]> code.ossystems Code Review - openembedded-core.git/log
openembedded-core.git
7 years agowic: partition plugin wrongly assumes it is rootfs
Paulo Neves [Sun, 25 Feb 2018 16:49:25 +0000 (17:49 +0100)]
wic: partition plugin wrongly assumes it is rootfs

The partition plugin is used as the base for other plugins.
One of the methods the plugins use, is the prepare_rootfs
method.

The prepare_rootfs method wrongly assumes that the value
ROOTFS_SIZE from bitbake datastore is relevant to every
invocation of prepare_rootfs, which it clearly is not, for
example in the bootimg-partition case.

This commit adds an optional argument to prepare_rootfs
where a caller can tell prepare_rootfs if it is an actual
rootfs and whether related rootfs information retrieved from
bitbake is valid. The default behavior of this optional
argument is to assume that the invocation is an actual
rootfs, to maintain compatibility with previous
implementations.

Signed-off-by: Paulo Neves <ptsneves@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agowic: Add post operation hook for SourcePlugin
Parthiban Nallathambi [Tue, 20 Feb 2018 15:49:27 +0000 (16:49 +0100)]
wic: Add post operation hook for SourcePlugin

do_post_partition hook is needed if some operations like security signing
the parition needs to be done. source plugins can make use of this to implement
post operatiosn in do_post_partition. do_post_partition is called after
do_prepare_partition if present.

Signed-off-by: Parthiban Nallathambi <pn@denx.de>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agokernel-yocto: check git config during checkout
Justin Bronder [Fri, 9 Feb 2018 21:05:52 +0000 (16:05 -0500)]
kernel-yocto: check git config during checkout

Initialize the git config prior creating a git repository in a plain
directory.

Signed-off-by: Justin Bronder <jsbronder@cold-front.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agopkgconf: revert changes to pkg-config-wrapper
Maxin B. John [Mon, 26 Feb 2018 08:58:48 +0000 (10:58 +0200)]
pkgconf: revert changes to pkg-config-wrapper

To maintain the status quo, revert changes to pkgconf wrappers.
This helps to keep it compatible with freedesktop.org pkg-config.

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibmpc: upgrade to 1.1.0
Oleksandr Kravchuk [Sat, 24 Feb 2018 23:19:39 +0000 (00:19 +0100)]
libmpc: upgrade to 1.1.0

Signed-off-by: Oleksandr Kravchuk <open.source@oleksandr-kravchuk.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agogstreamer-rtsp-server: remove unnecessary dependencies
Anuj Mittal [Sat, 24 Feb 2018 11:53:03 +0000 (19:53 +0800)]
gstreamer-rtsp-server: remove unnecessary dependencies

libcheck dep was removed from gst-rtsp-server some time back [1].

Disable building of examples and tests and remove dep on libcgroup which
was needed, if present, to build only one specific example.

[1] https://cgit.freedesktop.org/gstreamer/gst-rtsp-server/commit?id=43ec4e7c786227b85aec06599c3955626e241a01

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agogcc: Fix internal compiler error for PPC test case "gcc.dg/vmx/7d-02.c"
Mark Hatle [Fri, 23 Feb 2018 17:41:24 +0000 (12:41 -0500)]
gcc: Fix internal compiler error for PPC test case "gcc.dg/vmx/7d-02.c"

Fix an internal compiler error on PPC from building a specific test:

$ $CC -S 7d-02.c
7d-02.c: In function â:
7d-02.c:11:5: internal compiler error: in copy_to_mode_reg, at explow.c:612
     vec_st(v, i*16,p);
     ^~~~~~

The failure appears to happen on all optimizations levels as well.

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agogcc: Fix test case issue when SSE is not enabled
Mark Hatle [Fri, 23 Feb 2018 17:41:23 +0000 (12:41 -0500)]
gcc: Fix test case issue when SSE is not enabled

Whenever "-mno-sse" is used, "-mfpmath" should be set to 387.

The test case should be modified accordingly as below:

/* { dg-additional-options "-mno-sse -mfpmath=387 " { target { i?86-*-* x86_64-*-* } } } */

Original patch from: RAGHUNATH LOLUR <raghunath.lolur@kpit.com>

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agocmake: refactor compile and install for easier re-use
André Draszik [Fri, 23 Feb 2018 14:50:13 +0000 (14:50 +0000)]
cmake: refactor compile and install for easier re-use

cmake_do_compile() and cmake_do_install() basically do the
same, except they use a different --target, and at the
moment this is copy/pasted code with a minor modification.

Other recipes which e.g. might want to support compilation
as part of ptest have to do the same. This is a bit
inconvenient.

By factoring out all of this into a common helper,
cmake_runcmake_build(), this is easily re-used. An
(imaginary) recipe can compile ptest support simply by
using
  cmake_runcmake_build --target buildtest-TESTS
(assuming such a build target exists).

Also, this now is very similar to oe_runmake().

Signed-off-by: André Draszik <andre.draszik@jci.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoicecc: Remove several getVar() expand arguments
Joshua Watt [Thu, 22 Feb 2018 22:59:31 +0000 (16:59 -0600)]
icecc: Remove several getVar() expand arguments

Several of the calls to getVar() were either superfluously passing
True for the expand argument, or were wrongly passing False

Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agou-boot-mkimage: Add dependency on dtc
Otavio Salvador [Thu, 22 Feb 2018 20:54:57 +0000 (17:54 -0300)]
u-boot-mkimage: Add dependency on dtc

The dtc is used when generating images which use Device Tree and we
must use the OE version to avoid relying on the host one.

Reported-by: Renato Caldas <rm.santos.caldas@gmail.com>
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoopkg-utils: upgrade to version 0.3.6
Alejandro del Castillo [Wed, 28 Feb 2018 20:36:45 +0000 (14:36 -0600)]
opkg-utils: upgrade to version 0.3.6

0.3.5 -> 0.3.6

Patches:
 - Add support for tar versions that don't support --sort
 - Use local time when setting the modication time on the archives

Signed-off-by: Alejandro del Castillo <alejandro.delcastillo@ni.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoopkg: upgrade to version 0.3.6
Alejandro del Castillo [Thu, 22 Feb 2018 19:57:35 +0000 (13:57 -0600)]
opkg: upgrade to version 0.3.6

0.3.5 -> 0.3.6

Signed-off-by: Alejandro del Castillo <alejandro.delcastillo@ni.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agogit: 2.15.0 -> 2.16.1
Robert Yang [Thu, 22 Feb 2018 09:19:42 +0000 (17:19 +0800)]
git: 2.15.0 -> 2.16.1

Add /usr/share/git-core/templates/hooks/fsmonitor-watchman.sample to PERLTOOLS to fix:
ERROR: git-2.16.1-r0 do_package_qa: QA Issue: /usr/share/git-core/templates/hooks/fsmonitor-watchman.sample contained in package git requires /usr/bin/perl, but no providers found in RDEPENDS_git? [file-rdeps]
ERROR: git-2.16.1-r0 do_package_qa: QA run found fatal errors. Please consider fixing them.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agokbd: build tests only when ptest is enabled
Anuj Mittal [Thu, 22 Feb 2018 09:11:14 +0000 (17:11 +0800)]
kbd: build tests only when ptest is enabled

Also include the libcheck dependency only when tests are to be built.

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agobluez5: remove dependency on libcheck
Anuj Mittal [Thu, 22 Feb 2018 05:45:50 +0000 (13:45 +0800)]
bluez5: remove dependency on libcheck

It was removed a while back and isn't needed anymore.

https://git.kernel.org/pub/scm/bluetooth/bluez.git/commit/?id=292a37990cf0b9517135a4d58c8c1df6b9123cb4

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoglibc: Enable static PIE support when security_flags are enabled
Khem Raj [Thu, 1 Mar 2018 18:26:34 +0000 (18:26 +0000)]
glibc: Enable static PIE support when security_flags are enabled

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agosdk: generate locale archive and remove packages
Ross Burton [Thu, 1 Mar 2018 18:26:33 +0000 (18:26 +0000)]
sdk: generate locale archive and remove packages

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agopopulate_sdk: install UTF-8 locales in SDKs
Ross Burton [Thu, 1 Mar 2018 18:26:32 +0000 (18:26 +0000)]
populate_sdk: install UTF-8 locales in SDKs

As glibc 2.27 can't read older locale-archives, SDKs using glibc 2.27 on hosts
using glibc earlier than 2.27 won't be able to find any locales, so bitbake
won't start and Python can't use UTF-8.

So by default install all locales into the SDK.  Special-case Extensible SDKs by
installing no locales as they ship glibc in a buildtools, and that will have the
locales.

Locale installation requires cross-localedef, so add that to DEPENDS.

Also remove the explicit en_US addition in buildtools-tarball as it is now
redundant.

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agosdk: install specified locales into SDK
Ross Burton [Thu, 1 Mar 2018 18:26:31 +0000 (18:26 +0000)]
sdk: install specified locales into SDK

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agopackage-manager: add install_glob()
Ross Burton [Thu, 1 Mar 2018 18:26:30 +0000 (18:26 +0000)]
package-manager: add install_glob()

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agopackage_manager: improve install_complementary
Ross Burton [Thu, 1 Mar 2018 18:26:29 +0000 (18:26 +0000)]
package_manager: improve install_complementary

- No need to use bb.utils.which() as subprocess will search $PATH
- Clarity flow by moving the install inside the try/except

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agocross-localedef-native: add way to specify which locale archive to write
Ross Burton [Thu, 1 Mar 2018 18:26:28 +0000 (18:26 +0000)]
cross-localedef-native: add way to specify which locale archive to write

localedef has no way to specify which locale archive to use, and the
compile-time default isn't useful as it points to the work directory.

Add support to read an environmental variable for the path, and don't fail to
write a new locale archive.

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agoglibc: relocate locale paths in nativesdk
Ross Burton [Thu, 1 Mar 2018 18:26:27 +0000 (18:26 +0000)]
glibc: relocate locale paths in nativesdk

nativesdk is built with a specific prefix but this will be different at install
time, however glibc hard-codes the path to locale files. Expand these strings to 4K and move them to a magic segment which we can relocate when the SDK is installed.

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agoglibc: don't use host locales in nativesdk
Ross Burton [Thu, 1 Mar 2018 18:26:26 +0000 (18:26 +0000)]
glibc: don't use host locales in nativesdk

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agodefault-distrovars: don't rename locales for nativesdk
Ross Burton [Thu, 1 Mar 2018 18:26:25 +0000 (18:26 +0000)]
default-distrovars: don't rename locales for nativesdk

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agoglibc: Upgrade to 2.27 release
Ross Burton [Thu, 1 Mar 2018 18:26:24 +0000 (18:26 +0000)]
glibc: Upgrade to 2.27 release

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agodbus-test_1.12.2: various fixes
Juro Bystricky [Tue, 27 Feb 2018 20:19:26 +0000 (12:19 -0800)]
dbus-test_1.12.2: various fixes

The result of running dbus-test-ptest was a series of
various segfaults, interpreted as FAILs. This was a direct consequence
of the test suite loading the installed shared library libdbus-1.so, not the
one built along the test suite.

While we normally want to test against the installed libraries, we cannot
do this in this case as the test suite expects a library that is configured/compiled
differently from the installed one. We could configure the installed library
identically as the test suite expects, (and there should be no issues), however
this is not desirable for performance reasons.

Hence we need to use the library built along with the test suite.
Of course, running the test suite against its own library does not
test the installed library, however they are both built from the same
sources so that can give us some kind of indication.

The following changes were made:

1. Configure the test library as close as possible to the installed one,
   with some additional configuration options that are needed for testing.
   (Use dbus_1.12.2.bb recipe as a template)
2. Include the shared libraries in the package, use LD_LIBRARY_PATH during
   testing to load them instead of the installed ones.
3. Add a few more tests. (There are still some additional tests built that
   are not used, but they would have to be special-cased).
4. When evaluating the test results, differentiate between "FAIL" and "SKIP"

[YOCTO #10841]
[YOCTO #12277]

(From OE-Core rev: 5d148aa9c3c338fabab1e60e2ca64d09c9b8477f)

Signed-off-by: Juro Bystricky <juro.bystricky@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agopackage.bbclass: fetch PRIVATE_LIBS twice as we iterate twice
Ross Burton [Wed, 28 Feb 2018 23:06:30 +0000 (23:06 +0000)]
package.bbclass: fetch PRIVATE_LIBS twice as we iterate twice

The shlibs detection/handling iterates the package list twice, but PRIVATE_LIBS
is only fetched in the first loop which means the second loop only considers the
value set.

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agometa-world-pkgdata: This recipe is machine specific, mark as such
Richard Purdie [Wed, 28 Feb 2018 11:22:37 +0000 (11:22 +0000)]
meta-world-pkgdata: This recipe is machine specific, mark as such

The recipe depends on many machine specific tasks and should be marked
as machine specific itself. This fixes signature tests after some dependency
issues were fixed at the bitbake level which exposed that issue.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agodevtool: search: tweak help text
Paul Eggleton [Mon, 26 Feb 2018 01:50:01 +0000 (14:50 +1300)]
devtool: search: tweak help text

* We now match on more than just target recipes, so don't specify that
  only target recipes are searched.
* We're printing the SUMMARY value in addition to the name, so mention
  that so it's clear where that text is coming from.
* Remind users that they should use quotes around the keyword to avoid
  shell expansion when using regular expressions.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agodevtool: search: also look in recipe cache
Paul Eggleton [Mon, 26 Feb 2018 01:50:00 +0000 (14:50 +1300)]
devtool: search: also look in recipe cache

If pkgdata isn't present or is incomplete, then you get either a
traceback or you don't see the results you were hoping for. The recipe
cache that bitbake collects during startup contains some useful
information for each recipe that we could search through as well, and
we can access it easily using tinfoil's all_recipes() API function,
so add some code that does that. (We still show a warning if pkgdata
isn't present, as there are certain dynamic packages that are generated
at packaging time that won't show up in the cache).

One side-effect of this is that we will start showing non-target
recipes - that's actually a good thing, since seeing those is useful,
however we exclude nativesdk recipes when in the eSDK to avoid confusion
since nativesdk isn't directly applicable there.

Fixes [YOCTO #12356].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agolib/oe/recipeutils: add .txz extension to archive list
Paul Eggleton [Mon, 26 Feb 2018 01:49:59 +0000 (14:49 +1300)]
lib/oe/recipeutils: add .txz extension to archive list

Prompted by bitbake commit 2ba8a6b25ccc12e7b543e8450121e5311c7a701d,
add .txz to the list of archives used within get_recipe_local_files()
here as well.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agodevtool: deploy-target: don't specify ssh/scp port unless user does
Paul Eggleton [Mon, 26 Feb 2018 01:49:58 +0000 (14:49 +1300)]
devtool: deploy-target: don't specify ssh/scp port unless user does

If the user doesn't specify a port then we should avoid specifying one
on the ssh/scp command line in case the user has configured one for the
host they are connecting to, which was being overridden unnecessarily.

Fixes [YOCTO #12381].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agodevtool: finish: fix erroneously creating bbappend for relative paths
Paul Eggleton [Mon, 26 Feb 2018 01:49:57 +0000 (14:49 +1300)]
devtool: finish: fix erroneously creating bbappend for relative paths

After OE-Core rev 5e3fe00a0233d563781849a44f53885b4e924a9c we call
os.path.abspath() on the original layer path, but we later compare that
to the destination layer path. If that layer path isn't absolute but is
effectively the same path, it should be writing to the original recipe
but because we weren't making it absolute we were writing a bbappend
instead. Call os.path.abspath() on the destination path as well to avoid
that.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agodevtool: reset: delete bbappend file if _check_preserve() doesn't
Paul Eggleton [Mon, 26 Feb 2018 01:49:56 +0000 (14:49 +1300)]
devtool: reset: delete bbappend file if _check_preserve() doesn't

If the .devtool_md5 file doesn't contain a reference to the bbappend
file (e.g. because devtool was interrupted before it could write that
out) then _check_preserve() won't delete it, so we need to delete it
separately because otherwise the recipe won't actually be reset.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agodevtool: fix poor handling of upgraded BBCLASSEXTENDed recipes
Paul Eggleton [Mon, 26 Feb 2018 01:49:55 +0000 (14:49 +1300)]
devtool: fix poor handling of upgraded BBCLASSEXTENDed recipes

Fix two aspects of handling BBCLASSEXTENDed targets (e.g.
openssl-native) that have been run through "devtool upgrade":

* Fix recipe name not showing up in "devtool status"
* Fix "devtool reset" not deleting empty directories under the recipe
  directory within the workspace, which may lead to problems if you
  subsequently run "devtool upgrade" on the same target again

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agolib/oe/path: implement is_path_parent()
Paul Eggleton [Mon, 26 Feb 2018 01:49:54 +0000 (14:49 +1300)]
lib/oe/path: implement is_path_parent()

In a few places we have checks to see path B is the parent of path A, by
adding / to the end of the path B and then seeing if path A starts with
the suffixed path B. Unfortunately there are two potential flaws:
(1) path A needs to be suffixed with / as well or the directory itself
won't match (semantics perhaps, but in a lot of scenarios returning True
is correct); (2) you need to run os.path.abspath() on both paths first
or you will wrongly return False for some relative paths where you
should return True. Let's solve this once and for all by writing a
function that takes care of these and put it in oe.path.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agooe-init-env: search 'bitbake' outside of oe-core
Enrico Scholz [Tue, 27 Feb 2018 18:46:48 +0000 (19:46 +0100)]
oe-init-env: search 'bitbake' outside of oe-core

atm, 'oe-init-build-env' expects 'bitbake' to be within the OE core git
repository.  This complicates the project setup because you have to
manage the 'bitbake' directory or symlink manually (or specify the
bitbake location explicitly).

Looking for 'bitbake' outside the main git repository will ease project
management significantly.  Now, you can put everything into git submodules,
clone the project with

| git clone --recursive ...

and continue immediately with

| ..../oe-init-build-env

E.g. when you had previously

| .
| |-- build/
| `-- sources/
|     `-- org.openembedded.core/
|         `-- bitbake -> ../bitbake

(where 'bitbake' must be created manually after cloning the project),
you can have now

| .
| |-- build/
| `-- sources/
|     |-- bitbake/
|     `-- org.openembedded.core/

which is completely managed by 'git'.

Patch adds $OEROOT/.. to the search path for 'bitbake'

Signed-off-by: Enrico Scholz <enrico.scholz@sigma-chemnitz.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agosstatesig: Move hardcoded native tools
Joshua Watt [Tue, 27 Feb 2018 22:47:53 +0000 (16:47 -0600)]
sstatesig: Move hardcoded native tools

Now that SIGGEN_EXCLUDE_SAFE_RECIPE_DEPS supports the wildcard syntax,
these exclusions no longer need to be hardcoded

Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agosstatesig: Add recipe wildcard
Joshua Watt [Tue, 27 Feb 2018 22:47:51 +0000 (16:47 -0600)]
sstatesig: Add recipe wildcard

The special string "*" on the left hand side of the dependency
specification matches all recipes except the one on the right hand side.

Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agolinux-firmware: fix typo
Nicolas Dechesne [Tue, 27 Feb 2018 06:10:07 +0000 (07:10 +0100)]
linux-firmware: fix typo

Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agosstatesig: Ignore icecc-create-env in task sigs
Joshua Watt [Tue, 27 Feb 2018 03:01:35 +0000 (21:01 -0600)]
sstatesig: Ignore icecc-create-env in task sigs

Changes to the icecc-create-env recipe should not cause all recipes to
rebuild just because the have inherited icecc.

Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agoicecc.bbclass: Ignore more icecream vars in hashes
Joshua Watt [Tue, 27 Feb 2018 21:52:45 +0000 (15:52 -0600)]
icecc.bbclass: Ignore more icecream vars in hashes

Changing ICECC_ENV_VERSION or select variables that the user can set to
control the behavior of icecc should not cause recipes to rebuild

Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agopuzzles: upgrade to latest revision
Maxin B. John [Tue, 27 Feb 2018 13:16:16 +0000 (15:16 +0200)]
puzzles: upgrade to latest revision

License-Update: checksum change is due to standardization of
character encoding of source tree on UTF-8

1. Remove upstreamed patch:
        0001-signpost-Add-paranthesis-to-remove-doubt-about-prece.patch
2. Bump PE to prevent version going backwards

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agopiglit: upgrade to current git master
Maxin B. John [Tue, 27 Feb 2018 15:40:41 +0000 (17:40 +0200)]
piglit: upgrade to current git master

Upgrade to revision:
4ce0887e2f7f848d2be2e435a2d0f3c80e44ea3b

remove backported patch:
        0001-tests-egl-egl-context-priority.c-Use-piglit_egl_get_.patch

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agowebkit: add missing Upstream-Status to patch
Maxin B. John [Tue, 27 Feb 2018 15:40:40 +0000 (17:40 +0200)]
webkit: add missing Upstream-Status to patch

Add missing Upstream-Status to the following patch:
        0001-WebKitMacros-Append-to-I-and-not-to-isystem.patch

Upstream-Status is fetched from the first commit of the patch:
commit id: 700b846ba0e0f31b06459aee3ebb3ce5f1715918

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agosstatesig/staging/package_manager: Create common sstate manifest code
Richard Purdie [Tue, 27 Feb 2018 17:22:32 +0000 (17:22 +0000)]
sstatesig/staging/package_manager: Create common sstate manifest code

Create a common function for locating task manifest files rather than
several implementations with missing pieces.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agopackage_manager: Filter to only rpms we depend upon
Richard Purdie [Mon, 18 Sep 2017 16:51:58 +0000 (17:51 +0100)]
package_manager: Filter to only rpms we depend upon

Currently do_rootfs gets to see all rpms in the deploy directory. This filters
that view to only rpms which the image recipe has actual depends upon which
potentially removes some sources of confusion in the image construction.

This makes builds more reproducibile and also fixes contamination issues
where dnf picks up packages it shouldn't be able to 'see'.

[YOCTO #12039]

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agobuildhistory-diff: honour report_all flag
Anuj Mittal [Fri, 23 Feb 2018 10:55:59 +0000 (18:55 +0800)]
buildhistory-diff: honour report_all flag

Make sure that we're passing a bool value. Without this, buildhistory
shows all the output for all the keys/fields when it shouldn't be by
default.

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agolinux-yocto-dev: update to v4.16-rcX
Bruce Ashfield [Thu, 22 Feb 2018 15:03:12 +0000 (10:03 -0500)]
linux-yocto-dev: update to v4.16-rcX

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
7 years agosstatetests: pass the right files when comparing sigfiles
Ross Burton [Thu, 22 Feb 2018 12:47:20 +0000 (12:47 +0000)]
sstatetests: pass the right files when comparing sigfiles

Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agotime: 1.7 -> 1.8
Robert Yang [Thu, 22 Feb 2018 09:19:43 +0000 (17:19 +0800)]
time: 1.7 -> 1.8

* Remove debian.patch which is already in the source.
* License-Update: The license is changed to GPLv3, and move v2 one to meta-gplv2.
* Merge time.inc into time_1.8.bb.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agobinutils: Upgrade to 2.30 release
Khem Raj [Wed, 21 Feb 2018 03:12:48 +0000 (19:12 -0800)]
binutils: Upgrade to 2.30 release

Additionally cherry-pick

5ffeba4a09 Revert "PowerPC PLT speculative execution barriers"
b01452b1d4 [PR22764][LD][AARCH64]Allow R_AARCH64_ABS16 and R_AARCH64_ABS32 against absolution symbol or undefine symbol in shared object.
a985e9b9de Import patch from mainline to remove PROVODE qualifiers around definitions of __CTOR_LIST__ and __DTOR_LIST__ in PE linker scripts.
eec4607fc5 Add support for DWARF-4 line number tables.

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoglibc: Update to tip of 2.26
Khem Raj [Wed, 21 Feb 2018 03:12:49 +0000 (19:12 -0800)]
glibc: Update to tip of 2.26

This will make it easy to backport to rocko if needed after 2.27 is landed in master
plus it fixes the aarch64 build issue seen with binutils 2.30

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibva-utils: upgrade to 2.1.0
Anuj Mittal [Wed, 21 Feb 2018 04:30:16 +0000 (12:30 +0800)]
libva-utils: upgrade to 2.1.0

Major changes:

* Refine gtest conformance cases
* vp9enc: add support low power mode
* vavpp: add support for RGBA/RGBX surface
* vainfo: add support new profile/entrypoint pairs

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibva: upgrade to 2.1.0
Anuj Mittal [Wed, 21 Feb 2018 04:30:15 +0000 (12:30 +0800)]
libva: upgrade to 2.1.0

Major changes:

* Bump VA-API version to 1.1.0
* Add API for multi-frame processing
* Add entrypoint VAEntrypointStats for Statistics
* Add data structures for HEVC FEI support
* Add new attributes for decoding/encoding/video processing
* Add new VPP filter for Total Color Correction
* Add blending interface in VPP
* Add rotation interface in VPP
* Add mirroring interface in VPP
* Add Chroma siting flags in VPP
* Add new color standard definitions
* Add new interface for exporting surface
* Add message callbacks for drivers to use

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agologging.bbclass: Enclose the tr string in quotes
Jason Wessel [Tue, 20 Feb 2018 23:53:29 +0000 (15:53 -0800)]
logging.bbclass: Enclose the tr string in quotes

On some linux hosts image recipes will fail to build as follows:

ERROR: build-essential-0.3-r0 do_image_ext3: Usage: bbdebug [123] "message"
ERROR: build-essential-0.3-r0 do_image_ext3: Function failed: do_image_ext3 (log file is located at /opt/build/tmp/work/intel_corei7_64-wrs-linux/build-essential/0.3-r0/temp/log.do_image_ext3.43744)
ERROR: Logfile of failure stored in: /opt/build/tmp/work/intel_corei7_64-wrs-linux/build-essential/0.3-r0/temp/log.do_image_ext3.43744
ERROR: Task (/opt/layers/meta-overc/meta-build/recipes-core/images/build-essential_0.3.bb:do_image_ext3) failed with exit code '1'

Running with bitbake -v -v -v -D we get in the log file:
+ bbdebug 1 Executing 'dd if=/dev/zero of=/opt/build/tmp/work/intel_corei7_64-wrs-linux/build-essential/0.3-r0/deploy-build-essential-image-complete/build-essential-intel-corei7-64-20180220190510.rootfs.ext3 seek=484486 count=0 bs=1024'
+ USAGE='Usage: bbdebug [123] "message"'
+ '[' 3 -lt 2 ']'
+ DBGLVL=1
+ shift
++ echo 1

++ echo 1
++ tr -d t

++ tr -d t
+ NONDIGITS=1
+ '[' 1 ']'
+ bbfatal 'Usage: bbdebug [123] "message"'

The debug output tells us that the NONDIGITS check failed to remove
the digits using the tr expression.  Enclosing the expression in
quotes causes it to work properly.

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agompfr: merge bb/inc and fix license
Ross Burton [Tue, 20 Feb 2018 23:17:47 +0000 (23:17 +0000)]
mpfr: merge bb/inc and fix license

Merge the bb/inc as there's no reason to split them.

Remove redundant S assignment.

Fix the LICENSE assignment to LGPLv3+.  The source of mpfr is Lesser GPL v3 or
higher, the GPL is assigned to some test data that isn't shipped.

Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoxcb-proto: don't compile for Python 2
Ross Burton [Tue, 20 Feb 2018 22:43:53 +0000 (22:43 +0000)]
xcb-proto: don't compile for Python 2

Remove the do_install_append to create the Python 2 pyc files, as nothing in the
build is using Python 2 anymore (libxcb is the only user, and that uses Python
3).

Also use variables instead of a patch to control what Python binary and path the
modules are installed to.

Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibpthread-stubs: clean up recipe
Ross Burton [Tue, 20 Feb 2018 20:28:58 +0000 (20:28 +0000)]
libpthread-stubs: clean up recipe

No need to inherit pkgconfig as configure doesn't use it, remove commented
DEPENDS which clearly came from another recipe, stop deploying an empty PN.

Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibrsvg: provide RECIPE_NO_UPDATE_REASON
Maxin B. John [Tue, 20 Feb 2018 15:44:04 +0000 (17:44 +0200)]
librsvg: provide RECIPE_NO_UPDATE_REASON

Rustification of librsvg from version 2.41.0 onwards creates the need
for Rust compiler to build it. We need to push the upgrade of
librsvg until we have Rust support in oe-core.

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agogtk-icon-utils-native: upgrade to version 3.22.28
Maxin B. John [Tue, 20 Feb 2018 15:44:03 +0000 (17:44 +0200)]
gtk-icon-utils-native: upgrade to version 3.22.28

3.22.26 -> 3.22.28

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agogtk+3: upgrade to version 3.22.28
Maxin B. John [Tue, 20 Feb 2018 15:44:02 +0000 (17:44 +0200)]
gtk+3: upgrade to version 3.22.28

3.22.26 -> 3.22.28

Bug-fix release

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoglib-2.0: upgrade to version 2.54.3
Maxin B. John [Tue, 20 Feb 2018 15:44:01 +0000 (17:44 +0200)]
glib-2.0: upgrade to version 2.54.3

2.54.2 -> 2.54.3

Remove upstreamed patch:
        1. glib-mkenums-replace-and-warn-decoding.patch

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agomesa: Upgrade 17.3.3 -> 17.3.5
Otavio Salvador [Tue, 20 Feb 2018 13:28:07 +0000 (10:28 -0300)]
mesa: Upgrade 17.3.3 -> 17.3.5

The 17.3.5 release, published in February 19th, 2018 to fix a critical
regression from 17.3.4 release. They fix a number of issues since
17.3.3 release.

The release notes can be seen at:

  https://www.mesa3d.org/relnotes/17.3.4.html
  https://www.mesa3d.org/relnotes/17.3.5.html

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agomeson.bbclass: Add support for nativesdk
Peter Kjellerstedt [Tue, 20 Feb 2018 13:16:34 +0000 (14:16 +0100)]
meson.bbclass: Add support for nativesdk

We need to use the meson.cross file when building for nativesdk.
Additionally, we need to trick meson's sanity tests, just as it is
done for target builds.

Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agogst-validate: inherit upstream-version-is-even
Anuj Mittal [Tue, 20 Feb 2018 06:47:06 +0000 (14:47 +0800)]
gst-validate: inherit upstream-version-is-even

gst-validate follows GStreamer versioning. Inherit class to make
sure development releases aren't marked as latest.

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibtirpc: refresh patches
Ross Burton [Tue, 20 Feb 2018 00:46:52 +0000 (00:46 +0000)]
libtirpc: refresh patches

Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibtirpc: stop dropping in NIS headers
Ross Burton [Tue, 20 Feb 2018 00:39:57 +0000 (00:39 +0000)]
libtirpc: stop dropping in NIS headers

libtirpc prior to 1.0.2 assumed that the system provided nis.h but this isn't
always true.  Until now we've been using a tarball of the missing files from
Gentoo, but libtirpc 1.0.2 added a copy of nis.h to the sources so this isn't
required anymore.

Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agokernel.bbclass: explicitly depend on bison-native for deterministic builds
Denys Dmytriyenko [Mon, 19 Feb 2018 23:54:15 +0000 (18:54 -0500)]
kernel.bbclass: explicitly depend on bison-native for deterministic builds

Explicitly depend on bison-native for deterministic builds, as it is required
for the build:

|   HOSTCC  scripts/basic/fixdep
|   GEN     ./Makefile
|   HOSTCC  scripts/kconfig/conf.o
|   YACC    scripts/kconfig/zconf.tab.c
| /bin/sh: bison: command not found
| scripts/Makefile.lib:217: recipe for target 'scripts/kconfig/zconf.tab.c' failed

In most cases, this dependency comes indirectly via toolchain dependencies,
specifically binutils-cross, which pulls bison-native. Different setups,
such as with external toolchain, would expose this problem, since correct
dependency is not marked explicitly.

Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoicecc.bbclass: Remove icecream from uninative
Joshua Watt [Mon, 19 Feb 2018 22:30:22 +0000 (16:30 -0600)]
icecc.bbclass: Remove icecream from uninative

The icecream native tools should not be included in uninative tarballs
even though it is nativesdk

Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoicecc.bbclass: Add environment version
Joshua Watt [Mon, 19 Feb 2018 22:30:21 +0000 (16:30 -0600)]
icecc.bbclass: Add environment version

Adds a version to the environment which can be used to invalidate any
previous environments on the remote compile nodes

Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoicecc.bbclass: Fix combining with ccache
Joshua Watt [Mon, 19 Feb 2018 22:30:20 +0000 (16:30 -0600)]
icecc.bbclass: Fix combining with ccache

Fixes the case where ccache is enabled along with Icecream. In these
cases, there is the danger that Icecream will accidentally add the
ccache executable to the toolchain, which prevents it from working. In
particular, Fedora enables ccache by default via symbolic links in PATH.

Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agonativesdk-icecc-toolchain: Fix LIC_FILES_CHKSUM md5
Martin Jansa [Sat, 17 Feb 2018 00:18:24 +0000 (00:18 +0000)]
nativesdk-icecc-toolchain: Fix LIC_FILES_CHKSUM md5

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agomaintainers: update for change to icecc-create-env recipe name
Ross Burton [Thu, 22 Feb 2018 10:49:07 +0000 (10:49 +0000)]
maintainers: update for change to icecc-create-env recipe name

Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibunwind: Fix multilib header conflict - libunwind.h
Jagadeesh Krishnanjanappa [Fri, 16 Feb 2018 13:49:12 +0000 (19:19 +0530)]
libunwind: Fix multilib header conflict - libunwind.h

libunwind.h conflicts between 32-bit and 64-bit versions

This patch solves below error:
-- snip --
file /usr/include/libunwind.h conflicts between attempted installs of libunwind-dev-1.2-r0.core2_64 and lib32-libunwind-dev-1.2-r0.i586
-- snip --

Signed-off-by: Jagadeesh Krishnanjanappa <jkrishnanjanappa@mvista.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibsolv: upgrade to version 0.6.32
Maxin B. John [Mon, 19 Feb 2018 15:56:34 +0000 (17:56 +0200)]
libsolv: upgrade to version 0.6.32

0.6.31 -> 0.6.32

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agommc-utils: Upgrade to latest revision
Maxin B. John [Mon, 19 Feb 2018 15:56:33 +0000 (17:56 +0200)]
mmc-utils: Upgrade to latest revision

updates:

mmc-utils: manpage: fix arguments for TH macro
mmc-utils: remove unused #includes
mmc-utils: move offsetof from mmc.h to only user
mmc-utils: expand .gitignore
mmc-utils: make use of dependency information
mmc-utils: drop unused header
mmc-utils: drop macro CHECK

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agocve-check-tool: correctly exported web proxies
Konstantin Shemyak [Mon, 19 Feb 2018 14:35:54 +0000 (16:35 +0200)]
cve-check-tool: correctly exported web proxies

The binary 'cve-check-update' downloads the CVE database from the Internet.
If the system is behind a web proxy, the download fails, as proxy-related
variables are not exported.
In turn, 'cve-check-tool' does not connect to the network and correspondingly
does not need exported proxies.

Exported all proxy-related environment variables to 'cve-check-update' and
removed the unneeded export from 'cve-check-tool'.

Signed-off-by: Konstantin Shemyak <konstantin.shemyak@ge.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agopkgconf: upgrade to version 1.4.2
Maxin B. John [Mon, 19 Feb 2018 08:36:52 +0000 (10:36 +0200)]
pkgconf: upgrade to version 1.4.2

pkgconf improved compatibility with freedesktop.org pkg-config.
So, we don't need to unset PKG_CONFIG_SYSROOT_DIR in wrappers.

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agocurl: upgrade to 7.58.0
Oleksandr Kravchuk [Mon, 19 Feb 2018 01:32:09 +0000 (02:32 +0100)]
curl: upgrade to 7.58.0

Signed-off-by: Oleksandr Kravchuk <open.source@oleksandr-kravchuk.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibmpc: fix SRC_URI
Oleksandr Kravchuk [Mon, 19 Feb 2018 01:32:08 +0000 (02:32 +0100)]
libmpc: fix SRC_URI

Signed-off-by: Oleksandr Kravchuk <open.source@oleksandr-kravchuk.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibdrm: Upgrade 2.4.89 -> 2.4.90
Otavio Salvador [Sun, 18 Feb 2018 22:21:20 +0000 (19:21 -0300)]
libdrm: Upgrade 2.4.89 -> 2.4.90

The 2.4.90 was announced in Feb 17th 2018 and had following changes
are listed:

,----
| Andrey Grodzovsky (2):
|       amdgpu: Update deadlock test to not assert on ECANCELED
|       amdgpu: Fix segfault in deadlock test.
|
| Anuj Phogat (1):
|       intel: Add more Coffeelake PCI IDs
|
| Bas Nieuwenhuizen (1):
|       drm: Fix 32-bit drmSyncobjWait.
|
| Christian König (5):
|       amdgpu: fix 32bit VA manager max address
|       headers: sync up amdgpu_drm.h with drm-next
|       amdgpu: use the high VA range if possible v2
|       test/amdgpu: fix compiler warnings
|       amdgpu: fix high VA mask
|
| Christoph Haag (1):
|       meson: fix the install path of amdgpu.ids
|
| Chunming Zhou (5):
|       fix return value for syncobj wait
|       amdgpu: fix inefficient vamgr algorithm
|       amdgpu: clean up non list code path for vamgr
|       tests/amdgpu: add bo eviction test
|       amdgpu: clean up non list code path for
|       vamgr v2
|
| Dylan Baker (7):
|       Add meson build system
|       autotools: Include meson.build files in tarball
|       README: Add note about meson
|       meson: set proper pkg-config version for
|       libdrm_freedreno
|       meson: set the minimum version correctly
|       meson: fix libdrm_nouveau pkgconfig include directories
|       meson: include headers in root directory in ext_libdrm
|
| Emil Velikov (1):
|       tests/amdgpu: add missing config.h include
|
| Eric Engestrom (25):
|       remove unnecessary double-semicolon
|       tests/amdgpu: add parentheses to make operation priority explicit
|       tests/amdgpu: drop unused variables
|       tests/util: fix signed/unsigned comparisons
|       tests/util: drop unused parameters
|       tests/etnaviv: drop unused `return 0`
|       meson: add missing HAVE_RADEON
|       configure: remove unused HAVE_CUNIT define
|       configure: remove unused HAVE_INSTALL_TESTS define
|       meson,configure: remove unused HAVE_OMAP define
|       meson,configure: remove unused HAVE_TEGRA define
|       meson,configure: remove unused HAVE_FREEDRENO define
|       meson,configure: remove unused HAVE_ETNAVIV define
|       meson,configure: always define HAVE_{INTEL,VMWGFX,NOUVEAU,EXYNOS,VC4,RADEON}
|       always define HAVE_FREEDRENO_KGSL
|       always define HAVE_CAIRO
|       always define HAVE_VALGRIND
|       meson: sort HAVE_* defines
|       xf86atomic: fix -Wundef warning
|       meson: cleanup whitespace
|       meson,configure: add warning when using undefined preprocessor tokens
|       xf86drmHash: remove always-false #if guards
|       configure: always define HAVE_LIBDRM_ATOMIC_PRIMITIVES and HAVE_LIB_ATOMIC_OPS
|       exynos/tests: use #ifdef for never-defined token
|       meson,configure: turn undefined preprocessor tokens warnings into errors
|
| Hawking Zhang (3):
|       tests/amdgpu: execute write linear on all the available rings
|       tests/amdgpu: execute const fill on all the available rings
|       tests/amdgpu: execute copy linear on all the available rings
|
| Marek Olšák (2):
|       amdgpu: add amdgpu_query_sw_info for querying high bits of 32-bit address space
|       configure.ac: bump version to 2.4.90
|
| Michel Dänzer (7):
|       amdgpu: Don't print error message if parse_one_line returned -EAGAIN
|       amdgpu: Don't dereference device_handle after
|       amdgpu_device_deinitialize
|       amdgpu: Symlink .editorconfig to tests/amdgpu
|       amdgpu: Disable deadlock test suite by default for SI ASICs
|       amdgpu: Disable VM test suite by default for SI ASICs
|       Revert "amdgpu: clean up non list code path for vamgr"
|       amdgpu: Add amdgpu_query_sw_info to amdgpu-symbol-check
|
| Rob Clark (1):
|       freedreno: clamp priority based on # of rings
|
| Robert Foss (5):
|       android: Move gralloc handle struct to libdrm
|       android: Add version variable to gralloc_handle_t
|       android: Mark gralloc_handle_t magic variable as const
|       android: Remove member name from gralloc_handle_t
|       android: Change gralloc_handle_t members to be fixed width
|
| Seung-Woo Kim (2):
|       amdgpu: fix not to add amdgpu.ids when building without amdgpu
|       modetest: Fix to check return value of asprintf()
`----

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agomtd-utils: Bump revision to 1bfee866
Otavio Salvador [Sun, 18 Feb 2018 22:21:19 +0000 (19:21 -0300)]
mtd-utils: Bump revision to 1bfee866

This bumps the version to 2.0.1 release plus some more changes.

The following patches were merged upstream:

 - 0001-Fix-build-with-musl.patch
 - 010-fix-rpmatch.patch
 - fix-armv7-neon-alignment.patch

The `mtd-utils-fix-corrupt-cleanmarker-with-flash_erase--j-command.patch`
patch has been removed as the issue seems to have been addressed
in the Linux MTD subsystem. The discussion can be seen at:

 http://lists.infradead.org/pipermail/linux-mtd/2011-August/037320.html

Also a research to see if any other distribution had a similar patch
included and they don't. So there is no clear need to keep diverging
from upstream.

See the logs below:

,----[ Changes from 2.0.0 -> 2.0.1 ]
81049e5 Release mtd-utils-2.0.1
4458ad6 mtd-utils: tests: Avoid using less than two blocks in nandpagetest
76a55b1 mtd-utils: ubinfo: add parameter check
454a3d0 mkfs-ubifs: fix inclusion of uuid.h
df913e4 mtd-utils: tests: Add Erased Pages Bit Flip Test
f8bc7d9 Add const modifier to read only strings and string constants
a7c8cb3 Silence warnings about unused arguments
e511691 Remove unused variables and functions
ca517e8 Eliminate warnings about missing prototypes
3029b65 Enable compiler warnings
9eae306 Move libfec declarations to public header in global include directory
022ce62 Use autoconf header detection correctly for libmissing
1a3e3af Restructure autoconf configure.ac
e772dc8 Replace defunct ubifs_assert
b0b1bc5 mtd-utils: tests: Fix nandbiterrs Failure Check
76e27a7 ubi-utils: Return error code if command line option is unknown
b8e785a nor-utils: Return error code if command line option is unknown
12e27a1 jffsX-utils: Return error code if command line option is unknown
9a357ee ftl_format: Use return directly to leave main function
6890608 flashcp: Drop exit code defines
ccffc10 Use defines for exit code values
871c2f1 mkfs.ubifs: Add support for symlinks in device table
c0972a5 libiniparser: remove unused function needing float
5c57076 libubi.c: add klibc specific fixes for ioctl
dede98f Return correct error number in ubi_get_vol_info1
a2eeedb Fix libmtd behaviour if MTD is not present on the system
4dab9be Fix build with musl
82839c3 Replace rpmatch() usage with checking first character of line
7d026a8 Fix alignment trap triggered by NEON instructions
d7e8612 mtd-utils: Support jffs2 flash-erase for large OOB (>32b)
fc7aa01 ubi: tests: Speedup io_paral by using rand_r()
fdec8a4 ubi: tests: Support up to 65k NAND page size
d600419 ubi: tests: Replace variable-length array with malloc()
1710e0c Remove README.udev from ubi-tests extra dist
e3e010c ubirename: trivial fixes to the help text
1fed9b4 Remove UDEV_SETTLE_HACK
22089b0 mtd-utils: serve_image: Use PRIdoff_t as format specifier.
5a80691 mtd-utils: Fix format specifier definitions for off_t and loff_t.
44558d1 nanddump: Add --skip-bad-blocks-to-start option
2fc8058 nandwrite: Add --skip-bad-blocks-to-start option
6df83fd nandwrite: fix/cleanup bad block skipping
7b8981a nandwrite: replace erase loop with mtd_erase_multi
dffaa1d nandwrite: add stricter sanity checking for blockalign
`----

,----[ Changes from 2.0.1 -> 9c61735 ]
9c61735 mtd: unittests: Stop testing stat() calls
daef4f1 mtd: unittests: Decode arg size from ioctl request
fd0b70b mtd: unittests: Use proper unsigned long type for ioctl requests
c75d104 ubi-utils: ubiformat.c: convert to integer arithmetic
61afcc6 mtd-utils: common.c: convert to integer arithmetic
6766178 Run unit test programs through "make check"
098e91a mtd: tests: Fix check on ebcnt in nandpagetest
9cfb02c libscan: fix a comment typo in libscan.h
7086393 libmtd: fix a comment typo in dev_node2num
3c54e4e ubi-utils: ubicrc32: process command line arguments first
ce50e6d nandbiterrs: Fix copy & paste fail
03dafe3 Enable further warning flags, address new warnings
0555cb2 jittertest: Use the appropriate versions of abs()
475bf70 Mark or fix switch cases that fall through
cb2fcfc Add no-return attribute to usage() style functions
81fd176 Remove self-assignments of unused paramters
3b421ab mkfs.ubifs: ignore EOPNOTSUPP when listing extended attributes
030a63b tests: checkfs: Remove unused source file from makefiles
40dd609 misc-utils: initialize "ip" in docfdisk to NULL
94f6573 ubi-tests: io_update: fix missleading indentation
afba4fc mkfs.ubifs: Apply squash-uids to the root node
3a34784 Add ctags files to .gitignore
`----

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agositeinfo: add aarch64_illp32 decode
Armin Kuster [Sat, 17 Feb 2018 15:44:20 +0000 (07:44 -0800)]
siteinfo: add aarch64_illp32 decode

Signed-off-by: Armin Kuster <akuster808@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoupdate-rc.d: QA regression.
Armin Kuster [Sat, 17 Feb 2018 15:44:19 +0000 (07:44 -0800)]
update-rc.d: QA regression.

I noticed many new QA warning with arm64 mulitlib
suspicious values 'initd-functions-dev' in RRECOMMENDS [multilib]

I believe this is a regression via commit
http://cgit.openembedded.org/openembedded-core/commit/meta/classes/update-rc.d.bbclass?id=cdcebd81c872cb7386c658998e27cf24e1d0447c

Signed-off-by: Armin Kuster <akuster808@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agogo-dep: Upgrade 0.3.2 -> 0.4.1
Otavio Salvador [Sat, 17 Feb 2018 01:36:23 +0000 (23:36 -0200)]
go-dep: Upgrade 0.3.2 -> 0.4.1

The release was announced in Jan 24th 2018 and provides a number of
features and bug fixes.

The 0.4.1 restores the `dep prune` command support, which was dropped
on 0.4.0, so existing scripts will continue to work. That said, it is
advisable to change the scripts to use the new features of `dep
ensure` as `dep prune` will be removed in future.

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoutils.py: Fix parallel_make limit
Joshua Watt [Fri, 16 Feb 2018 20:30:00 +0000 (14:30 -0600)]
utils.py: Fix parallel_make limit

parallel_make_argument() was incorrectly taking the maximum of the limit
and the calculated value instead of the minimum.

Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agorecipes-devtools: Bump QEMU to 2.11.1
Alistair Francis [Fri, 16 Feb 2018 19:48:41 +0000 (11:48 -0800)]
recipes-devtools: Bump QEMU to 2.11.1

Bump the QEMU version to the bug fix release of 2.11.1 and remove the
patches that are no longer required.

Signed-off-by: Alistair Francis <alistair.francis@xilinx.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agowebkitgtk_2.18.6.bb: Fix configure failure for aarch64 build
Manjukumar Matha [Fri, 16 Feb 2018 03:26:30 +0000 (19:26 -0800)]
webkitgtk_2.18.6.bb: Fix configure failure for aarch64 build

Fix configure failure for aarch64

| -- Performing Test CXX_COMPILER_SUPPORTS_-fcolor-diagnostics - Failed
| -- Performing Test CXX_COMPILER_SUPPORTS_-fdiagnostics-color=always
| -- Performing Test CXX_COMPILER_SUPPORTS_-fdiagnostics-color=always -Success
| CMake Error at Source/cmake/OptionsCommon.cmake:42 (WEBKIT_PREPEND_GLOBAL_COMPILER_FLAG):
|   Unknown CMake command "WEBKIT_PREPEND_GLOBAL_COMPILER_FLAG".
| Call Stack (most recent call first):
|   Source/cmake/WebKitCommon.cmake:58 (include)
|   CMakeLists.txt:166 (include)

Signed-off-by: Manjukumar Matha <manjukumar.harthikote-matha@xilinx.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agocmake: Upgrade 3.10.1 -> 3.10.2
Otavio Salvador [Thu, 15 Feb 2018 17:11:16 +0000 (15:11 -0200)]
cmake: Upgrade 3.10.1 -> 3.10.2

This is a minor stable update to CMake. This includes following
changes:

,----
c1e087a9d CMake 3.10.2
d592bfc9f Autogen: Ignore not existing source files in cmMakefile
d49986b3b KWIML: Backport RISC-V update to CMake 3.10
9417a6d32 FindPostgreSQL: Add support for PG10
683e9023a Autogen: Fix for problematic nested list separator
9b4d31dc7 cmGraphVizWriter: Updated to create and follow dependers for
|           interface targets
`----

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agogo: Upgrade 1.9 to 1.9.4 stable release
Otavio Salvador [Thu, 15 Feb 2018 17:11:14 +0000 (15:11 -0200)]
go: Upgrade 1.9 to 1.9.4 stable release

The 1.9.4 fixes a number of issues in the Go compiler and is important
to get in before we start working on 1.10 inclusion.

 - go1.9.1 (released 2017/10/04) includes two security fixes.

 - go1.9.2 (released 2017/10/25) includes fixes to the compiler,
   linker, runtime, documentation, go command, and the crypto/x509,
   database/sql, log, and net/smtp packages. It includes a fix to a
   bug introduced in Go 1.9.1 that broke go get of non-Git
   repositories under certain conditions.

 - go1.9.3 (released 2018/01/22) includes fixes to the compiler,
   runtime, and the database/sql, math/big, net/http, and net/url
   packages.

 - go1.9.4 (released 2018/02/07) includes a security fix to “go get”.

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoeglinfo-fb: Pass -DMESA_EGL_NO_X11_HEADERS to cxxflags
Khem Raj [Thu, 15 Feb 2018 07:49:25 +0000 (23:49 -0800)]
eglinfo-fb: Pass -DMESA_EGL_NO_X11_HEADERS to cxxflags

This helps in compiling on graphics drivers which use
this define to differentiate between eglfs and x11 headers
e.g. mali or mesa

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoopenssl: remove patch from 1.0.2m left behind after update to 1.0.2n
Denys Dmytriyenko [Wed, 14 Feb 2018 21:30:11 +0000 (16:30 -0500)]
openssl: remove patch from 1.0.2m left behind after update to 1.0.2n

Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agosudo: 1.8.21P1 -> 1.8.22
Huang Qiyu [Tue, 13 Feb 2018 05:48:41 +0000 (13:48 +0800)]
sudo: 1.8.21P1 -> 1.8.22

1.Upgrade sudo from 1.8.21P1 to 1.8.22.
2.Update the checksum of LIC_FILES_CHKSUM.
  The following content is appended to doc/LICENSE, plugins/sudoers/redblack.c.

    Todd C. Miller <Todd.Miller@courtesan.com> -> Todd C. Miller <Todd.Miller@sudo.ws>

Signed-off-by: Huang Qiyu <huangqy.fnst@cn.fujitsu.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agooe-selftest: add a test for recipes without maintainers
Alexander Kanavin [Mon, 12 Feb 2018 11:05:21 +0000 (13:05 +0200)]
oe-selftest: add a test for recipes without maintainers

'bitbake -c checkpkg world' is moved to class initializer to avoid
it being run twice in a row.

The no-maintainers test checks only oe-core recipes, as other layers
may be be configured, and assigning maintainership to specific people via
maintainers.inc is known to be used only in oe-core.

Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agooe-selftest: fix tests to reflect man->man-db transition
Alexander Kanavin [Mon, 12 Feb 2018 11:05:19 +0000 (13:05 +0200)]
oe-selftest: fix tests to reflect man->man-db transition

Rather than apply a patch twice, an incorrect patch is applied
(as the new man-db recipe does not have any patches yet).

Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agoman-db: replace the outdated man implementation
Alexander Kanavin [Mon, 12 Feb 2018 11:05:18 +0000 (13:05 +0200)]
man-db: replace the outdated man implementation

Until now oe-core has been using a well obsolete implementation of man.

Man-db on the other hand is used by all modern Linux distros, is actively
maintained, has a standard build system, and does not require 30 patches
to build.

Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
7 years agolibpipeline: add a recipe
Alexander Kanavin [Mon, 12 Feb 2018 11:05:17 +0000 (13:05 +0200)]
libpipeline: add a recipe

This is a hard dependency of man-db, which will be added in the following commit.

Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>