bb.fetch2.bzr: add bzr urldata_init

move the bzr specific urldata init from localpath to urldata_init
so that it can be called early

Signed-off-by: Yu Ke <ke.yu@intel.com>

bb.fetch2.hg: add hg urldata_init

move the hg specific urldata init from localpath to urldata_init
so that it can be called early

Signed-off-by: Yu Ke <ke.yu@intel.com>

bb.fetch2.svn.py: add urldata_init

move the svn specific urldata init from localpath to urldata_init
so that it can be called early

Signed-off-by: Yu Ke <ke.yu@intel.com>

bb.fetch2.git.py: add git urldata_init

move the git specific urldata init from localpath to urldata_init
so that it can be called early

Signed-off-by: Yu Ke <ke.yu@intel.com>

bb.fetch2: add urldata_init call for Fetch class

FetchData has some fetch method specific data, and only fetch method knows how
to initialize it. originally it is mostly initialized in Fetch.localpath().
But now there is requirement to call Fetch.latest_revision() before
Fetch.localpath(), thus require another earlier place for initialization. so
urldata_init is introduced for this purpose. it will be called in FetchData:__init__
and make all the Fetch functions useable after that.

Signed-off-by: Yu Ke <ke.yu@intel.com>

BBHandler: remove bb.fetch referrence

BBHandler.py no longer use bb.fetch, so remove its import statement

Signed-off-by: Yu Ke <ke.yu@intel.com>

machines: move PACKAGE_EXTRA_ARCHS to tune files

This gets us closer to making including tune-<arch>.inc "just work".
Moving the TARGET_ARCH definitions is something for a follup patch.

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: For poky only, force the use of fetch2 codebase

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

bitbake: Switch to fetch2 fetcher code based on the environment variable BBFETCH2

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

bb.fetch2: replace bb.fetch with bb.fetch2 in the bb.fetch

bb.fetch2 is copied from bb.fetch, and has many bb.fetch referrence.
Fix these referrence with bb.fetch2 referrence

Signed-off-by: Yu Ke <ke.yu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake/fetch/git: Add backwards compatibility code for branch name handling

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

bitbake: copy bb.fetch to bb.fetch2 as initial code base for fetcher overhaul

Signed-off-by: Yu Ke <ke.yu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake/poky.conf: Add TARGET_LINK_HASH_STYLE option to TARGET_LDFLAGS as per OE.dev

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

bitbake/codeparser: Fix import to match upstream bitbake

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

utils: show the actual exception in better_exec

(Bitbake rev: a148e6a63c842ac586ac1dddbd9008f93cdea297)

Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

runqueue.py: Drop unneeded import and improve quieterrors handling as per upstream bitbake

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

bitbake: Update event queue handling to match upstream bitbake

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

bitbake/runqueue.py: Sync with changes in upstream bitbake

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

bitbake/fetch: Sync up logger/whitespace improvements

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

bitbake/utils.py: Drop unused function extend_deps()

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

bitbake: Sync environment handling function names and comments with upstream alterations

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

parse: pass filename, lineno into the ast

We will be needing this information to improve the tracebacks of python code
from the metadata, as well as to give the user information about where
variables were defined, so they know how it ended up the way it is.

(Bitbake rev: 9615c538b894f71a2d1a0ba6b3f260db91e75786)

Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cache: don't expand variables for skipped recipes

Errors can result from these expansions, but for skipped recipes, we
shouldn't care about those failures.  This fixes the same issue which
Richard Purdie fixed in poky, commit 847b717.

(Bitbake rev: 96ee6840010c1ae1080e6bf7ff0f4eb2d361e84b)

Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake/knotty.py: Don't show log messages of NOTE and below for task processes

This cleans up the knotty console messages to be a lot quieter and cleaning,
in keeping with the expectations of most users.

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

bitbake/__init__.py: Add taskpid to all LogRecords (subclassed to be BBLogRecords)

This allows us to identify which task messages are from.

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

bitbake/msg.py: Add named definitions for all logging levels

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

xserver-xf86-common: Remove COMPATIBLE_HOST as it was wrong solution

Signed-off-by: Saul Wold <sgw@linux.intel.com>

insane.bbclass: Enable the package_qa_hash_style check

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

bitbake/runqueue.py: Remove duplicate TaskStarted event fire()

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

xvinfo: Fix LIC_FILES_CHKSUM

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

libcanberra: Fix LIC_FILES_CHKSUM

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

bitbake/build.py: Fix del_stamp work correctly after recent stamp function changes

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

bitbake/cooker.py: Fix merge conflict with upstream which was preventing clean bitbake shutdown

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

bitbake/cache.py: Expansion of variables in skipped recipes should not be attempted as they can trigger errors

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

bitbake/data.py: corrected the output for shell syntax.

[BUGID# 645], modify the emit_var()
1. Added "#" to the beginning of each line if the comment contains
multiple lines.

2. Added "\" to the end of each line if the shell variable value
contains multiple lines.

Signed-off-by: Lianhao Lu <lianhao.lu@intel.com>

bison-native: create a wrapper script for sstate installation

bison-native encodes M4 and PKGDATA staging path in its binary, which
breaks sstate installation in a new build environment. Use create_wrapper
similar as flex-native to solve this issue

Signed-off-by: Kevin Tian <kevin.tian@intel.com>

flex-native: create a wrapper script for sstate installation

flex-native encodes M4 staging path in its binary, which breaks sstate
installation in a new build environment. Use create_wrapper to create
a wrapper script which explicitly set M4 environmental variable to the
new path

Signed-off-by: Kevin Tian <kevin.tian@intel.com>

perl: use ${CCLD} instead of its own LD definition

perl explicitly assigns LD to a bogus value:
    oe_runmake perl LD="${TARGET_SYS}-gcc"
which breaks sstate-based build when toolchain is not generated locally,
due to lacking of --sysroot option. Use ${CCLD} instead to make this
relocatable.

Signed-off-by: Kevin Tian <kevin.tian@intel.com>

bitbake/siggen.py: Add a signature generator which adds the hash to the stamp files

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

base.bbclass: Use the new stampfile function in bitbake to determine the path to the stampfile

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

bitbake/runqueue.py: Fix a bug where do_setscene dependencies would be ignored

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

bitbake build/siggen/runqueue: Fix stampfile parameters

The current parameters are not useful to the stampfile generator function
as they can't uniquely define a task. This updated things so the
parameters can identify unique tasks.

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

bitbake build.py: Stamp handling improvements

* Move stamp file deletion out of the internal stamp helper function
* Add a new function to return the path to a stamp for a given task

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

bitbake build.py: Use localdata for stamp handling, not d

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

poky-ref-manual: Fix my email address

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

bitbake/build.py: Ditch the extract_stamp function as its just going to complicate improving the stamp handling functions

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

parse.ast: drop __word__ regular expression

We can use the string split method for this instead.

(Bitbake rev: aa9646717b3ee1006628246a7c495f601e62391c)

Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

parse.ast: avoid code duplication for inherit

(Bitbake rev: 0b11a3d4eab84b372fd45b9537cf0327008daf8d)

Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

event: use BBLogFormatter in print_ui_queue

This ensures that when a failure occurs very early on in bitbake startup, the
message formatting ematches that used by the UIs.

(Bitbake rev: c8ff0fd3e9f050a668f1a069cf37ee37db3664fa)

Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

build: use logger.isEnabledFor, not getEffectiveLevel

I think this is less confusing, and avoids needing to know about the *range*
of logging levels, instead simply asking what we really want to know.

(Bitbake rev: dc2264387617586b5c0a61e126c75edde5e7abcd)

Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

recpies: use SRCPV instead of SRCREV for PV

SRCPV is intended being used by PV, some recipes still use
SRCREV for PV, which is not correct. This patch fix all the
misusage.

Signed-off-by: Yu Ke <ke.yu@intel.com>

qemuimagetest: Remove connman test for lsb image and fix one warning in Test_Create_Qemu

There is no connman in LSB image. So we need to remove connman test from it.
And when we check if ip address fetched by Test_Fetch_Target_IP is valid,
we should use "==" operator for string comparison.

Signed-off-by Jiajun Xu <jiajun.xu@intel.com>

insane.bbclass: Fix message for LIC_FILES_CHKSUM Errors

Signed-off-by: Saul Wold <sgw@linux.intel.com>

python-dbus: upgrade from 0.83.1 to 0.83.2

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>

byacc: upgrade recipe from 20100610 to 20101127

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>

git: upgrade from 1.7.3.2 to 1.7.3.4

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>

distro-tracking: update recently updated recipes

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>

modutils: Extend gcc4 patch for obj_ppc.c

Signed-off-by: Saul Wold <sgw@linux.intel.com>

xserver: Add COMPATIBLE_MACHINE

Signed-off-by: Saul Wold <sgw@linux.intel.com>

rpm: Bump PR to force rebuild due to new beecrypt

Signed-off-by: Saul Wold <sgw@linux.intel.com>

distro_tracking_fields.inc: update apmd, blktool, dbus-glib, libgdbus, and pax-utils

Signed-off-by: Scott Garman <scott.a.garman@intel.com>

dbus-glib: upgrade to version 0.92

Includes some minor recipe formatting changes.

Signed-off-by: Scott Garman <scott.a.garman@intel.com>

dbus-glib: delete unused patch

Signed-off-by: Scott Garman <scott.a.garman@intel.com>

libgdbus_git: metadata fixes and minor formatting changes

* Added DESCRIPTION, SECTION, PRIORITY, and PR fields
* Updated SUMMARY and HOMEPAGE fields to reflect more accurate info

Signed-off-by: Scott Garman <scott.a.garman@intel.com>

blktool: minor formatting changes

Signed-off-by: Scott Garman <scott.a.garman@intel.com>

apmd: added HOMEPAGE, minor formatting changes

Signed-off-by: Scott Garman <scott.a.garman@intel.com>

apmd: delete unused patch

Signed-off-by: Scott Garman <scott.a.garman@intel.com>

pax-utils: upgrade to version 2.2.2

* Upgrade to latest upstream release
* Minor cosmetic recipe changes

Signed-off-by: Scott Garman <scott.a.garman@intel.com>

binutils: use default ${CC} instead of defining it own

binutils.inc exports its own ${CC} which is similar to default ${CC}
but missing ${TOOLCHAIN_OPTION}, which makes --sysroot option lost
when compiling target binutils which then further breaks sstate.

Signed-off-by: Kevin Tian <kevin.tian@intel.com>

kernelshark: add ftrace gui viewer

The trace-cmd source includes a graphical trace viewer, but we
don't want to make the trace-cmd recipe depend on gtk+. This
patch adds a second recipe in the meta/recipes-kernel/trace-cmd
directory which uses the trace-cmd SRCREV and RDEPENDS on trace-cmd
to ensure compatibility.

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Tom Zanussi <tom.zanussi@intel.com>
CC: Richard Purdie <richard.purdie@linuxfoundation.org>
CC: Saul Wold <sgw@linux.intel.com>

glib-2.0: fix g_once_init_enter compile failure

Kernelshark trips over a compile bug in glib-2.0 addressed by Open Embedded
back in the 2.22 timeframe. It appears to still be present in 2.27.

http://www.mail-archive.com/openembedded-devel@lists.openembedded.org/msg00529.html

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Tom Zanussi <tom.zanussi@intel.com>

trace-cmd: license file updates

Signed-off-by: Darren Hart <dvhart@linux.intel.com>

trace-cmd: fix cross-compilation

The trace-cmd Makefile forces certain variables, such as CC and AR.
It was using the host gcc and loader, fix it to use the poky-built
cross-compiler and linker.

inherit pkgconfig to ensure we don't use the host pkg-config.

Remove unecessary variables from the oemake commands, such as ARCH
and LD which aren't used by the Makefile.

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Tom Zanussi <tom.zanussi@intel.com>

distro_tracking_fields.inc: update for lastest package updates

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

preferred-xorg-versions.inc: updated for lastest package updates

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

librsvg: upgrade from 2.32.0 to 2.32.1

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

libxxf86misc: upgrade from 1.0.2 to 1.0.3

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

libdmx: upgrade from 1.1.0 to 1.1.1

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

pixman: upgrade from 0.20.0 to 0.21.2

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

scrnsaverproto: upgrade from 1.2.0 to 1.2.1

Add "--enable-specs=no" to disable doc generation -- I doubt we really need
it, and it doesn't work well due to

SEVERE: Error with opening URL 'http://docbook.sourceforge.net/release/images/draft.png'.
Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

resourceproto: upgrade from 1.1.0 to 1.1.1

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

compositeproto: upgrade from 0.4 to 0.4.2

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

xproto: upgrade from 7.0.19 to 7.0.20

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

inputproto: upgrade from 2.0 to 2.0.1

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

renderproto: upgrade from 0.11 to the latest version 0.11.1

Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>

matchbox-wm-2: Upgrade to svn r2106

Signed-off-by: Zhai Edwin <edwin.zhai@intel.com>

metacity: Upgrade to 2.30.3

Revised one patch from openembed to make host native binary for build.

Signed-off-by: Zhai Edwin <edwin.zhai@intel.com>

libcanberra: Upgrade to 0.26

Remove nofallbackfix.patch it as already in upstream. Remove
autoconf_version.patch as already have new version of autoconf and no inbuild
libltdl now.

Signed-off-by: Zhai Edwin <edwin.zhai@intel.com>

xvinfo: Upgrade to 1.1.1

Signed-off-by: Zhai Edwin <edwin.zhai@intel.com>

beecrypt: Update to 4.2.1

Rebased disable-icu-check.patch & fix-security.patch. Removed x64fix.patch as
it is already in upstream. Also some configure changes: removing "--with-arch"
as it can't cowork with expert mode, adding "--disable-openmp" to avoid
unnecessary dependency.

Signed-off-by: Zhai Edwin <edwin.zhai@intel.com>

rootfs_rpm: fix rootfs generation rm error

During a poky-image-minimal build I ran into the following error:
rm: cannot remove `/vol/1/dvhart/poky.git/build/tmp/work/atom-pc-poky-linux/poky-image-minimal-initramfs-1.0-r0/rootfs/install': Is a directory

This is caused by a missing -r (recursive) argument to rm. This patch
adds that to the rm command of rootfs_rpm.bbclass.

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Bruce Ashfield <bruce.ashfield@windriver.com>

rootfs_rpm.bbclass: Remove temporary work files in install directory, saving the manifests as logfiles into ${T}

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

bitbake: build.py: Logging of None was losing the logs so force this to stdout so we can see function execution failures

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

scripts/bitbake: Skip building pseudo if necessary.

Fixed [BUGID# 625]. Skip building pseudo if necessary by parsing the
options to bitbake.

Signed-off-by: Lianhao Lu <lianhao.lu@intel.com>

usbutils: Don't install the uncompressed version of usb.ids, the compressed version will suffice

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

pciutils: Don't install the uncompressed version of pci.ids, the compressed version will suffice

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

bitbake: Misc syncups with bitbake upstream, mostly whitespace and a missed line from the last UI commit

Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>

Resurrect alternative UIs

The various alternative UIs have been updated to once again be functional
with the latest bitbake internals.  Each of the UIs still have much room for
functional improvement.

In particular, they have been updated to:
 - interact with the new process based server
 - handle the current set of events and notifications fired from the server
   and its associated subsystems

(Bitbake rev: b947e7aa405966262c0614cae02e7978ec637095)

Signed-off-by: Bob Foerster <robert@erafx.com>
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>

bitbake: build.py logging updates

python tasks calling shell functions using exec_func() would show the log
file as /dev/null. It makes most sense for all the task logging to be setup
centrally by exec_task(), at least with the current code base in Poky.

This commit will need discussion in relation to upstream bitbake and the
IO redirection could be better handled using a context manager (although
task contexts shouldn't ever nest).

Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>

bitbake: Fix process.py to accept None as the log option

Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>