sanity.bbclass: quote path passed to stat in get_filesystem_id()

Although get_filesystem_id() is a private API and never gets passed
a path containing spaces or other special characters, etc, quote the
path anyway for consistency.

Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

glibc: Replace strncpy with memccpy to fix -Wstringop-truncation.

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

glibc-2.27: Update to bring in 2.27 bug fixes since release

Here is full list of fixes

https://sourceware.org/git/?p=glibc.git;a=shortlog;h=df3ff4e49d4ee3cbbdaeb0b1cb5dc2344c08be98;hp=23158b08a0908f381459f273a984c6fd328363cb

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

tcp-wrappers: Fix build with clang

Fix non-void function 'fix_options' should return a value.
Add function prototype to tcpd.c and miscd.c.

Signed-off-by: Oleksiy Obitotskyy <oobitots@cisco.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

systemtap: bring in PR22551 fix that addresses build against 4.15 kernel issue

With 4.15 kernel systemtap needs update to address systemtap module
compilation issues. It is fixed in later version of systemtap by
PR22551.

Upstream-Status: Backport
Signed-off-by: Victor Kamensky <kamensky@cisco.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

systemtap: supports mips starting with version 3.1

Signed-off-by: Victor Kamensky <kamensky@cisco.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

elfutils: Fix gcc compile time alignment errors

Allow devtool to organize the SRC_URI

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

initrdscripts: format rootfs partition as ext4

Use ext4 filesystem instead of ext3 when using the live image to install
on target. wic defaults to ext4 as well.

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

libxml-namespacesupport: use stable v1.12 release; inherit ptest-perl

* Fix RDEPENDS
* Upstream v1.12_9 is a development version, not a stable release
* Add UPSTREAM_CHECK_REGEX skip development releases
* Drop anonymous python function to "fix" version, which breaks
  auto-upgrade-helper (AUH)
* Use LICENSE file for checksum rather than ephemeral META.yml
* License remains the same

Fixes: [YOCTO #12581]
License-Update: use LICENSE file for checksum

Signed-off-by: Tim Orling <timothy.t.orling@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

pseudo: explicitly enable xattr support

Pseudo is using a custom configure script that detects if it shall build with
extended file attribute support or not. The check is done by simply calling
'getfattr' provided by attr-native which is not part of the dependency list.
Due to the recent changes (recipe specific sysroot & cleanup of $PATH) this
call fails now when the recipe is being build for the first time (at least
when being build for nativesdk case). Explicitly setting up a dependency to
attr-native just to satisfy configure would be wrong also since the real
dependency is to attr/nativesdk-attr which are already part of the dependency
list (see DEPENDS). Therefore bypass the test in the configure by explicitly
enabling xattr using a configure option available in any case.

Signed-off-by: Andreas Kaufmann <andreas.kaufmann.79@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

virtual/libgbm: create

The 'glamor' PACKAGECONFIG in xserver-xorg creates a dependency on libgbm
which can be satisfied in some cases by mesa, in others by blobs such as mali.

Signed-off-by: Trevor Woerner <twoerner@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

mesa.inc: make PROVIDES conditional on PACKAGECONFIG

Mesa only PROVIDES these features if they are enabled via PACKAGECONFIG.
Therefore make the PROVIDES conditional depending on whether or not these
features have been enabled.

Signed-off-by: Trevor Woerner <twoerner@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

mesa: update Upstream-Status of a patch

replace_glibc_check_with_linux.patch was accepted upstream with
modifications.

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

mesa: Upgrade 17.3.6 -> 17.3.7

This version has been published at March 21, 2018, and it is a bugfix
only release. It includes several important fixes that were made as
part of 18.0.0 development cycle.

Full list of bug fixes can be see online at:

 https://www.mesa3d.org/relnotes/17.3.7.html

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>

distcc: Change SRC_URI

ERROR: distcc-3.2-r0 do_fetch: Fetcher failure: Unable to find revision d8b18df3e9dcbe4f092bed565835d3975e99432c in branch 3.2 even from upstream
ERROR: distcc-3.2-r0 do_fetch: Fetcher failure for URL: 'git://github.com/distcc/distcc.git;branch=3.2'. Unable to fetch URL from any source.
ERROR: distcc-3.2-r0 do_fetch: Function failed: base_do_fetch

[v2]
upstream deleted the branch and the hash no longer exists.

Took the git snapshot from yocto and created a copy on my github.
There was no offical 3.2 release, only rc versions.

Signed-off-by: Armin Kuster <akuster@mvista.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

waf.bbclass: Throw error if waf doesn't exist

Before, waf.bbclass would fail to catch FileNotFoundError. Now, it will
catch this error and say that waf doesn't exist.

Fixes [YOCTO 12553]

Signed-off-by: Amanda Brindle <amanda.r.brindle@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

reproducible_build.bbclass: support for binary reproducibility

Setup environment for builds requiring binary reproducibility.
Determine and export SOURCE_DATE_EPOCH per each recipe.
This is a crucial step to achieve binary reproducibility.
The value for this variable (timestamp) is obtained after source code for
a recipe has been unpacked, but before it is patched. If the code sources
come from a GIT repo, we get the timestamp from the top commit. (GIT repo
does not preserve file mktime timestamps). Otherwise, if GIT repo is not
present, we try to get mtime from known files such as NEWS, ChangeLog, etc.
If this also fails, we go through all files and get the timestamp from the
youngest one. We create an individual timestamp for each recipe.
The timestamp is stored in the file '__source_date_epoch.txt' (in the folder
source-date-epoch_). Later on, each task reads this file and sets
the exported value of SOURCE_DATE_EPOCH to the value found in the file.

Uasge:

INHERIT += "reproducible_build"

[YOCTO#11178]
[YOCTO#11179]

Signed-off-by: Juro Bystricky <juro.bystricky@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

systemd: link udev statically with systemd internal libraries

This was the default behavior with autotools, but is not with meson.
Otherwise, udev package will pull in the rest of systemd even
that is not desired.

[YOCTO #12618]

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

systemd: change PACKAGECONFIG 'resolve' back to 'resolved'

When systemd was upgraded from 234 to 237, the PACKAGECONFIG item
'resolved' is changed to 'resolve', this is because meson_options.txt
uses the word 'resolve' instead of 'resolved'.

However, this causes trouble for users. Backward compatibility is obviously
more important, because we might have bbappend files in other layers
using this PACKAGECONFIG item.

So change the name back to 'resolved'.

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

externalsrc: do not call make clean for recipes with CLEANBROKEN = "1" set

ERROR: distrho-ports-0.0.0+git999-r0 do_buildclean: oe_runmake failed
ERROR: distrho-ports-0.0.0+git999-r0 do_buildclean: Function failed: do_buildclean
ERROR: Logfile of failure stored in: <...>/temp/log.do_buildclean.17285
Log data follows:
| DEBUG: Executing shell function do_buildclean
| NOTE: make clean
| make clean -C libs/drowaudio
| make[1]: Entering directory '/home/a.mueller/data/oe-core/workspace/sources/distrho-ports/libs/drowaudio'
| make clean -C build-drowaudio
| make[2]: Entering directory '/home/a.mueller/data/oe-core/workspace/sources/distrho-ports/libs/drowaudio/build-drowaudio'
| make[2]: *** No rule to make target 'clean'.  Stop.
| make[2]: Leaving directory '/home/a.mueller/data/oe-core/workspace/sources/distrho-ports/libs/drowaudio/build-drowaudio'
| make[1]: *** [Makefile:7: clean] Error 2
| make[1]: Leaving directory '/home/a.mueller/data/oe-core/workspace/sources/distrho-ports/libs/drowaudio'
| make: *** [Makefile:73: clean] Error 2
| ERROR: oe_runmake failed

Signed-off-by: Andreas Müller <schnitzeltony@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

gtk-doc: don't regenerate gtk-doc in do_install

In out-of-tree builds gtk-doc's setup-build target copies all the content from
$srcdir to $builddir. However, if some of this content is regenerated at
configure time this can happen:

1) configure writes new build/version.xml
2) make compile copies content, including the tarball's src/version.xml
   to build/version.xml, and generates gtk-doc.
3) make install notices build/version.xml is older than configure.status,
   so regenerates gtk-doc.

gtk-doc generation is a slow process at the best of times, so doing it twice
isn't good.

Signed-off-by: Ross Burton <ross.burton@intel.com>

librepo: disable building of tests and docs

Also remove libcheck dependency which was required only for tests.

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

ncurses: Abstract out termlib

termlib needs to be disabled on some targets e.g. mingw
this change paves the way for doing that. Functionally
it does not change anything for other platforms

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

systemd: Fix build failures with glibc 2.27 + kernels without memfd

Backport a fix that is needed for systemd to build with latest glibc
and kernel being old.

see
https://github.com/systemd/systemd/issues/8099

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

site/risc-v: Cache common variables to build libIDL

These variables force runtime tests during configure
they are already cached for other architectures

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

glide.bbclass: Add class to easy Glide use

To use 'glide' this class does the integration and reduces code
duplication.

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>

shadow.inc: run postinst only for target

* fails for nativesdk-shadow with:
  pwconv: /etc/passwd.29063: No such file or directory
  pwconv: cannot lock /etc/passwd; try again later.

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

iputils: change default PACKAGECONFIG to disable libidn

* wrong revision of this patch, where the commit message didn't match
  with the default PACKAGECONFIG value, was merged to master, update
  it to avoid confusion

* it got enabled by default, but without the dependency on libidn in:
  commit 5997981fa2c22609a88b8cbb595dbf7758b2f7c2
  Author: Alexander Kanavin <alexander.kanavin@linux.intel.com>
  AuthorDate: Thu Feb 1 20:02:08 2018 +0200
  Subject: iputils: update to 20161105

* https://github.com/iputils/iputils/blob/master/RELNOTES.old
  mentiones that IDN was enabled by default in:
  [s20160308] and surprisingly the same in [s20150815]
  but there are no release notes for s20151218 version we were using until
  now, don't know how it really relates to [s20150815].

* but there are some issues with libidn as described in:
  https://github.com/iputils/iputils/commit/f3a461603ef4fb7512ade3bdb73fe1824e294547
  so disable it by default.

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

maintainers.inc: drop lsb4 perl modules no longer in core

The following perl module packages were removed from oe-core
with commit: 30fb4c8f329fe3aa3c528ffeba60ee7d702e873e
  - libclass-isa-perl
  - libenv-perl
  - libdumpvalue-perl
  - libfile-checktree-perl
  - libi18n-collate-perl
  - libpod-plainer-perl

Remove these from the maintainers list

Fixes: [Yocto #12582]
Signed-off-by: Tim Orling <timothy.t.orling@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

maintainers: remove obsolete entries

Delete entries of removed packages.

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

libvorbis: CVE-2018-5146

Prevent out-of-bounds write in codebook decoding. The bug could allow
code execution from a specially crafted Ogg Vorbis file.

References:
https://www.debian.org/security/2018/dsa-4140
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2018-5146

Signed-off-by: Tanu Kaskinen <tanuk@iki.fi>
Signed-off-by: Ross Burton <ross.burton@intel.com>

libvorbis: CVE-2017-14632

Xiph.Org libvorbis 1.3.5 allows Remote Code Execution upon freeing
uninitialized memory in the function vorbis_analysis_headerout() in
info.c when vi->channels<=0, a similar issue to Mozilla bug 550184.

References:
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2017-14632

Signed-off-by: Tanu Kaskinen <tanuk@iki.fi>
Signed-off-by: Ross Burton <ross.burton@intel.com>

libvorbis: CVE-2017-14633

In Xiph.Org libvorbis 1.3.5, an out-of-bounds array read vulnerability
exists in the function mapping0_forward() in mapping0.c, which may lead
to DoS when operating on a crafted audio file with vorbis_analysis().

References:
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2017-14633

Signed-off-by: Tanu Kaskinen <tanuk@iki.fi>
Signed-off-by: Ross Burton <ross.burton@intel.com>

xserver-xorg: import distro patch to use modesetting driver on newer hardware

Debian and Fedora both carry this patch, and the xf86-video-modesetting
driver seems better on recent hardware.

As an example, on a NUC6CAYS, the x11perf -aa10text and -rgb10text tests
see around a 20x increase.

[YOCTO #12019]
[YOCTO #12390]

Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

tiff: Security fixes

Fix CVE-2017-99935, CVE-2017-18013, CVE-2018-5784

References:
https://nvd.nist.gov/vuln/detail/CVE-2017-9935
https://nvd.nist.gov/vuln/detail/CVE-2017-18013
https://nvd.nist.gov/vuln/detail/CVE-2018-5784

Patches from:
CVE-2017-9935:
https://gitlab.com/libtiff/libtiff/commit/3dd8f6a357981a4090f126ab9025056c938b6940
CVE-2017-18013:
https://gitlab.com/libtiff/libtiff/commit/c6f41df7b581402dfba3c19a1e3df4454c551a01
CVE-2018-5784:
https://gitlab.com/libtiff/libtiff/commit/473851d211cf8805a161820337ca74cc9615d6ef

Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

systemd-boot: upgrade to 237

Upgrade systemd-boot to 237.

As systemd has dropped autotools support, fix configure and compile
failures related to meson.

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>

systemd: fix build failure for qemux86 and qemuppc with musl

Remove the 'fstack-protector' and 'fstack-protector-strong' flags
as a workaround to fix the following error when building for qemux86
and qemuppc with musl.

  undefined reference to `__stack_chk_fail_local'

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>

systemd: upgrade to 237

Upgrade systemd to 237.

Note that this version has dropped autotools support.

The following patches are rebased:
0004-Use-getenv-when-secure-versions-are-not-available.patch
0005-binfmt-Don-t-install-dependency-links-at-install-tim.patch
0007-use-lnr-wrapper-instead-of-looking-for-relative-opti.patch
0015-Revert-udev-remove-userspace-firmware-loading-suppor.patch
0018-check-for-uchar.h-in-configure.patch
0019-socket-util-don-t-fail-if-libc-doesn-t-support-IDN.patch
0001-add-fallback-parse_printf_format-implementation.patch
0002-src-basic-missing.h-check-for-missing-strndupa.patch
0007-check-for-missing-canonicalize_file_name.patch
0008-Do-not-enable-nss-tests.patch
0010-test-sizeof.c-Disable-tests-for-missing-typedefs-in-.patch
0011-nss-mymachines-Build-conditionally-when-HAVE_MYHOSTN.patch

The following backported patches are dropped:
0001-core-evaluate-presets-after-generators-have-run-6526.patch
0001-main-skip-many-initialization-steps-when-running-in-.patch
0001-meson-update-header-file-to-detect-memfd_create.patch
0003-fileio-include-sys-mman.h.patch

The following patch is dropped as autotools support is dropped:
0002-configure.ac-Check-if-memfd_create-is-already-define.patch

The following patches are newly added to fix problems:
0027-remove-nobody-user-group-checking.patch
0028-add-missing-FTW_-macros-for-musl.patch
0030-fix-missing-of-__register_atfork-for-non-glibc-build.patch
0031-fix-missing-ULONG_LONG_MAX-definition-in-case-of-mus.patch

Other changes are mostly autotools/meson related.

This new version has dropped ptest support, as there's no easy
way to do this in the framework of meson.

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>

libsolv: refresh the patches

fixes:

WARNING: libsolv-0.6.33-r0 do_patch:
Some of the context lines in patches were ignored. This can lead to
incorrectly applied patches.
The context lines in the patches can be updated with devtool:

devtool modify <recipe>
devtool finish --force-patch-refresh <recipe> <layer_path>

Then the updated patches and the source tree (in devtool's workspace)
should be reviewed to make sure the patches apply in the correct place
and don't introduce duplicate lines (which can, and does happen
when some of the context is ignored). Further information:
http://lists.openembedded.org/pipermail/openembedded-core/2018-March/148675.html
https://bugzilla.yoctoproject.org/show_bug.cgi?id=10450
Details:
Applying patch
0001-Add-fallback-fopencookie-implementation.patch
patching file ext/CMakeLists.txt
patching file ext/solv_xfopen.c
Hunk #1 succeeded at 12 with fuzz 1 (offset -1 lines).
Hunk #2 succeeded at 25 (offset -18 lines).
Hunk #3 succeeded at 34 (offset -18 lines).
Hunk #4 succeeded at 46 (offset -18 lines).
patching file ext/solv_xfopen_fallback_fopencookie.c
patching file ext/solv_xfopen_fallback_fopencookie.h

Now at patch 0001-Add-fallback-fopencookie-implementation.patch

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

iputils: add PACKAGECONFIG for libidn and disable it by default

* it got enabled by default, but without the dependency on libidn in:
  commit 5997981fa2c22609a88b8cbb595dbf7758b2f7c2
  Author: Alexander Kanavin <alexander.kanavin@linux.intel.com>
  AuthorDate: Thu Feb 1 20:02:08 2018 +0200
  Subject: iputils: update to 20161105

* https://github.com/iputils/iputils/blob/master/RELNOTES.old
  mentiones that IDN was enabled by default in:
  [s20160308] and surprisingly the same in [s20150815]
  but there are no release notes for s20151218 version we were using until
  now, don't know how it really relates to [s20150815].

* but there are some issues with libidn as described in:
  https://github.com/iputils/iputils/commit/f3a461603ef4fb7512ade3bdb73fe1824e294547
  so disable it by default.

* fails with:
  | In file included from ping_common.c:1:0:
  | ping.h:39:10: fatal error: idna.h: No such file or directory
  |  #include <idna.h>
  |           ^~~~~~~~

* Easiest way to reproduce this failure is to remove libidn from gnutls
  PACKAGECONFIG or to use gnutls which doesn't have libidn PACKAGECONFIG
  at all (like the one in meta-gplv2).

* First it leads to following QA issue:
  http://errors.yoctoproject.org/Errors/Build/53212/
  ERROR: iputils-s20161105-r0 do_package_qa: QA Issue: iputils-ping rdepends on libidn, but it isn't a build dependency, missing libidn in DEPENDS or PACKAGECONFIG? [build-deps]
  ERROR: iputils-s20161105-r0 do_package_qa: QA Issue: iputils-traceroute6 rdepends on libidn, but it isn't a build dependency, missing libidn in DEPENDS or PACKAGECONFIG? [build-deps]
  ERROR: iputils-s20161105-r0 do_package_qa: QA run found fatal errors.
  Please consider fixing them.
  ERROR: iputils-s20161105-r0 do_package_qa: Function failed:
  do_package_qa
  ERROR: Logfile of failure stored in: /OE/build/oe-core/tmp-glibc/work/core2-64-oe-linux/iputils/s20161105-r0/temp/log.do_package_qa.7627
  ERROR: Task (/OE/build/oe-core/openembedded-core/meta/recipes-extended/iputils/iputils_s20161105.bb:do_package_qa) failed with exit code '1'

* But if you cleansstate iputils as well (after removing libidn from
  gnutls PACKAGECONFIG) to empty iputils RSS, then you get the error about
  missing idna.h:
  http://errors.yoctoproject.org/Errors/Build/53213/

* Adding the libidn dependency explicitly in iputils recipe fixes the
  issue.

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

package-index: index also subdirectories when using rpm

Previously only the top-level index was created, which did not
work if PACKAGE_FEED_ARCHS whitelisting (or explicitly listing
architectures in dnf repo files by hand) was in use:
https://lists.yoctoproject.org/pipermail/yocto/2018-March/040327.html
https://bugzilla.yoctoproject.org/show_bug.cgi?id=12419

[YOCTO #12419]

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

glibc: fix upstream version check

Exclude x.y.90* pre-release versions in particular.

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

dpkg: Backport riscv support

Refresh patches with devtool

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

mtd-utils: Explicitly add pthread options to cflags

Some architectures e.g. riscv gcc does not add -D_REENTRANT
when enabling pthreads. Help it here by adding these options
while gcc gets fixed

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

linux-firmware: upgrade to 4c0bf11 revision

License-Update: new releases and copyright years updated.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>

watchdog: remove interdependencies of watchdog and wd_keepalive

Since watchdog and watchdog-keepalive packages can't be installed
together, move wd_keepalive.service to watchdog-keepalive package.

Remove the inter-dependencies of watchdog and wd_keepalive
services as well.

[YOCTO #12565]

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

opkg-utils: use multithreaded xz when building packages

Signed-off-by: Ross Burton <ross.burton@intel.com>

perf: enable support for libaudit

perf needs audit-python to be able to show syscall names and for
'perf trace' to work.

Enable dependency on audit-python if present in PACKAGECONFIG. It's
disabled by default since audit as of now is in meta-selinux.

Fixes [YOCTO #3343]
Fixes [YOCTO #3358]

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

image_types.bbclass: Rename lz4_legacy to lz4

LZ4 format currently used by the Linux kernel is the 'legacy' format.

In order to avoid creating an image that can't be used as a compressed
initial ramdisk with Linux kernel, rename lz4_legacy to lz4.

[YOCTO #12461]
[YOCTO #12149]

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

patch.py: Use git format-patch with --no-signature --no-numbered params

* --no-signature saves unnecessary .patch modifications when executed on
  host with different git version
* --no-numbered saves unnecessary .patch modifications when number of the
  applied patches is changed (the number is still in the filename so the
  order how they should be applied is still preserved)
* both options exist for very long time, I've tested them with git 1.9.1
  from Ubuntu 14.04 and I'm quite sure they were available even in much
  older releases, so there shouldn't be any issue on relatively new sanity
  tested distros

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

package_manager: format pydoc comments properly

Signed-off-by: Ross Burton <ross.burton@intel.com>

kernel.bbclass: set HOSTLDFLAGS in KCONFIG_CONFIG_COMMAND

Kernel v4.14 and newer contain the following in their Makefile:

HOST_LFS_LDFLAGS := $(shell getconf LFS_LDFLAGS)
HOSTLDFLAGS  := $(HOST_LFS_LDFLAGS)

This breaks our menuconfig, because it can no longer find ncurses if its
not on the host machine. This can be seen in linux-yocto-dev, for
example:

[clsulliv@clsulliv build]$ bitbake virtual/kernel -c menuconfig

  GEN     ./Makefile
  HOSTLD  scripts/kconfig/mconf
/home/clsulliv/yocto/poky/build/tmp/hosttools/ld: cannot find -lncurses
/home/clsulliv/yocto/poky/build/tmp/hosttools/ld: cannot find -ltinfo
collect2: error: ld returned 1 exit status
make[3]: *** [scripts/Makefile.host:99: scripts/kconfig/mconf] Error 1
make[2]: *** [/home/clsulliv/yocto/poky/build/tmp/work-shared/intel-corei7-64/kernel-source/Makefile:504: menuconfig] Error 2
make[1]: *** [Makefile:146: sub-make] Error 2
make: *** [Makefile:24: __sub-make] Error 2
Command failed.
Press any key to continue...

Fix this by setting HOSTLDFLAGS to ${BUILD_LDFLAGS} in our
'make menuconfig' command.

Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

python-numpy: update to 1.14.2

Update to the latest stable release

Signed-off-by: Derek Straka <derek@asterius.io>
Signed-off-by: Ross Burton <ross.burton@intel.com>

python3-pip: update to 9.0.2

Update to the latest stable release

Signed-off-by: Derek Straka <derek@asterius.io>
Signed-off-by: Ross Burton <ross.burton@intel.com>

python3-pygobject: update to 3.28.1

Update to the latest stable release

Tested in qemux86-64 running core-image-minimal

Signed-off-by: Derek Straka <derek@asterius.io>
Signed-off-by: Ross Burton <ross.burton@intel.com>

python*-setuptools: update to 39.0.0

Update the python{3}-setuptools to the latest stable version

Tested on the qemu with core-image-minimal

Signed-off-by: Derek Straka <derek@asterius.io>
Signed-off-by: Ross Burton <ross.burton@intel.com>

python-numpy: Fix build for riscv64

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

python3-native: update to version 3.5.5 to fix security issues

License-Update: checksum change is due to bump in copyright year

Resolves CVE-2017-1000158 and other potential security issues

See https://docs.python.org/3.5/whatsnew/changelog.html#python-3-5-5-final

Signed-off-by: Derek Straka <derek@asterius.io>
Signed-off-by: Ross Burton <ross.burton@intel.com>

python3-native: fix build races in 'make regen-all'

Python 2.x and target Python 3.x do not have this issue.

[YOCTO #12596]

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

gnome-desktop: do not assume time_t is long

Replace use of atol() to set a time_t variable.

Signed-off-by: Joe Slater <joe.slater@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

Revert "lttng-tools: Upgrade 2.9.5 -> 2.10.2"

None of the test suite works, so something quite fundamental broke.

[ YOCTO #12606 ]

This reverts commit 406ea737f1ce5bb5fed46c8cbeb5c7c13bf3b8ad.

Signed-off-by: Ross Burton <ross.burton@intel.com>

slang: 2.3.1a -> 2.3.2

Upgrade slang from 2.3.1a to 2.3.2.

Signed-off-by: Huang Qiyu <huangqy.fnst@cn.fujitsu.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

python3: add readline to PACKAGECONFIG

Provide a way to make readline dependency optional in case someone
might want to use BSD alternative editline instead.

Using editline would need some changes though (python issue13501).

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

ca-certificates: run postinst script only for -target package

Nativesdk package has a special arrangement where the same thing is done
in do_install(). It was assumed (in the comment) that postinsts don't run when
installing nativesdk packages, but this was incorrect: they are run, but
any failures were previously silently ignored. Now this missing failure reporting has
been fixed, and so we get to see the failures.

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

libpng: fix upstream version check

Sourceforge for some reason claims an older version is 'the latest'.

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

distcc: remove UPSTREAM_VERSION_UNKNOWN

Upstream has finally released a new version.

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

checklayer: remove reference to undefined class

LayerError doesn't exist and will lead to an error when this failure
code path is hit.

Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

gcc: Do not use --with-linker-hash-style if LINKER_HASH_STYLE is empty

We allow to set LINKER_HASH_STYLE to be empty so this would fail
since --with-linker-hash-style needs an argument and cant be empty

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

gcc: backport patch to fix ICE in MIPS64 target in G++

Backport a patch to fix an ICE when compiling for MIPS64.

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

buildperf: measure the size of core-image-sato rootfs

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

scripts/oe-build-perf-report: fix comparing arbitrary commits

Fix a crash when generating a txt report and the two commits to be
compared were not consecutive (but there were some tested commits
between them).

Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto/4.12: backport bugfixes for x86

Integrating the following commits:

  60b649971940 x86/hibernate/64: Mask off CR3's PCID bits in the saved CR3
  cec3c008ec8f drm/i915/cfl: Coffee Lake works on Kaby Lake PCH.
  073873cb152c brd: remove unused brd_mutex
  912c53b1b346 audit: fix memleak in auditd_send_unicast_skb.

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

maintainers: add maintainer for make-mod-scripts

I introduced the recipe, so I get to fix any bugs!

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

linux-yocto/4.14/4.15: revert "features/systemtap/systemtap.cfg: enable CONFIG_KERNEL_DEBUG"

The following commit:

   Author: California Sullivan <california.l.sullivan@intel.com>
   Date:   Wed Feb 17 16:47:10 2016 -0800

       ktypes: add developer ktype

       The developer ktype enables EMBEDDED, EXPERT, and DEBUG_KERNEL,
       opening up more kernel options and setting some defaults.

Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Was created to address the kernel footprint related issues that are related
to many of the kernel debug options.

When this commit was merged, it re-enabled CONFIG_DEBUG_KERNEL for the
standard kernel, since it includes the systemtap fragment. The correct thing
to do is to move systemtap properly into the developer kernel type.

For now, you can build the developer kernel type, or add the developer kernel
configuration fragment via a bbappend, and you'll have a functional systemtap.

[YOCTO #12603]

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

btrfs-tools: add RECIPE_NO_UPDATE_REASON

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

python3: Fix purelib install and runtime paths

oe-core commit: 45afadf0b6 fixed the pip problem with purelib for
python2, even though the the patch stated it was for python3.  This
patch addresses the purelib problem for python3.

If you install the package python3-pip you will have a pip3 binary
where you can see the problem on the device easily where the modules
install into the incorrect area and are not able to be referenced by
python3 at all.

Example error:
   pip3 install imutils
   pip3 list |grep imutils || echo ERROR no imutils
      ERROR no imutils
   python3 -c 'import imutils'
     Traceback (most recent call last):
       File "<string>", line 1, in <module>
     ImportError: No module named 'imutils'

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

insane.bbclass: Fix typos in 32bit risc-v machine type

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

insane.bbclass: add support for RISC-V baremetal

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

gtk-doc: use --srcdir instead of cd

Signed-off-by: Ross Burton <ross.burton@intel.com>

gtk-doc: always inherit python3native

My attempt at only inheriting python3native if it was needed was broken and
didn't work, so back it out and always inherit.

Signed-off-by: Ross Burton <ross.burton@intel.com>

libgpg-error: Fix multilib header conflict - pg-error.h and gpgrt.h

Header file conflict between 32-bit and 64-bit versions.

Signed-off-by: Zhang Xiao <xiao.zhang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

tiff: Fix multilib header conflict - tiffconf.h

Header file conflict between 32-bit and 64-bit versions.

Signed-off-by: Zhang Xiao <xiao.zhang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

valgrind: Fix multilib header conflict - valgrind/config.h

Header file conflict between 32-bit and 64-bit versions.

Signed-off-by: Zhang Xiao <xiao.zhang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

kmod: don't call gtkdocize twice

The gtk-doc class already calls gtkdocize, so we don't need to do it again

Signed-off-by: Ross Burton <ross.burton@intel.com>

maintainers: remove obsolete entry

Signed-off-by: Ross Burton <ross.burton@intel.com>

build-recipe-list: improvements

risc-v: Pin oldest kernel to 4.15 for the new kid

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

site/riscv: Shunt the realloc configure test for startup-notification

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

pcmanfm: Drop COMPATIBLE_HOST setting

We have all core architectures in there its almost redundant

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

tcmode-default.inc: Define QEMUVERSION and softly pin it

This provides ability to surgically override qemu for certain
machines which are provided by external layers.

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>

python-numpy: reorganize numpy recipes to use a common .inc file to reduce duplication

Signed-off-by: Derek Straka <derek@asterius.io>
Signed-off-by: Ross Burton <ross.burton@intel.com>

ncurses: fix deletion of /usr/lib/terminfo

Ncurses doesn't honour ${libdir} for terminfo, so try more options to remove it.

Signed-off-by: Koen Kooi <koen.kooi@linaro.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>

openssl: fix libdir logic to allow multiarch style paths

The recipes were using 'basename' to turn '/usr/lib' into 'lib', which breaks when libdir is '/usr/lib/tuple', leading to libraries ending up in '/usr/tuple', which isn't in FILES_*. Change the logic to use sed to strip the prefix instead.

Signed-off-by: Koen Kooi <koen.kooi@linaro.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>

ncurses: fix do_install failure when base_libdir has more than one level

Other sections of the .inc already use mkdir -p, so use it here as well.

Signed-off-by: Koen Kooi <koen.kooi@linaro.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>

uninative: Add compatiblity version check

If glibc is newer on the host than in uninative, the failure mode is
pretty nasty for clusters where the sstate is shared, including the Yocto
Project autobuilder.

This check aborts the use of uninative in such scenarios where a newer
glibc version appears and avoids corruption of sstate caches.

We use ldd to check the glibc version since that is included in libc-bin
(or equivalent) which locales use so it should always be present.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

build-recipe-list: build universe instead of world

Building world means recipes that are excluded from world build for whatever
reason get skipped from the manifests, which isn't useful.  Instead building
universe and pass -k so that the expected dependency failures are not fatal.

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

meta: remove some EXCLUDE_FROM_WORLD assignments

Now that we have recipe-specific-sysroots we don't need to exclude recipes from
world builds because they conflict with other recipes, as they'll all be built
with their own sysroots.

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

iproute2: fix rebuild failures

When rebuilding iproute2, many such instances of the following build failure
occur:

| make[1]: Entering directory '.../iproute2/4.14.1-r0/iproute2-4.14.1/lib'
| Makefile:1: ../config.mk: No such file or directory
| make[1]: *** No rule to make target '../config.mk'.  Stop.

Signed-off-by: Trevor Woerner <twoerner@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

btrfs-tools: set CLEANBROKEN to 1 to avoid rebuild failure

When rebuilding btrfs-tools, we would sometimes meet the following error.

  Makefile:43: *** Makefile.inc not generated, please configure first.

Set CLEANBROKEN to "1" to solve this problem.

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

python3: update to version 3.5.5 to fix security issues

License-Update: checksum change is due to bump in copyright year

Resolves CVE-2017-1000158 and other potential security issues

See https://docs.python.org/3.5/whatsnew/changelog.html#python-3-5-5-final

Signed-off-by: Derek Straka <derek@asterius.io>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>