packagegroup-self-hosted: move graphics packages to graphics group

The following packages are gui based tools:
  xdg-utils       # A set of common interfaces for desktop
  leafpad         # A GTK+ text editor used in Build Appliance
  midori          # A web browser used in Build Appliance
  pcmanfm         # PCManFm file manager used in Build Appliance
  vte             # Virtual terminal emulator GTK+ widget library

Move them to packagegroup-self-hosted-graphics so that we can make
a graphics free self-hosted image by excluding the
packagegroup-self-hosted-graphics.

(From OE-Core rev: bb7d081ccb5a352d3089b6b30ab740e62c05e791)

Signed-off-by: Jackie Huang <jackie.huang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

packagegroup.bbclass: Use immediate expanded PACKAGE_ARCH value

When the PACKAGE_ARCH is dependant of a override, the expanding is
done too late triggering the false-positive error of wrong inclusion
order.

To fix this we use immediate expansion operator to evaluate the
PACKAGE_ARCH value in order to choose to load 'allarch' class or not.

(From OE-Core rev: 2c9b1d304daade7b0907320aeb9c522e7ab9dcab)

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

oeqa/runtime: Automatic test for ptest

For images without ptest the packages are automatically installed alongside ptest-runner. Log results are saved in ./results folder.
No cleanup is done for packages after the test is finished.

(From OE-Core rev: f8e99fa8baa020c6414da19428b73c1fd30c9523)

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

oeqa/utils/logparser.py: results based log parser utility

A module for parsing results based logs like ptest, compliance and performance.
Supports breaking the logs into multiple sections and also provides a result object to use the parser with.
The parser is initialized with the regex required to identify results and section statements in the target log file.

(From OE-Core rev: 72308d030fc98982e88b121b052cd2438f6b031c)

Signed-off-by: Corneliu Stoicescu <corneliux.stoicescu@intel.com>
Signed-off-by: Lucian Musat <georgex.l.musat@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

oeqa: Added package installer to oetest to aid in future automatic install of packages

(From OE-Core rev: e0af8a3c751f938faf7deb60b51f42450ae58e27)

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

linux-yocto-dev: bump to v3.17+

Update the -dev recipes PV to reflect that the development tree is
now at v3.17-rc4.

(From OE-Core rev: 0a5353ef42814afea33ce5d69c1fe80e3b3257ba)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto/3.14: configuration updates and feature merges.

Updating the 3.14 meta and machine SRCREVs to pick up the following changes:

meta/config:

    a94680bfeb0e drm-gma500: build GMA500 DRM driver as kernel module
    21783e2e6be2 intel-common: intel-core-32: use tunnelcreek.scc
    0b46240ac3b3 soc: tunnelcreek: create tunnelcreek scc
    c7207d9c0e6a qemuppc64: create bsp config
    8f553f77e0ad meta: Create kernel config and scc for CRIU
    dacb1c1ab011 meta: common-pc: add support for booting via EFI
    b85e4cb574d6 meta: Add CONFIG_TTY to common-pc-drivers.cfg
    33f2c8d2ba33 meta: Add CONFIG_TTY to 8250.cfg
    e6f31e4894d5 GRE: enable gre feature.

Features:

    b85edae6fd61 net: e1000e calls skb_set_hash
    b45e6dec1972 net: ppp: use sk_unattached_filter api
    d310945fb6d8 tracing: accelerate tracing filters with BPF
    6742a0d5e218 net: filter: x86: internal BPF JIT
    66f2b151dd09 net: filter: x86: split bpf_jit_compile()
    3c82c5d1fc49 net: filter: Fix redefinition warnings on x86-64.
    5ad74ef546a4 net: filter: additional BPF tests
    f097814fc305 net: filter: BPF testsuite
    1bcefe39e229 net: filter: make BPF conversion more readable
    e75a3abd0c6f net: filter: misc/various cleanups
    f5cd96317979 net: filter: make register naming more comprehensible
    2f485870e68b net: filter: simplify label names from jump-table
    d381512d96f0 bpf_dbg: fix wrong register usage
    d99d91c2c5a9 sched, cls: check if we could overwrite actions when changing a filter
    8a03c23319dc net: filter: initialize A and X registers
    77a8a3fb86cb filter: added BPF random opcode
    a9bb9bcd5a04 net: filter: seccomp: fix wrong decoding of BPF_S_ANC_SECCOMP_LD_W
    724096236a68 filter: prevent nla extensions to peek beyond the end of the message
    41bdf9a8c75f net: filter: be more defensive on div/mod by X==0
    2f908136e311 net: filter: minor: fix kdoc in __sk_run_filter
    2c733311d285 net: filter: rework/optimize internal BPF interpreter's instruction set
    40155402aff6 net: filter: move filter accounting to filter core
    2635af59e1b6 net: filter: keep original BPF program around
    97147dbf46f8 net: filter: add jited flag to indicate jit compiled filters
    676d895e105a net: Rename skb->rxhash to skb->hash
    1e4f73c14307 tools: bpf_dbg: various misc code cleanups
    47956b735bf6 timerfd: Implement timerfd_ioctl method to restore timerfd_ctx::ticks, v3
    26b3ef638e06 docs: Procfs -- Document timerfd output
    af3430a5fdc8 timerfd: Implement show_fdinfo method
    5dd4e59bf04b mm: softdirty: clear VM_SOFTDIRTY flag inside clear_refs_write() instead of clear_soft_dirty()
    c2a643b6da96 mm: softdirty: don't forget to save file map softdiry bit on unmap
    8a98c67f7124 mm/memory.c: don't forget to set softdirty on file mapped fault
    5886c8ea1f26 mm: softdirty: make freshly remapped file pages being softdirty unconditionally
    8f4adad45df3 proc: show mnt_id in /proc/pid/fdinfo
    d7e814681dd4 gre6: Call skb_checksum_simple_validate
    f17fe1ff70a1 net: Generalize checksum_init functions
    b6d250d7bead ip6gre: add x-netns support
    aebe8ead74e5 gre: allow changing mac address when device is up
    98268e587a24 gre: add x-netns support
    923b14472945 yaffs: fix yaffs_vfs.c warnings.
    1cc6b7d49394 yaffs: Avoid setting any ACL releated xattr

(From OE-Core rev: 7c9a0f87cfe07ddc94e63eb6b8b3b547895ada80)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto/3.10: update valleyisland-io merge branch

Updating the topic branch that is used for valleyisland-io merges.

(From OE-Core rev: 4d804dc8d683024ef2a0a07200492f458f52b535)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto: introduce v3.17

With this commit, we introduce the v3.17-rcX kernels, which will be
released as part of the 1.7 yocto project.

All architectures have been compiled and booted with this revision
of the kernel.

(From OE-Core rev: 07315861450ebf3cd224f4700eb125b74057aa0f)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

perf: fix v3.17 powerpc compilation issues

The powerpc perf arch port has two issues in the current 3.17-rc
seriers. Undefined dwarf symbols and undefined pr_debug calls.

commit a60335ba [perf tools powerpc: Adjust callchain based on DWARF debug info]
introduces skip-callchain-idx.c, and breaks the perf build.

Until fixes get into the mainline kernel, we temporarily fix the issues by
adding a forced linke to -ldw and modifying skip-callchain-idx.c to include
the proper debug files for pr_debug.

(From OE-Core rev: 9f204a02f07f8e378e8a078db103a6ba94ec55cb)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

lttng-modules: fix compilation for 3.17-rcX

Updating the lttng-modules 2.5 recipe with backports from the master branch to
enable builds on 3.17 based kernels.

(From OE-Core rev: cc2ef9b0f6371837faa5e430c0b6c48bf24a8c4c)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto/3.10: baytrail i/o updates

Updating the 3.10 kernel with baytrail feature backports:

   e313fe1374d spi/pxa2xx: fix incorrect SW mode chipselect setting for BayTrail LPSS SPI
   da387aea86a8 spi/pxa2xx: change default supported DMA burst size to 1
   119e77242d80 spi/pxa2xx: fix runtime PM enabling order
   ac2ff9de705c spi/pxa2xx: Prevent DMA from transferring too many bytes
   2260921e859c spi: pxa2xx: remove unnecessary OOM messages
   59f141f4a558 spi/pxa2xx-pci: Add PCI mode support for BayTrail LPSS SPI
   25c8232cbb66 spi: remove DEFINE_PCI_DEVICE_TABLE macro
   6eb21efb3641 i2c: designware-pci: set ideal HCNT, LCNT and SDA hold time value
   4e2bd3da1739 i2c: designware-pci: add 10-bit addressing mode functionality for BYT I2C
   7b813400820f i2c: designware-pci: Add Baytrail PCI IDs
   5fe4585c09d1 usb: gadget: udc-core: move sysfs_notify() to a workqueue
   cdbdda47494a usb: gadget: don't fail when DMA isn't present
   8c1914c2e3bc mmc: sdhci: add DDR50 1.8V mode support for BayTrail eMMC Controller
   5ad460bf8d4a mmc: sdhci: Preset value not supported in Baytrail eMMC
   3359f2b4d2d4 mmc: sdhci: Allow for irq being shared
   0663d41d3b8c 8250_dw: Support all baudrates on baytrail
   1866808f61bf serial: 8250_pci: add support for Intel BayTrail
   472c007ce197 serial: 8250: don't change the fifo trigger level when using dma
   fadb6c538a2b serial: 8250_dma: check the result of TX buffer mapping
   e48701bee23b dmaengine: dw: don't perform DMA when dmaengine_submit is called
   9fda0d3bfdd8 dmaengine: dw: introduce dwc_dostart_first_queued() helper
   a0b8d0199e8f dma: dw: Add suspend and resume handling for PCI mode DW_DMAC.
   f1c1bd979f71 pwm: lpss: Fix const qualifier and sparse warnings
   d6aa3e78c2ea pwm: lpss: Add support for PCI devices
   a8fe7a1badcd pwm: add support for Intel Low Power Subsystem PWM
   cd9e2e0a810c ACPI / LPSS: Add Intel BayTrail ACPI mode PWM
   b8dab93f3e79 pwm: Add sysfs interface

(From OE-Core rev: 2c01ce4341b6745e2a778505d4840b64eb053e74)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: update acinclude.m4

Some macros have been added upstream and were not being executed, so refresh our
fork of this file.

(From OE-Core rev: 6988baed623c507aa44164801c5359411f462d84)

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

e2fsprogs: remove manual build of util/subst

The build system uses BUILD_CC itself now, so we don't need to do it manually.

(From OE-Core rev: 32ddff201eb00b1ffa244904251ad0ad39760b62)

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

e2fsprogs: enable verbose build

(From OE-Core rev: e29c5947c2618341a46dc125a155b80c07d2f9ae)

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

e2fsprogs: remove chmod

This chmod doesn't appear to be needed, and history fades away in 2005.

(From OE-Core rev: bade04d62468acbec2b54f38b5071a3b1070ab9a)

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

pcihp: fix possible array out of bounds, CVE-2014-5388

Prevent out-of-bounds array access on acpi_pcihp_pci_status.

[From QEMU: fa365d7cd11185237471823a5a33d36765454e16]

(From OE-Core rev: 43ba8f84fb03a8f645995b43966cb87152b69f4c)

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

bitbake.conf: use ??= for IMAGE_ROOTFS_SIZE

Previously, when building core-image-minimal, the rootfs size would
default to 64M because we use '?=' in bitbake.conf and also '?=' in
core-image-minimal.bb.

The thing is, we'd like to have a default value for all images set
in bitbake.conf but still allow each image recipe to set its own default
value which could be overridden by users in local.conf.

(From OE-Core rev: 18f499df6bcbf79d7bd0a99c4c8693268683485f)

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

kernel.bbclass: handle .dsp firmware

A few firmware files use this filename extension, like korg/k1212.dsp .
Create respective packages.

(From OE-Core rev: 4a4877c2e7cdcb8a1d0a54add600c0cd4e92e647)

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

classes/buildhistory: fix for sstate class change

SSTATEPOSTINSTFUNCS is now set with = in sstate.bbclass, and because
the line here in buildhistory.bbclass ends up being parsed before that
we now need to use _append or the value we are adding will be wiped out.
This fixes buildhistory no longer recording package information since
OE-Core revision 9d659c6f20fa4a141b491c62a3ef0dfb1f896d9c.

(From OE-Core rev: b5c05b474554925524073023f3aac2532c87537f)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

glibc: Incremental bump to 2.20 release branch

Now that glibc 2.20 has been released. We switch
to use release branch and remove the already applied
patch

(From OE-Core rev: 70bfccd8e13af712381b4feea3ef882369951264)

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gdb: allow compilation for __aarch64__

asm/ptrace.h must be included in two files.

(From OE-Core rev: dd776e2aacb4e75f2195b8943a9594131592a0da)

Signed-off-by: Joe Slater <jslater@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

wic: Completely remove all urlgrabber references

The previous commit, 'wic: Remove 3rdparty/urlgrabber' didn't actually
remove all references to urlgrabber - this does however.

(From OE-Core rev: b4367af770611174d9de82c63c697d4347e30dcf)

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

valgrind: backport enabling for glibc 2.20

Now that we're running glibc 2.20 not 2.19, backport the configure.ac fixes from
upstream.

(From OE-Core rev: 3a98dd15f19034a1837210f764310c195f9906e1)

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

systemd: disable resolved and networkd

These two services are fairly immature and need careful integration into the
rest of the system, so disable them for now.

(From OE-Core rev: 23efb1755c3074e17c547f9b9fd13564cc6d8566)

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

alsa-utils: hard-code alsa-utils in SRC_URI

This recipe is used by alsa-utils-alsaconf so don't use ${BP} in the SRC_URI.

(From OE-Core rev: f14ce47604cfbcc930d57567354000845aefb739)

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

qemu.bbclass: add -r ${OLDEST_KERNEL} parameter

* in some cases (e.g. with external toolchain which doesn't respect our
  reasonably old version set in OLDEST_KERNEL) it's possible to have libc
  which requires newer kernel than what we have on builders, qemu supports
  -r param to use different uname than what's returned by host system.
* change qemu_run_binary to pass -r ${OLDEST_KERNEL} and add the same to
  QEMU_OPTIONS which are used by qemuwrapper-cross
* maybe we should eventually convert all qemu_run_binary usages always include
  qemuwrapper-cross dependency and always call qemu through that (it
  seems very strange that qemu_target_binary is called from qemuwrapper
  and for allarch recipes it can return qemu-allarch as qemu binary).

  qemu_run_binary is used by:
  meta/classes/gtk-immodules-cache.bbclass:            ${@qemu_run_binary(d, '$D', '${bindir}/gtk-query-immodules-$maj_ver.0')} \
  meta/classes/qemu.bbclass:def qemu_run_binary(data, rootfs_path, binary):
  meta/recipes-core/systemd/systemd_213.bb:         ${@qemu_run_binary(d, '$D', '${base_bindir}/udevadm')} hwdb --update \
  meta/recipes-graphics/pango/pango.inc:    ${@qemu_run_binary(d, '$D','${bindir}/${MLPREFIX}pango-querymodules')} \

  and qemuwrapper directly by:
  scripts/postinst-intercepts/update_font_cache:PSEUDO_UNLOAD=1 qemuwrapper -L $D -E LD_LIBRARY_PATH=$D/${libdir}:$D/${base_libdir}\
  scripts/postinst-intercepts/update_pixbuf_cache:PSEUDO_UNLOAD=1 qemuwrapper -L $D -E LD_LIBRARY_PATH=$D/${libdir}:$D/${base_libdir}\

(From OE-Core rev: 1f0bff320077f4d9f2ee51096a1438e8cae9dd0d)

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

shadow: add Qualcomm and STMicroelectronics SoCs

Reported-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
Reported-by: Peter Griffin <peter.griffin@linaro.org>
(From OE-Core rev: 8c811541125fff75429d09249299add5ffe80912)

Signed-off-by: Fathi Boudra <fathi.boudra@linaro.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

glib-2.0: fix mount detection

Currently a USB stick gets mounted on /run/media/sdX, where X is the
letter corresponding to a certain drive. In older builds of core-image-sato
it got mounted in /media/sdX. This change made the drives to not be
displayed in graphical applications which use glib-2.0 such as the
File Manager (pcmanfm). The reason is that the gio component of glib-2.0
which handles mounts contains a function which determines if a mount
is worthy to be displayed in the UI or not. The function is called
g_unix_mount_guess_should_display().

It expects a drive to be mounted in /run/media/<username>/sdX, but
in the current build the username part is missing in case a root user
is authenticated in the system.

The easiest solution is to allow the display of drives mounted in the
path used by the current configuration and that is /run/media/sdX.

[YOCTO #6492]

(From OE-Core rev: da9bbcd2233057a0529005b9b2cc34aff65bc6e2)

Signed-off-by: Marius Avram <marius.avram@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

python-native: disable user site support

The user site-packages gets inserted into sys.path ahead of the
system site directories, so a site package installed there will
be used in preference to what's in our sysroot, causing less
deterministic builds, and potential build breakage, depending
on what the user has installed there. Disable it for our native
python, so they don't affect our builds.

(From OE-Core rev: c448bf3629ab5d930ed845d4ba48e37e4a85d2a3)

Signed-off-by: Christopher Larson <kergoth@gmail.com>
Signed-off-by: Yasir-Khan <yasir_khan@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gstreamer: Set DEPENDS dynamically for plugins-ugly and plugins-bad

Only add plugins-ugly or plugins-bad to the DEPENDS if they are
added to the COMMERCIAL_AUDIO_PLUGINS or COMMERCIAL_VIDEO_PLUGINS.

(From OE-Core rev: e8eb01fb49bf18dedb62ef05712374cec3bd5efd)

Signed-off-by: Drew Moseley <drew_moseley@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bluez5: Allow method calls over dbus for bluetooth daemon

This patch changes the dbus policy settings in order to allow
calling methods over dbus. bluez5 and pulseaudio rely on this
mechanism to configure media end points. Previously bluetoothd
was failing to configure the media end points with error
org.freedesktop.DBus.Error.AccessDenied, due to which bluetooth
headset and smartphone were unable to connect in A2DP mode.

(From OE-Core rev: eb3a380301b8f7ec34d3649c386214d39355eaa0)

Signed-off-by: Yasir-Khan <yasir_khan@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

alsa-utils: upgrade to 1.0.28

Changelog:
http://www.alsa-project.org/main/index.php/Changes_v1.0.27.2_v1.0.28

0001-alsactl-don-t-let-systemd-unit-restore-the-volume-wh.patch updated.

(From OE-Core rev: 5a3331b7ee8bb565ebb3823fc23123120ccd653d)

(From OE-Core rev: 162fa884c02b77517ebb131de314b38cf9a1244f)

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

systemd: Adding RRECOMMENDS for os-release

Systemd depends on the os-release hence adding
RRECOMMENDS for the systemd package.

[ RB - updated for 216 ]

(From OE-Core rev: 003d16fbd54f59422aab917b8f3bf2aec4ab50b6)

Signed-off-by: Christopher Larson <kergoth@gmail.com>
Signed-off-by: Sujith H <Sujith_Haridasan@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

os-release: Adding a new recipe for operating system identification

The /etc/os-release will have the operating system identification data.
Tested on target with systemd enabled. Here is the sample file looks
in the file:

ID=poky-ivi-systemd
NAME=Yocto GENIVI Baseline (Poky/meta-ivi)
VERSION=6.0+snapshot-20140721 (daisy)
VERSION_ID=6.0+snapshot-20140721
PRETTY_NAME=Yocto GENIVI Baseline (Poky/meta-ivi) 6.0+snapshot-20140721 (daisy)

(From OE-Core rev: 8a12350e00b5443a24f3d6c7693fdde9fcb7319f)

Signed-off-by: Christopher Larson <kergoth@gmail.com>
Signed-off-by: Sujith H <Sujith_Haridasan@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pixman: upgrade to 0.32.6

Revert a patch to the MIPS-specific code that was causing compile failures.

(From OE-Core rev: ace56dcd6eed1ab92af9bdf5bd6d9b11e35a0917)

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

gtk+3: update to 3.12.2

fix-build-when-wayland-backend-enabled.patch removed as the issue
is fixed in upstream.

(From OE-Core rev: 59fe6f91b39c4aa67e7e1734c3731d49c7ea8376)

Signed-off-by: Marko Lindqvist <cazfi74@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pango: update to 1.36.6

(From OE-Core rev: 715f62c5960263ecb0b181cab16ed28345f05189)

Signed-off-by: Marko Lindqvist <cazfi74@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libpng: update to 1.6.13

License terms have not changed. In case of LICENSE and png.h lines 0-15
the version numbers and release dates have changed. In case of png.h lines
229-343 the text has not changed at all, but old line numbers were not
pointing to correctly to the licensing terms segment.

(From OE-Core rev: efb760b19f3f88f509bfbc44a08c9c8c03baf6bc)

Signed-off-by: Marko Lindqvist <cazfi74@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libpng: drop unnecessary automake requirement lowering patch

Automake recipe provides automake 1.14.1 anyway

(From OE-Core rev: 3e8ce8508356112e07e37734d228494a840cf04c)

Signed-off-by: Marko Lindqvist <cazfi74@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gtk+2: update to 2.24.24

(From OE-Core rev: 64a433278d525859c661d8a1e50cc0817be12ac3)

Signed-off-by: Marko Lindqvist <cazfi74@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libxkbcommon: Add missing dependency on xkeyboard-config

libxkbcommon depends on xkeyboard-config

(From OE-Core rev: ca14b0c0b5a6653336c79d8ec80e2cdef3f9239e)

Signed-off-by: Mikhail Durnev <mikhail_durnev@mentor.com>
Signed-off-by: Abhijit Potnis <abhijit_potnis@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sstate: Add extra intercept functions SSTATEPOSTCREATEFUNCS

In some cases we do either need to add extra sstate manipulation
functions, or change the existing modification functions. This patch
parametrizes it to SSTATEPOSTCREATEFUNCS after sstate_create_package

(From OE-Core rev: ed5fb9c69fcf0f6e8e98f72ac753c66119468ea5)

Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_manager.py: use RPM_PREFER_ELF_ARCH

Use RPM_PREFER_ELF_ARCH to instead of RPM_PREFER_COLOR as rpm upstream
suggested, and use "4" to instead of "3" since it is a bit mask (not
enumeration), so we need "4" here.

(From OE-Core rev: 8dcfa7143a6690455c3bd5772621f142f2d9e559)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

perf: add libunwind support

Add a new feature named 'perf-libunwind'. Adding this support to perf allows it
to do stack traces on ARM - thumb2 and MIPS targets. PERF_FEATURES variable in
perf-features.inc will enable the perf-libunwind.

(From OE-Core rev: 10dfa4ee4e05841be3d3caaa28778aa40b782f97)

Signed-off-by: Chong Lu <Chong.Lu@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libunwind: add recipes

Add libunwind from meta-oe to oe-core, since perf depends on it.

(From OE-Core rev: aefddc23eeda9b46c1bb3ac776c8bff15f89707c)

Signed-off-by: Chong Lu <Chong.Lu@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

lighttpd: install config file without execute permissions

(From OE-Core rev: de8d70ccd846167df31037b7a89f7d13f6731cf6)

Signed-off-by: Dan McGregor <dan.mcgregor@usask.ca>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sudo: make sudoers a config file

(From OE-Core rev: c840369a65ee5041154e12fdd8919938856a0a6c)

Signed-off-by: Dan McGregor <dan.mcgregor@usask.ca>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

insane.bbclass: add QA check: file-rdeps

The ipk or deb can't depend on file such as "/bin/bash" or
"/usr/bin/perl", so it knows nothing about the pkg depends bash or perl,
thus there would be dependencies problems when we run "apt-get
install/remove <pkg>" on the target, this check can help us find the
issues and then fix them manually.

* Benefits:
  - Help to fix file rdepends issues for ipk and deb
  - Help to fix abnormal rdepends.
  - Help to check whether the rdepends is OK or not after build each
    recipe (don't have to install to the image), for example, a recipe may
    generate 10 binary packages, only a part of them will be installed to
    the image by default, we can know whether the rdepends are OK or
    not for the ones which are installed, but can't know the ones which
    are not installed, this patch can help check all the 10 binary
    packages' rdepends.

* Basic designs:
  - Get all the RDEPENDS on the chain.

  - Get the pkg's FILERPROVIDES from oe.packagedata.read_subpkgdata()
    and save to set filerdepends.

  - Get each RPDEPENDS' FILERPROVIDES, RPROVIDES and FILERPROVIDESFLIST,
    and save to set rdep_rprovides.

  - Do the set "filerdepends -= rdep_rprovides" and QA issue if
    filerdepends is not null.

[YOCTO #1662]

(From OE-Core rev: cd5e0f01cdb4e7c759c01cf3f87952a20253737c)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_manager: RpmPkgsList: determine rpm version

Do not set the rpm_version to 5 if it is not provided, instead
determine the real rpm version.

(From OE-Core rev: 48beaee2e5b2b4ae35c596c19f8a38e0ff4427e9)

(From OE-Core rev: fe4ad726d0dc63be73f885598e216b8eb927668a)

Signed-off-by: Kévin THIERRY <kevin.thierry@open.eurogiciel.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: tinfoil: add a means of enabling variable history tracking

Unfortunately it seems like the external use of the cooker
enableDataTracking() function broke at some point since the code that
reads it now runs within BBCooker's constructor. Since this now has to
be done early, add a parameter to Tinfoil's constructor to allow
enabling variable history tracking.

Fixes [YOCTO #6676].

(Bitbake rev: a9439b136f55f3f0e80ff053cd3b159da69ba362)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: fix some code spacing issues

Code is related to the basetable templates files.
It had mixed tabs and spaces and was miss aligned in various places,
making it hard to read.

(Bitbake rev: cdaea8951df6b707afd1fefbf22295088256dd6f)

Signed-off-by: Marius Avram <marius.avram@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: use cookies for count and sorting in templates tables

Until now cookies were used to save which columns were shown and which
were hidden in toaster tables. The tables from the templates also have
functionalities like sorting the entries on a certain column and
limiting the number of entries displayed on a page. The later however
were not saved using cookies. This patch brings this new feature.

The cookies are not saved only in the front-end. They are saved both
in the frontend in case the user uses the inputs/buttons to change
a parameter and also in the backend in case the user specifies manually
using GET variables the value of the parameters.

When no GET parameters are given the views will redirect the url to one
containg the parameters saved as cookies. When no cookies exist, default
values will be used.

[YOCTO #6126]

(Bitbake rev: 880b58c845e3a501fa90d24e1bd89c87ca84b709)

Signed-off-by: Marius Avram <marius.avram@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: do not save objects in session

In order to avoid problems when using JSON serializer for
saving sessions, we move from storing the objects themselves
in the session to storing the object id and reloading the
object when retrieved.

This allows, for example, to use cookie-storage sessions if the
infrastructure owner so desires.

(Bitbake rev: 39d0f0c2e87d4b161f1eeaa2657e61b5a6bc9ee2)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: enable SSH-based remote build support

We enable support for starting builds on remote machines
through SSH. The support is limited to poky-based distributions.

We refactor localhost build support and we update
bldcontrol application tests to uniformely test the APIs
of localhost and SSH build controllers.

[YOCTO #6240]

(Bitbake rev: c2ad9c9bb83f61c171434324df8c4d5ee655a556)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: bitbake server listen on all interface

We change the toaster starting script to make the
bitbake server listen on all interfaces on the local machine.

This is needed to be able to receive a controlling client
running on a remote machine.

(Bitbake rev: 137179eafca8d1a5a69b6302f8cc8961be3b45c4)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: rename bldviewer projecttags custom tagset

We rename the projecttags in bldviewer.templatetags to
simple_projecttags in order to avoid conflict with the
similarly named tagset in toastergui.

The conflict leads to an intermittent bug where proper
tags are not read correctly since Django uses only the
module name as global tag library identificator.

(Bitbake rev: a37f2c194d7e59611177cb8755524b7ad702fe91)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

alsa-lib: libasound should runtime depends on alsa-conf

Make libasound runtime depends on alsa-conf to install alsa-conf to rootfs.

(From OE-Core rev: e72aac2ae9e5fce1715fa04b7e94034fd06892d9)

Signed-off-by: Zidan Wang <b50113@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

rpmresolve: ignore null requires

It is not a problem if a package requires nothing (similar to RDEPENDS
is null), for example, these packages depends on nothing:

[snip]
alsa-conf-base
base-files
eglibc-binary-localedata-en-us
xserver-xf86-config
[snip]

The rpmresolve-native's algorithm is:
===fake code
for pkg in pkg1, pkg2, pkg3:
    rc = get_req(pkg)
return rc
===fake code

Suppose of the 3 pkgs requires null:
- We are lukcy if pkg1 or pkg2 is null, nothing happend.
- We are *not* lukcy if pkg3 is null, and will get the error when "INHERIT +=
  'buildhistory'":
  ERROR: Cannot get the package dependencies. Command
  '/path/to/x86_64-linux/usr/bin/rpmresolve -t /path/to/var/lib/rpm' returned 1:

This patch fixes the problem.

(From OE-Core rev: 2f234160ff4e9eccd8794a31df851b96328f6b74)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemtap: Cross compilation fix

This is a cross compilation fix. It allows systemtap to find the
kernel map file in the right place, i.e. in the kernel build tree.
Without this fix it takes a map file from the build host, if available.

(From OE-Core rev: 787bed708676fc04aee2850825e803273152f657)

Signed-off-by: Mikhail Durnev <mikhail_durnev@mentor.com>
Signed-off-by: Muzaffar Mahmood <muzaffar_mahmood@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemtap: Fixed probe syscall.sendfile failure

Marked __syscall.compat_sendfile as optional

(From OE-Core rev: 74bfb662246411aa062cd8b2acd37be1bbed39d3)

Signed-off-by: Mikhail Durnev <mikhail_durnev@mentor.com>
Signed-off-by: Muzaffar Mahmood <muzaffar_mahmood@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

useradd-staticids.bbclass: Fix for Bug 6633

When using the useradd-staticids.bbclass under meta/classes,
this error occurs:
"<username> - <username>: Username  does not have a static uid defined."
There was a problem with the regular expression for parsing parameters,
it was sometimes returning an empty string.

I have fixed this by skipping empty strings.

(From OE-Core rev: f249ef32709069a2680b92dc5a5b4f6545d014b7)

Signed-off-by: Fabrice Coulon <fabrice@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_rpm: Add optional improved directory handling

During spec generation, ideally directories should not be auto
packaged under the %file section of rpm packages but take ownership of
specific directories.

* packages only empty directories or explict directory.
   See:
       - http://www.rpm.org/max-rpm/s1-rpm-inside-files-list-directives.html
       - "The %dir Directive"

* This will prevent the overlapping of security permission.
   For example, in Tizen the directory /etc have smack label 'System::Shared'
   So Only one package should own and set the label of /etc to prevent
   the overwriting of the smack label.

Existing behaviour is maintained if DIRFILES is not set. If it is set,
the modified behaviour is used. If can be set to an empty value by
core recipes to trigger the modified behaviour.

[RP: Modified to allow optional usage of DIRFILES]
(From OE-Core rev: 0e33d232916125ba5305ced7200cc00f8b5f7b22)

Signed-off-by: Ronan Le Martret <ronan@fridu.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: enable forwarding messages to syslog daemon

In systemd_216, journald by default doesn't forward messages to
syslog daemon. This breaks the oeqa testing cases. So we enable
forwarding the messages for now.

(From OE-Core rev: 67ddca4a5533328ff95ab61610b6a1db24f57e35)

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

run-postinsts.service: remove redundant line

The basic.target itself requires sysinit.target, so there's no
need to write this redundant line in the run-postinsts unit file.

(From OE-Core rev: 28835fcd78fd35db6d3bd7ee2be16904ddf11991)

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

modutils-initscripts: mask modutils in case of systemd

We need to prevent the init script from running via systemd because
its functionality has been implemented interanlly in systemd.

In a systemd system, we can add configuration files under
/etc/modules-load.d/ to make things work.

(From OE-Core rev: 8fc304a59bb3977146645fe2c3570a3aaa4aac39)

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

psplash: mask psplash in case of systemd

We need to prevent the psplash init script from running via systemd
because the script is supposed to be used by sysvinit system and it
doesn't have any real effect in a systemd based system.

(From OE-Core rev: 2e847db3ac3333a68b721b11624f4f515bda7ccf)

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

v86d: mask fbsetup when necessary

When 'sysvinit' and 'systemd' are both in DISTRO_FEATURES, we need
to prevent the init script from running via systemd.
This is because that the functionality has been implemented internally
in systemd.

(From OE-Core rev: 2236678e38b22b763de8322d90585cbf648a448c)

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

keymaps: mask keymap when necessary

When 'sysvinit' and 'systemd' are both in DISTRO_FEATURES, we need
to prevent the init script from running via systemd.
This is because that the functionality of the init script has implemented
in systemd internally.

(From OE-Core rev: 8cfba07e24dae3d1837ccb5cb04e11f362519b0a)

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

initscripts: mask several init scripts

We now support executing scripts under /etc/rcS.d. So we need to
mask several more init scripts here to prevent them from running
at boot time.

(From OE-Core rev: c8f9527f18e5dd813c0330ba409875d34c36f6ab)

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

packagegroup-core-boot: conditionally rdepend on VIRTUAL-RUNTIME_initscripts

If we are building a systemd image with 'sysvinit' not in DISTRO_FEATURES,
the initscripts should not be installed into the image, as they are useless.
What's more, we as now support executing scripts under /etc/rcS.d, the
boot time will be increased by these scripts as the systemd will try to
translate them into temporary service files.

These init scritps are actually needed only when 'sysvinit' is in
DISTRO_FEATURES.

(From OE-Core rev: 74b931909e58fc581bb52c1af1f3ae26602980f3)

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

keymaps: remove dependency on initscripts

The 'keymaps' package actually doesn't have runtime dependency on
VIRTUAL-RUNTIME_initscripts, thus removing this dependency.

(From OE-Core rev: 2cf66a552a4966cb9c2f2475c622560cbe9db07b)

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

at: inherit update-rc.d to handle sysv init script

Inherit update-rc.d.bbclass to handle sysv init script.
Also make the 'at' daemon start in runlevel 2,3,4,5 instead of S.

(From OE-Core rev: a761d5fcd1880c8cee96f1f3198093fd92d08c77)

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

at: remove dependency on initscripts

The 'at' package actually doesn't have runtime dependency on
VIRTUAL-RUNTIME_initscripts, so remove this dependency.

(From OE-Core rev: 38359227215ad9f2d55740e25405f187771131d0)

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

rpcbind: avoid entering failed status after stopping daemon

Exiting with '2' is actually expected with rpcbind, because it catches
SIGTERM and exits with '2' explicitly.

The related code is as follows.

(void) signal(SIGTERM, terminate);

/*
 * Catch the signal and die
 */
static void
terminate(int dummy /*__unused*/)
{
        close(rpcbindlockfd);
        unlink(_PATH_RPCBINDSOCK);
        unlink(RPCBINDDLOCK);
        syslog(LOG_ERR,
                "rpcbind terminating on signal. Restart with \"rpcbind -w\"");
        write_warmstart(); /* Dump yourself */
        exit(2);
}

(From OE-Core rev: 1bfc6a45e7ba81d3537ea3ae8b176f5a9c206eaa)

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

cups: add systemd support

Add systemd unit files.

Also remove sysvinit related files if 'sysvinit' is not in
DISTRO_FEATURES.

(From OE-Core rev: 020065d3c57ccdc86c47cd0fc288071cdd194bbc)

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

cups: make cups daemon start correctly

The cups daemon needs the lpadmin group to start correctly.

Also add 'procps' to RDEPENDS if 'sysvinit' is in DISTRO_FEATURES,
because the init script shipped with cups needs it.

(From OE-Core rev: a8b2f086034585f3e115db3055575833922e3a59)

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

acpid: upgrade to 2.0.22 and add systemd support

For acpid needs to upgrade to work correctly with the new kernel.
The new version now uses autotools.

Also add systemd unit file.

(From OE-Core rev: 93805abedb5563a91886f330153b179033a29d11)

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

dhcp: add systemd service files

Add dhcpd.service and dhcrelay.service to support systemd systems.

(From OE-Core rev: 5f40081b0b9867bc167057712b04b3c1afb8a20d)

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

systemd: add support for executing scripts under /etc/rcS.d

This patch adds support for systemd to execute scripts under /etc/rcS.d.

To be compitable, all services translated from /etc/rcS.d/ scripts would
run before services translated from /etc/rcN.d scripts.

[YOCTO #5159]

(From OE-Core rev: 90bb8e8f9bc2454590d230b209fc749ea7270b9e)

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

v86d: set INHIBIT_UPDATERCD_BBCLASS if 'sysvinit' not in DISTRO_FEATURES

Set INHIBIT_UPDATERCD_BBCLASS to "1" if 'sysvinit' is not in DISTRO_FEATURES.

The functionality of the init script 'fbsetup' is implemented internally in
systemd. So fbsetup is not installed if 'sysvinit' is in DISTRO_FEATURES.

That's why we need to set INHIBIT_UPDATERCD_BBCLASS to "1" to avoid
generation of update-rc.d related preinst/postinst scripts.

(From OE-Core rev: 640a2231597d74517527ebedb6fd141278ba06f4)

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

systemd: add kbd-keymaps to RRECOMMENDS of systemd-vconsole-setup

systemd-vconsole-setup.service not only configures font but also
keymaps.

(From OE-Core rev: ffcd4bf2198171c9663f72d4499d5d55c29abf40)

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

keymaps: fix for systemd

Only install 'keymap.sh' script if 'sysvinit' is in DISTRO_FEATURES.
This is because systemd internally provides a mechanism to implement
the functionality this script tries to implement. Please see
systemd-vconsole-setup.service for more details.

Also we set 'INHIBIT_UPDATERCD_BBCLASS' to "1" if 'sysvinit' is not
in DISTRO_FEATURES.

(From OE-Core rev: 5e0aa114241b956f61cf852e1f4ff41b576adff4)

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

alsa-state: fix pkg_postinst and set INHIBIT_UPDATERCD_BBCLASS

As alsa-state inherits update-rc.d, the update-rc.d part inpostinst script
has already been handled in update-rc.d.bbclass.

Besides, we need to set INHIBIT_UPDATERCD_BBCLASS to "1" if 'sysvinit'
is not in DISTRO_FEATURES. This is because the init script is not installed
in such situation, adding update-rc.d part in the preinst/postinst scripts
just makes no sense.

(From OE-Core rev: 9b9da6667b606a56c89abc1b458f52dcca9774e9)

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

update-rc.d: fix logic in populate_packages_updatercd

The 'and' logic for the check is wrong. To make things clear, please
see the example below.

Say that we have a recipe A, which has a sysv-init style init script named
initA and no corresponding service file. The recipe inherits update-rc.d,
but it doesn't inherit systemd.bbclass. The DISTRO_FEATURES has 'systemd'
inside it, but it doesn't have 'sysvinit'. Now if we build an image, with
the 'and' logic in the check, the symlinks for initA would not be installed
into /etc/rc?.d directory.
This is incorrect. Because there's no corresponding service. The symlinks
in /etc/rc?.d/ should be created so that the service would be correctly
started at boot time.

The logic should really be 'or' in the check. This is actually what the code
was when it was originally written.

Several different situations are listed below to prove the correctness of the
'or' logic.

If 'sysvinit' is in DISTRO_FEATURES, the initA script would always be installed
with corresponding preisnt/postinst generated and added.

If 'sysvinit' is not in DISTRO_FEATURES, we have three situations.
1) A has initA and A.service.
   In such situation, systemd.bbclass would set INHIBIT_UPDATERCD_BBCLASS,
   so no preinst/postinst about update-rc.d would be added and the symlinks
   for initA would not be created.

2) A has initA, and the functionality of initA is not implemented internally
   in systemd.
   In such situation, symlinks for initA would be installed.

3) A has initA, and the functionality of initA is implemented internally in
   systemd or in some other recipe.
   Examples for such situation are alsa-state and keymaps in OE.
   In such situation, we need to set INHIBIT_UPDATERCD_BBCLASS in the recipe
   so that there would be no preinst/postinst scripts about update-rc.d added.

(From OE-Core rev: 6272c81509bbdb5d602056ceaa6f2c02bbf47a41)

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

systemd: make runlevel work in non-runlevel targets

Previously, after booting into the targets like multi-user.target or
graphical.target, the output of `runlevel' command is 'unknown'.

This is confusing for users. Normally, we would expect mutli-user.target
would have a `runlevel' output of 'N 3'.
This is the behaviour of Fedora20.

This patch installs symlinks for systemd-update-utmp-runlevel.service
in do_install task to fix the above problem.

(From OE-Core rev: cb7e692ec895ff9c15966faf29c9a84c0e78cdf5)

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

systemd: add PACKAGECONFIG for 'journal-upload'

Add PACKAGECONFIG for 'journal-upload' and disable it by default.
Once enabled, it will add 'curl' to its dependency.

(From OE-Core rev: b52c028252270c7bddc71554089fb184e52c5870)

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

systemd: upgrade to 216

Upgrade systemd from 213 to 216.

systemd-older-kernel.patch is removed as it's fixed in 216.

0001-uClibc-doesn-t-implement-pwritev-preadv.patch is removed because
the file it patches has been removed from the project.

0001-util-Including-missing.h-to-get-MAX_HANDLE_SZ.patch is removed because
it has been merged.

0001-missing.h-add-fake-__NR_memfd_create-for-MIPS.patch is backported
to fix compilation error for mips.

Below are the required kernel versions for systemd 216.
        Linux kernel >= 3.0
        Linux kernel >= 3.3 for loop device partition support features with nspawn
        Linux kernel >= 3.8 for Smack support

(From OE-Core rev: 58c432b2725ad8b78d02cb91a6ab405d425c01cc)

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

qemu-native: only depends on libxext-native when x11

Fixed when build without x11 in DISTRO_FEATURES:
ERROR: libxext-native was skipped: missing required distro feature 'x11' (not in DISTRO_FEATURES)

(From OE-Core rev: 012c7fbe0a21e469db5112873e778ff7c3b17237)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

piglit: add bash to RDEPENDS_piglit

Bash:
piglit/usr/lib/piglit/tests/asmparsertest/make_test_list.sh:#!/bin/bash
piglit/usr/lib/piglit/tests/asmparsertest/vp-tex.sh:#!/bin/bash
piglit/usr/lib/piglit/tests/spec/glsl-1.10/variable-index-read.sh:#!/bin/bash
piglit/usr/lib/piglit/tests/spec/glsl-1.10/variable-index-write.sh:#!/bin/bash
piglit/usr/lib/piglit/tests/spec/arb_shader_texture_lod/compiler/make_tex_lod_tests.sh:#!/bin/bash
piglit/usr/lib/piglit/tests/spec/glsl-1.20/execution/outerProduct-const.sh:#!/bin/bash
piglit/usr/lib/piglit/tests/spec/glsl-1.20/execution/outerProduct.sh:#!/bin/bash
piglit/usr/lib/piglit/tests/spec/glsl-1.20/compiler/built-in-functions/outerProduct-invalid-parameters.sh:#!/bin/bash
piglit/usr/lib/piglit/tests/glslparsertest/glsl2/make_tex_lod_tests.sh:#!/bin/bash
piglit/usr/lib/piglit/tests/glslparsertest/glsl2/make_tex_rect_tests.sh:#!/bin/bash

The bash scripts are in the tests directory, we had planned to move the
tests into the ptest pkg, but that would make piglit unusable without
installing piglit-ptest.

(From OE-Core rev: ece99c85c27a546ea59a78875f630e82270c3177)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

scripts/contrib/build-perf-test: Update eglibc -> glibc

(From OE-Core rev: 928fc6c3652ba5db87e1f1c7923972d85c31bc08)

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

xserver-nodm-init: Only start in runlevel 5

Runlevel 3 is often multi-user without X, so don't start the X
server in runlevel 3. This allows one to start up a machine without
X running from the bootloader.

(From OE-Core rev: 116b8a48f0d701d8f0b7807144ffdb708aad215e)

Signed-off-by: Dan McGregor <dan.mcgregor@usask.ca>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

rpcbind: Make user's home directory /

(From OE-Core rev: fbd068e23e16746d89c0bb60c96edd705da27a35)

Signed-off-by: Dan McGregor <dan.mcgregor@usask.ca>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

classes/compress_doc.bbclass: compress man/info pages

A distribution compression policy should be established, and used by all
packages. It compressed man pages in ${mandir} and info pages in ${infodir}

1. The doc will be compressed to gz format by default, which is configured
in ${DOC_COMPRESS}

2. It will automatically correct the compressed doc which is not
in ${DOC_COMPRESS} but in ${DOC_COMPRESS_LIST} to the format
of ${DOC_COMPRESS} policy

3. It is easy to add a new type compression by editing
local.conf, such as:
...
DOC_COMPRESS_LIST_append = ' abc'
DOC_COMPRESS = 'abc'
DOC_COMPRESS_CMD[abc] = 'abc compress cmd ***'
DOC_DECOMPRESS_CMD[abc] = 'abc decompress cmd ***'
...

[YOCTO #1238]

(From OE-Core rev: 5c833e188bbf25c35c24f78eaa761a191d6b5801)

Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

perf.bb: fix multilib build

Fixed:
$ bitbake perf lib32-perf
ERROR: QA Issue: lib32-perf: Files/directories were installed but not shipped
  /usr/lib64
  /usr/lib64/traceevent
  /usr/lib64/traceevent/plugins
  /usr/lib64/traceevent/plugins/plugin_function.so
  /usr/lib64/traceevent/plugins/plugin_scsi.so
  /usr/lib64/traceevent/plugins/plugin_hrtimer.so
  /usr/lib64/traceevent/plugins/plugin_kmem.so
  /usr/lib64/traceevent/plugins/plugin_jbd2.so
  /usr/lib64/traceevent/plugins/plugin_xen.so
  /usr/lib64/traceevent/plugins/plugin_mac80211.so
  /usr/lib64/traceevent/plugins/plugin_cfg80211.so
  /usr/lib64/traceevent/plugins/plugin_sched_switch.so
  /usr/lib64/traceevent/plugins/plugin_kvm.so [installed-vs-shipped]

The perf.do_configure edits kernel's
${STAGING_KERNEL_DIR}/tools/perf/config/Makefile, there would be
problems since kernel doesn't have multilib, and the build result is
undetermined.

Previously, the sed command changed libdir to /usr/lib64 (or 32) in the
Makefile, so the build result was different if we build perf (64) first
or lib32-perf first.

Use the weak assignment "libdir ?=" to instead of "libdir =" will fix
the problem since the multilib builds are in different processes, and
they won't affect each other any more.

The sed command will match both $(prefix)/$(lib) and $(prefix)/lib since
the Makefile may has been modified before this patch.

(From OE-Core rev: c62d693e1341be4afcaaeb1ee37360ecac71f46a)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pango / ptest: clean CFLAGS for host binary

The binary gen-all-unicode needs to be compiled for the host
architecture, the CFLAGS passed to target system could cause
build issues for it.

(From OE-Core rev: b797cfbe605ab250a5eb714a7d5175861690ae6e)

Signed-off-by: Wenzong Fan <wenzong.fan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

scripts/create-recipe: minor fixes

Fixed:
- Remove PR. We don't need the "PR = r0" in the bb file.
- Function length can't get the array's size correctly.
  Should use function scalar.

(From OE-Core rev: efb9f9f1d9a85a22991f99277f77b87784762cbb)

Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_manager: Add rpm v4 support

Currently the package manager code only supports rpm v5. To be useful outside
of OE or with OE layers using v4, it makes sense to add in rpm v4 support. This
takes a patch from "Bartosh, Eduard" <eduard.bartosh@intel.com> and enhances
it to also include versions of the workarounds from poky-eurogiciel to
allow rpm v4 usage with the class for image construction.

(From OE-Core rev: fe21804c296bbb8b2b8b0c29e6e4890bc17f07fc)

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

package_rpm: Add builddir macro to define source directory

This allows its usage in other RPM macros so files in ${S} can be found.

(From OE-Core rev: 649ddd4b17e8a1f94ca9f4bc3a71560437c61909)

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

package_rpm: Add custom extension support for spec generation.

Add hooks to allow customisation of the rpm spec files. Since python functions
aren't visible in the data store, one variable is used to trigger the call to
a separately named function. A dummy function is not provided since this then
triggers various class ordering complexities which are best avoided.

Ultimately this will be replaced by a refactor of the code to generate the
spec file using a python class.

This allows the tizen layer to add hooks for the security manifests for
example.

(From OE-Core rev: 03ac91815013c0e85c4694b3ab849257e658aeba)

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