+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=327c75c49d2fb294cd863e77810d014acfe0c31f'/>
<id>urn:sha1:327c75c49d2fb294cd863e77810d014acfe0c31f</id>
<content type='text'>
As new compilers generate this error we disable it for now as a
WORKAROUND. This will be removed when openjdk-8 is updated to the latest
version

Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
<entry>
<title>java-library: make packages overriding PACKAGE_ARCH work again</title>
<updated>2018-12-18T15:05:15+00:00</updated>
<author>
<name>André Draszik</name>
<email>andre.draszik@jci.com</email>
</author>
<published>2018-11-17T09:57:38+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=b0c74493af1415dfd3b6caf77ebc677ed4e0fdaa'/>
<id>urn:sha1:b0c74493af1415dfd3b6caf77ebc677ed4e0fdaa</id>
<content type='text'>
Recent changes is OE have caused the traditional approach of inheriting
allarch and setting PACKAGE_ARCH not work anymore. Once allarch is
inherited, PACKAGE_ARCH can not be overridden afterwards.
See commit a23c482cab4f ("allarch: only enable allarch when multilib
is not used") d9ba0219b2f6 in poky.
    http://git.openembedded.org/openembedded-core/commit/?id=a23c482cab4f874f4a6a6889716123569eb5ece9

The error manifests itself with configure trying to
   --host=allarch-poky-linux --target=allarch-poky-linux
which fails.

To work around this we can make java-library's allarch inherit
conditional, as is done e.g. in OE-core for packagegroup.bbclass
    http://git.openembedded.org/openembedded-core/commit/?id=9c826962ec8fa45c2b035427442b90a41517144e
    http://git.openembedded.org/openembedded-core/commit/?id=2c9b1d304daade7b0907320aeb9c522e7ab9dcab

So this commit does exactly that, and fixes the two users of this to
follow the new approach.

Signed-off-by: André Draszik &lt;andre.draszik@jci.com&gt;
Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
<entry>
<title>openjdk-8-native: hotspot: handle format-overflow error for gcc &gt;= 7</title>
<updated>2018-10-23T22:34:29+00:00</updated>
<author>
<name>Andreas Obergschwandtner</name>
<email>andreas.obergschwandtner@skidata.com</email>
</author>
<published>2018-10-23T08:03:10+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=0cd349d14cb844075d6126f2a3285be614c54c3f'/>
<id>urn:sha1:0cd349d14cb844075d6126f2a3285be614c54c3f</id>
<content type='text'>
fixed the format-overflow warnings by patch affected files in
openjdk-8-hotspot

Signed-off-by: Andreas Obergschwandtner &lt;andreas.obergschwandtner@skidata.com&gt;
Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
<entry>
<title>openjdk: remove incorrect PROVIDES (java2-vm java2-runtime)</title>
<updated>2018-10-22T13:15:12+00:00</updated>
<author>
<name>André Draszik</name>
<email>andre.draszik@jci.com</email>
</author>
<published>2018-09-18T14:48:44+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=1fdd55930690c3c7abd7458b795c26dd7bfebc08'/>
<id>urn:sha1:1fdd55930690c3c7abd7458b795c26dd7bfebc08</id>
<content type='text'>
There are two issues here:
* PROVIDES is not package specific
* it doesn't look like the intention is to be able to
  DEPENDS on java2-vm or java2-runtime anyway

Regarding java2-vm:
-------------------
java2-vm was originally used in the OpenJDK 6 &amp; 7 recipes to be able to
select shark, zero, cacao, or jamvm VMs. OpenJDK-6 is not available any
more, and OpenJDK-7 has removed support for most of this in commit
38f4c1365c11 ("openjdk7: remove broken/unsupported VM's")

It is not clear why it was added to the OpenJDK-8 recipe either.

Given OpenJDK-7 has no way of using the VM compiled as part of OpenJDK-8,
and given that no part of the OpenJDK-8 makes use of the java2-vm part, the
correct solution here is to actually *remove* the incorrect PROVIDES as
well as *R*PROVIDES statements for java2-vm completely.

Regarding java2-runtime:
------------------------
Again, looking at the other uses of this:
java2-runtime is a virtual runtime package name, which is provided by
different *runtime* packages (created by the OpenJDK-7, OpenJDK-8, JamVM, or
Cacao recipes). Other recipes only ever RDEPEND on java2-runtime. It makes
no sense for the OpenJDK-8 recipe to PROVIDES java2-runtime given the above.

Remove the incorrect (R)PROVIDES statements.

Signed-off-by: André Draszik &lt;andre.draszik@jci.com&gt;
Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
<entry>
<title>openjdk-8: support rm_work disabled</title>
<updated>2018-10-22T13:15:12+00:00</updated>
<author>
<name>André Draszik</name>
<email>andre.draszik@jci.com</email>
</author>
<published>2018-09-18T15:10:43+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=634ac5baf998abd256309cc41f5f8a24045de8a3'/>
<id>urn:sha1:634ac5baf998abd256309cc41f5f8a24045de8a3</id>
<content type='text'>
Now that generation of the Java keystore actually works (again?),
OE-core commit 09bb7718d74 ("ca-certificates: use relative symlinks from $ETCCERTSDIR"),
do_compile() will fail when invoked multiple times.

The reason is that during do_compile(), the Java keytool is
used to create a Java keystore with the certificates provided
by ca-certificates. Before above OE-core commit, no
certificates were actually being added, but as certificates
are being added now, multiple do_compile() runs will end up
adding the same certificate twice when rm_work is disabled,
causing a keytool exception, as that is not allowed.

So let's remove any previously generated keystore before trying
to add certificates to it.

Signed-off-by: André Draszik &lt;andre.draszik@jci.com&gt;
Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
<entry>
<title>openjdk-7-common: fix jvm postinst scripts</title>
<updated>2018-10-02T05:59:06+00:00</updated>
<author>
<name>Richard Leitner</name>
<email>richard.leitner@skidata.com</email>
</author>
<published>2018-09-12T06:45:19+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=7f9dfe70b5990e0c244bb3826d5a0c19ebd0e69d'/>
<id>urn:sha1:7f9dfe70b5990e0c244bb3826d5a0c19ebd0e69d</id>
<content type='text'>
As the JVM postinst scripts fail during rootfs creation we postpone them
to the target using the "pkg_postinst_ontarget_" function. Failing the
build at failing postinst scripts was intentionally introduced by
oe-core.

Furthermore remove the no longer needed $D prefixes from those postinst
scripts.

Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
<entry>
<title>icedtea: disable error format-overflow for gcc 7</title>
<updated>2018-10-02T05:59:06+00:00</updated>
<author>
<name>Andreas Obergschwandtner</name>
<email>andreas.obergschwandtner@skidata.com</email>
</author>
<published>2018-08-22T13:41:48+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=f83992acbe737a1d8cc87e73c8983b7fa2f64a1b'/>
<id>urn:sha1:f83992acbe737a1d8cc87e73c8983b7fa2f64a1b</id>
<content type='text'>
As no patch has been found in debian and hotspot repo for this issue we
just disable this warning which was introduced with GCC 7.

Also known as: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=881824

Signed-off-by: Andreas Obergschwandtner &lt;andreas.obergschwandtner@skidata.com&gt;
Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
<entry>
<title>openjdk-7: merge postinst into openjdk-7-common</title>
<updated>2018-10-02T05:59:06+00:00</updated>
<author>
<name>Richard Leitner</name>
<email>richard.leitner@skidata.com</email>
</author>
<published>2018-08-20T21:16:15+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=817b836aa0ddda2d6d3ddb481c7b2b17c67897a0'/>
<id>urn:sha1:817b836aa0ddda2d6d3ddb481c7b2b17c67897a0</id>
<content type='text'>
As openjdk-7-common was the only consumer of openjdk-postinst.inc move
it's content and remove openjdk-postinst.inc.

Now all files in recipes-core/openjdk are named correctly according to
their versions again.

Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
<entry>
<title>openjdk-8: use openjdk-common</title>
<updated>2018-10-02T05:59:06+00:00</updated>
<author>
<name>Richard Leitner</name>
<email>richard.leitner@skidata.com</email>
</author>
<published>2018-08-20T21:16:14+00:00</published>
<link rel='alternate' type='text/html' href='https://git.enea.com/cgit/linux/meta-java.git/commit/?id=50d3e814a45363db81e446233c54895eebc5a999'/>
<id>urn:sha1:50d3e814a45363db81e446233c54895eebc5a999</id>
<content type='text'>
As openjdk-common.inc now serves all OpenJDK version let
openjdk-8-common require it. Furthermore remove the now duplicated
lines.

Signed-off-by: Richard Leitner &lt;richard.leitner@skidata.com&gt;
</content>
</entry>
</feed>
