bitbake: Fix bad merge of #2162

This patch fixes the bad merge of #2162 fixes on master.

(Bitbake rev: 5b84d88f2a47063197f9a20f8ebf0a7ccf22c2eb)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob: Error reports are done in a clearer way

For long errors (bigger than 200 letters),
the text box is scrollable and resizable
and text is selectable.

Additionaly, all message dialogs are modal.
Otherwise, a user could still interact with hob
even in an error case, leading to potential problems.

See design details in related bugs.

Fixes [YOCTO #2960], [YOCTO #2983]

(Bitbake rev: be8bf02f2b347edf5514cafc6cb6a44f71118736)

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

bitbake: hob/builddetailspage: fix failure_primary_action_button_clicked_cb

[YOCTO #3194]

(Bitbake rev: 660a449bdfd154fed556024f166e69c6931ff634)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob/builddetailspage: remove "back" button

When a build fails, there should not be a back button on the screen.
All available actions are provided within the failure notification,
so no back button is needed.

[YOCTO #3104]
(Bitbake rev: 03f978d21c7bfbf5f1afc741a43766030f2882a8)

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

yocto-bsp: fix dmaengine feature inclusion

The cfg/dmaengine/dmaengine feature changed location to cfg/dmaengine
in the 3.4 yocto kernel's meta branch.  Add template code to include
the appropriate version.

(From meta-yocto rev: b650fcb7781e1c6af6254c98ae64d5ea81b46abc)

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

yocto-bsp: make vfat feature inclusion conditional on kernel version

The linux-yocto-3.2 cfg/vfat feature changed location to cfg/fs/vfat
in the 3.4 yocto kernel's meta branch.  Add template code to include
the appropriate version depending on kernel version.

Fixes [YOCTO #3178].

(From meta-yocto rev: d574c56c51789ec56ff50518ac2057607740eaa8)

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

xserver-xorg: Remove RCONFLICTS against xserver-xorg

When merging the xserver-xorg fix the to use RDEPENDS in
xserver-xorg-module-exa the RCONFLICTS has not been removed by
mistake. This drops the RCONFLICTS to properly fix it.

(From OE-Core rev: d83e218dc480a09befddf8b934d774519cdbacb5)

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

siteconfig: Clear cache before rebuilding

This ensures consistent build results and avoids build failures when compiler flags
change for example.

(From OE-Core rev: a5ff8396cad130f809f8f8da49bb38e6f80f923c)

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

perl: Fix substitution madness

We're going around in circles trying to fix the sed expressions and making one case
work and others not work. This patch fixes the base configuration file so we have
non-overlapping substitutions. I've tried to significantly clean up various problems
that were occurring once and for all.

This will hopefully resolve all the issues people have been seeing with incorrect perl
paths.

(From OE-Core rev: 31ff70794ecc431431476f81c8934fff25383613)

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

flac_1.2.1.bb: use TUNE_FEATURES to enable/disable altivec

(From OE-Core rev: bbcc66dc2d69821adb5b39b3642c368fb74ad4d7)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

tune-ppce6500.inc: add e6500 tune files

Also supports a new altivec TUNE_FEATURE

(From OE-Core rev: 4586c24ad156773568cd38794936b8af62e862be)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

arch-powerpc.inc: add altivec as a valid tune feature

(From OE-Core rev: 026f8bc59b6c4cc23cc8a706117bf5b3555f2c7a)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

eglibc_2.16.bb: refresh fsl-ppc-no-fsqrt.patch for Freescale targets

(From OE-Core rev: 5635cf21520182e12c8a130707f8b47b5b4bec00)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

eglibc_2.16.bb: refresh ppc_slow_ieee754_sqrt.patch for Freescale targets

Make same changes for e6500 fpu as done with others

(From OE-Core rev: a39f8c19d0ea5dc92271cbe36a03d638cb806e04)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

eglibc_2.16.bb: refresh ppc-sqrt_finite.patch for Freescale targets

(From OE-Core rev: eba4de86e7e628690232f2f7912b321a9e22701b)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

eglibc_2.16.bb: replace patch with updated version that supports e6500

(From OE-Core rev: a68536b75cf93beaa1578b33d489d9f30b58ba2e)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

boost: Support enums in hash

Fixes builds which were building fine with previous versions of boost

(From OE-Core rev: 139b65affbf50c4230e344cd4d0205ef025bb7c3)

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

flex,bison: Add nativesdk variants

It is needed in some SDKs that we ship own
version of lex/yacc for sdk host

(From OE-Core rev: 536c9e42d316efb42651fdc2eba1b8548d74329d)

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

ltp: add dependency libcap

Similar to libaio, libcap is another dependency of ltp.

If libcap has been done populate_sysroot but rpm/ipk package is not
created, ltp will be compiled with libcap. So when install ltp to a
image, it complains that package libcap is not found.

[Yocto #2973]

(From OE-Core rev: bf5215f095e7e610508fcefe1224c9289c6c56cd)

Signed-off-by: Kang Kai <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: knotty: Allow displaying of status when no tasks are active

The console can appear to hang when no tasks are executing even if bitbake
is iterating through a large number of tasks behind the scenes.

This patch tweaks the footer code to display a status even when no tasks
are active to give the user better feedback about what is happening.

(Bitbake rev: 0a950ee14fce3a0cb938c22d7c717dc93ce6e25f)

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

bitbake: parse/ConfHandler: Add negative lookahead to spot some quoting problems

Syntax like:

FOO = "bar" # eek"

would result in FOO taking the value 'bar" #eek' which is clearly
not the intention. Whilst our metadata is riddled with mixtures of even
quotes like:

FOO = "d.getVar("X")"

odd numbers of quotes seem rare. This patch adds detection of one odd
quote which we don't have any of in OE-Core so it seems a valid sanity
improvement.

(Bitbake rev: 5f892d9b083550e20e37576070ec7d1a94cc88fe)

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

bitbake: hob: set modal flag on progress dialog

The entire interface isn't usable whilst the progress dialog is up so we might
as well set the modal flag so that some WMs (such as GNOME 3) can do nice things
with the dialog (such as pin it to the titlebar).

(Bitbake rev: 9a19fe8e8c65b75dbbb4ae5401df6d6838495bdd)

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

bitbake: hob/settings: implement a new tab in settings dialog to show SSTATE_MIRRORS

Add a new tab to show correctly SSTATE_MIRRORS variable. Now you can add new
mirrors or delete mirror. "info" image was also changed( it is smaller, so it
can be next to labels).
>From "Build environment" tab, SSTATE_DIR and SSTATE_MIRRORS vars were removed.

[YOCTO #2893]

(Bitbake rev: ae27a7cf4d31a1b99840a761a27fd6256cb1dd9b)

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sstate: Relax the duplicate file whitelist for now

do_package is a machine specific task at the moment due to packagedata. This means
do_package tasks and their dependencies rerun between different machines
with various duplicate file installations. There are plans to fix this but they're
too invasive before release.

This patch relaxes the whitelist for sstate duplicate file detection to account
for this. Post-release, we re-enable stricter settings once do_package is not
machine specific.

(From OE-Core rev: c858259ce1881c6284f1fc2790c225c81e4a751e)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

tcl: Fix dangerous do_install staging references

Nothing should ever be poking files directly into the staging/sysroot
directories, it should always go through ${D}.

This patch ensures this recipe does this and hence fixes various
potential build issues such as lack of sstate tracking of files.

(From OE-Core rev: 7642143760c0157aba80cea7b427024ee097cc2c)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xserver-xorg: Use rdepends to ensure xserver-xorg-module-exa match version

This fix the installation of xserver-xorg-module-exa package at rootfs
using opkg. It were failing as conflicts where not working properly.

(From OE-Core rev: 8fb19876215a8c7918361e8360c4342d1a933a93)

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

license.conf/bbclass: Move globals to license.conf

This requires the changes to bitbake.conf that allow parsing of
license.conf.

As we should now be parsing license.conf, we can move some globals
out of license.bblcass and into the conf file.

(From OE-Core rev: 03e6a7cd27ed109a011fac09cf04412f87f31c3a)

Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake.conf: parse license config

license.conf hasn't been being parsed. It probably should be.

(From OE-Core rev: b393b31fee3b4d42890c2bcbba09ea231c131dea)

Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

license.bbclass: Variable standardization

The variable mentioned in license.conf is LICENSE_PATH. The variable
used in license.bbclass is LICENSE_DIR. Conforming to what is in
license.conf

(From OE-Core rev: c6e13d9cd26d016ab06e7447b307d413e1331aa0)

Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sanity: Added explicit network error status in SanityCheckFailed event

If we fail a network test, a special flag is set in the SanityChekFailed
event. This helps Hob identify the network error properly and display
a special message to the user.

[YOCTO #3025]

(From OE-Core rev: 7877c4344db89237bba5f9a03342bfd9a03aebbf)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

initramfs: Make mkdir not fail

This patch make "mkdir foo" not fail if foo already exists.

(From OE-Core rev: 2bf5026933b733701d4d339e01a4f5e4468f368e)

Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

initrd: Spawn an emergency shell when something goes wrong

set -e allows to exit if a command fails. We install a trap and execute
emergency_shell() when either the init script exits or when ctrl-c is
typed (say if we are stuck somewhere and we want to debug it).

(From OE-Core rev: ae5e2bd994e3f60d3803ab56e6ed34d08fbc56f0)

Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sanity.bbclass: bblayers.conf should be updated automatically

- add check_bblayers_conf bitbake function which does the bblayers.conf
v4 -> v5 update if necessary (every layer should make its specific
bblayers.conf upgrades appending to the check_bblayers_conf function)
- we ask the user to re-run bitbake because we can't trigger reparsing
without being invasive

[YOCTO #3082]

(From OE-Core rev: 03ad4edace5db9c6e15ca776d06d20b7d4e42afc)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

runqemu-internal: don't append an empty element to PATH

Bitbake fails to run when an empty element exists in $PATH. Avoid
creating this situation when $CROSSPATH is not set.

This fixes bug [YOCTO #3101]

(From OE-Core rev: 1f7f590369eaa76dc970c9cffd1f0db53ce08c00)

Signed-off-by: Scott Garman <scott.a.garman@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

oe-find-native-sysroot: show bitbake errors to user

Ran into another bug that was masked by hiding a bitbake error message.
This catches this situation and displays the error to the user.

Also includes whitespace fixes.

(From OE-Core rev: 435ffeefe4a1df53335fd397ff404bed7deae2df)

Signed-off-by: Scott Garman <scott.a.garman@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gst-plugins-good: make pulseaudio support dependent on DISTRO_FEATURES

This should be no change to the previous situation unless you
explicitly have pulseaudio in DISTRO_FEATURES_BACKFILL_CONSIDERED
(currently).

(From OE-Core rev: 3d28d16042373e699fabd8576458f65d63463bb9)

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

xserver-xorg: merge version-specific .inc into .bb

The version-specific .inc was shared with the xserver-xorg-lite package, but
that doesn't exist anymore.

(From OE-Core rev: 09b1bf350384722127ac9f098a72371cf27c3822)

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

eglibc: Fix for dynamic linker broken offset

Solution provided by Donn Seeley in bug 1443:
https://bugzilla.yoctoproject.org/show_bug.cgi?id=1443

worked when testing with core-image-sato-sdk for qemuarm.

[YOCTO #2577]

(From OE-Core rev: 33ec4222c05c985b737e88850259218cf8336d46)

Signed-off-by: Andrei Dinu <andrei.adrianx.dinu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cogl: Add a missing depends on libxdamage

During a from scratch build test, cogl build failed due a missing
dependency on libxdamage.

(From OE-Core rev: 959a2f6d88d8fa6874fff83b7a1f0e7d4e36b887)

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

bitbake: Add missing file from previous commit

(Bitbake rev: 5bc0f0e70f7272ff84aaeca43dcf4bb4bc0f5c3f)

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

Revert "autotools.bbclass: using relative paths for acpaths"

This reverts commit aa66ef6598c84231577d139ec7be413e73fac2b1 since
bdwgc-native fails to build after it. Anything which runs with a
sub-configure will fail after this change. It therefore needs
rethinking.

(From OE-Core rev: f95a9e2c292a1551861220270838cf1eaaba85b9)

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

bitbake: hob/builder: When you stop a build, Hob should tell you stopping is happening

- use the progress bar text to indicate the stopping status
- the text should say: 'Stopping the build...'

[YOCTO #3152]

(Bitbake rev: 6f59db920ca4f527606670969c79afbf34eaff81)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: lib/bb/data.py: improve output for expansion errors

Instead of logging the function/variable separately as a NOTE when
failing to expand, re-raise ExpansionError with more contextual
information. This means that the full details are reported in Hob as
well as actually reporting the original error message in any UI where
we previously did not. For example, we used to get this with tab/space
indentation issues in a python function:

NOTE: Error expanding variable populate_packages
ERROR: Unable to parse /path/to/recipename.bb

Now, we will get this:

ERROR: ExpansionError during parsing /path/to/recipename.bb: Failure
 expanding variable populate_packages: IndentationError: unindent does
 not match any outer indentation level (<string>, line 4)

Fixes [YOCTO #3162].

(Bitbake rev: ce5c7a95a359cdaecab7c4a519ad4f9df029da82)

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

bitbake: cooker: adjust layer dependency error messages

Make these a little easier to understand.

(Bitbake rev: 84ab874c8818484d37ee438aab27486fff497705)

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

bitbake: hob/recipeselectionpage: recipes should not be shown brought in by themselves

[YOCTO #3107]
(Bitbake rev: 2cc5f517224cee8e2dd2b045a277423ce66ec886)

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: fetch2: add "-d" option to cpio

Add "-d" option to cpio since it is useful:
  -d
  --make-directories
  Create leading directories where needed.

[YOCTO #3137]

(Bitbake rev: a78f9ded7896432b107f34c0bb608b389fdb676a)

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

bitbake: Multiple user interface fixes in settings

1. Move the "Others" tab from advanced settings to simple settings
2. Organize content of tabs into headings
3. Match various labels in the settings with the design
4. Clean up config_md5 in both simple and advanced settings

This patch implements a part of the changes requested by Belen in the settings dialogs.
The second version of the patch addresses all the UI changes requested by Belen (more
details are in the bug description): alphabetical ordering of the image types and
warnings if no image type is selected.

[YOCTO #2162]

(Bitbake rev: b45438555ecf2e25ebb99324a18d31c812a2738a)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob/runningbuild: Add missing logging level argument

[YOCTO #3170]

(Bitbake rev: d11fe38a38adc10eedec3b4c251c575b95339774)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: Add sanity check progress screen

This patch adds a sanity check progress screen to hob. The screen
is displayed when Hob executes the sanity check procedure. The screen
is displayed for at least 5 seconds. If a network error is detected,
a special dialog is displayed which lets the user open the proxy
configuration page directly.
Note that currently bitbake triggers the network tests only when
the value of its TMPDIR variable changes, which happens fairly rare
on my system. This is the subject of another bug (#3026).
Version 2 of the patch splits the changes in two parts (sanity.bbclass
belongs to oe-core).

[YOCTO #3025]

(Bitbake rev: b48f1351271cc066ffe919db112b14834a6d8f8f)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob/builddetailspage: change branch field

When a user makes a build from a tarball, it shows fatal error in branch field.
Now it not complains as a fatal error. It is a normal message.

[YOCTO #3114]

(Bitbake rev: 53d5d542cd0197ca67c5f90a57808af2f19ff56d)

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob: Buttons width and height are taken for host

- removed all set_size_request calls for buttons in order to:
- force autosizing of buttons with regard to the text length
- use host theme default height for buttons
- modified buttons on image details page to have the same height (default host one) and the width of the button with the largest text
- modified Stop button on build details page to have the default height by directly attaching it to the containing table instead of hbox

(Bitbake rev: 9cdfaa17309d368c3bbae0f1cce0ad875d340e83)

Signed-off-by: Ioana Grigoropol <ioanax.grigoropol@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob/settings: implement a new tab in settings dialog to show SSTATE_MIRRORS

Add a new tab to show correctly SSTATE_MIRRORS variable. Now you can add new
mirrors or delete mirror. "info" image was also changed( it is smaller, so it
can be next to labels).
>From "Build environment" tab, SSTATE_DIR and SSTATE_MIRRORS vars were removed.

[YOCTO #2893]

(Bitbake rev: 1a81e27365d969e4ad4b4f0aec290aa967a8a35f)

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sstate-cache-management: hide error message when one of possible layer location doesn't exist

* fixes [YOCTO #3116]

(From OE-Core rev: b02d334e0e6a19a1bf3550add68f5770a835c772)

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gnome-keyring: include unpackaged files with PAM enabled

Include missing files/dirs:
    ${base_libdir}/security/*.la
    ${base_libdir}/security/*${SOLIBSDEV}
    ${base_libdir}/security/.debug/

[YOCTO #2805]

(From OE-Core rev: c5b6e3be0923b0af16f4854264fae104470bacf7)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

autotools.bbclass: using relative paths for acpaths

Fix autotools.bbclass to use relative paths for acpaths instead of
absolute ones. Since absolute paths may cause potential autoreconf
error like:

    Can't exec "/bin/sh": Argument list too long ...

This error occurs while building coreutils with long TMPDIR, because
it has bunch of m4 files need to be expanded.

[YOCTO #2766]

(From OE-Core rev: aa66ef6598c84231577d139ec7be413e73fac2b1)

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

perl: Fix nativesdk install path

Commit 38234f2e276356b1d77a87ceabc486107e336d19 tried to fix the sed
expressions by anchoring the left side of the search regexp to prevent
$prefix$prefix type expression in the perl config. For nativesdk this is
not enough. Adding anchors on both side fixes this.

(From OE-Core rev: bf84ec0fb9a4d01ea75447c2efe8e534ce975b53)

Signed-off-by: Christian Glindkamp <christian.glindkamp@taskit.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

dbus: disable all X11 when native

Without --without-x the X11 detection would still go ahead and find the host X11
headers, which seems to cause problems at link time.

(From OE-Core rev: d35d19b6d0844daf8ca8aa059c0aa6077c2f573a)

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

eglibc: Remove bogus PACKAGES_DYNAMIC setting

It transpires that eglibc has been setting PACKAGES_DYNAMIC = "libc6*" for
some time.  However, this is bogus for at least two reasons:

1. Bitbake interprets PACKAGES_DYNAMIC as a regex, not a glob, so this will
match against any package whose name starts "libc" plus zero or more sixes.
This is particularly toxic because the nativesdk variant picks up the same
value and will, consequently, start trying to build itself at the slightest
excuse.

2. eglibc doesn't actually build any packages named "libc6<anything>", other
than the ones that are named in PACKAGES anyway, so the dynamic provider
declaration is in any case useless.

Simply deleting the line is not sufficient since then we get the default
value from bitbake.conf which causes eglibc.bb to fight with eglibc-locale.bb.
So instead we must set it to the empty string for good results.

(From OE-Core rev: 0fbb2e0c1889ee34d7f96266615e891bb44b1d10)

Signed-off-by: Phil Blundell <pb@pbcl.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xserver-xorg: use EXTENDPKGV instead of PV in RCONFLICTS_${PN}-module-exa

* it doesn't make much sense with PV, because xserver-xorg-module-exa
  was introduced in
  http://git.openembedded.org/openembedded-core/commit/meta/recipes-graphics/xorg-xserver?id=1a666ee1cda3c0b74daba5881fc5f62e13deec66
  so our xserver-xorg-module-exa RCONFLICTS with xserver-xorg (<= 1.11.2-r4)
  and (< 1.11.2) is not good enough

* because we don't know how many PRINC are in BSP/DISTRO layers,
  then it's safer to RCONFLICTS with every older version then current
  EXTENDPKGV

Also fixes whitespace to work correctly with opkg

(From OE-Core rev: ed0216d29fc4355c5220f3ad51df04a63cacb0c3)

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
--

* I haven't really tested this with IPK, since it was changed from
  RREPLACES to RCONFLICTS (because of RPM) and all my installed devices
  are already upgraded
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cml1.bbclass: Convert tab indentation in python functions into four-space

Based on the previous commit 604d46c686d06d62d5a07b9c7f4fa170f99307d8
(Convert tab indentation in python functions into four-space), the
cml.bbclass was not converted, and in order to properly extend it
with external bbappend's it needs to be converted.

(From OE-Core rev: e4c1c37bb37e9eba635bc0a9308ab593abd299ec)

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

busybox: Fix misplaced quote

(From OE-Core rev: 938d07871bedd91f0d95ed6fe338ecbfafa5ebfe)

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

util-linux: Remove static libraries from -dev packages

Fixes the QA warnings:

WARNING: QA Issue: non -staticdev package contains static .a library: util-linux-libblkid-dev path '/work/mips32el-oe-linux/util-linux/2.21.2-r3micro3/packages-split/util-linux-libblkid-dev/lib/libblkid.a'
WARNING: QA Issue: non -staticdev package contains static .a library: util-linux-libuuid-dev path '/work/mips32el-oe-linux/util-linux/2.21.2-r3micro3/packages-split/util-linux-libuuid-dev/lib/libuuid.a'

(From OE-Core rev: 8b123ca10904386ec9d860a2908c713b1c6a73e8)

Signed-off-by: Phil Blundell <pb@pbcl.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

util-linux: Don't package chkdupexe

The chkdupexe utility is fairly worthless and drags perl in as a build dependency
of the whole util-linux recipe.  If anybody actually wants to use this script
then we should package it separately, but for the time being let's just delete it.

(From OE-Core rev: 19dd830ff8a1b87499b9a51599265dd436214708)

Signed-off-by: Phil Blundell <pb@pbcl.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_rpm.bbclass: change the arch's "-" to "_" for platform

The platform and platform_extra will be written to /etc/rpm/platform,
the rpm's arch has changed the "-" to "_", so the value in platform
should also be updated.

[YOCTO #3159]

(From OE-Core rev: 9880a5261ca509c69efbafa27cddd9b2b8ca08f0)

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

package_rpm.bbclass: no initial_solution in the second build

There is no initial_solution.manifest in the second build when
incremental rpm image generation, since the initial solution has been
skipped. So we should check it before cat it.

[YOCTO #3128]

(From OE-Core rev: ad17fa82a481ab3c9f17a8338ebad1eb07c0f9d8)

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

layer.conf: allow other layers to depend on this version

Set LAYERVERSION and rename the collection to "core". Given changes such
as the tabs to spaces cleanup for python functions in the current
version, this allows other layers to depend on this version of OE-Core
specifically should they choose to do so, by specifying the following
in their own layer.conf:

LAYERDEPENDS_layername = "core:1"

Where layername is whatever value is being added to BBFILE_COLLECTIONS.

(This change does nothing unless a layer has LAYERDEPENDS set.)

(From OE-Core rev: 7069c7ef829f56ae6dd0dab5e583342f351274ed)

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

iproute2: Use configured values for ${base_sbindir} and ${libdir}

These were previously being hard-coded to "/sbin" and "/usr/lib" respectively,
resulting in unpackaged files if the configured values were something else.

(From OE-Core rev: fab44bdb346533edd4f702a83523d7c2414f74e5)

Signed-off-by: Phil Blundell <pb@pbcl.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

SDK: trap any IO errors in the relocate script

If the files being relocated are already used by other processes the
relocate script will fail with a traceback. This patch will trap any IO
errors when opening such a file and gracefully report them to the user.

Also change the exit code from 1 to -1 for a better adt-installer user
experience (like pointing the user to the adt_installer.log).

[YOCTO #3164]

(From OE-Core rev: 26daec758b2eaeb208356d5aa8a9a191bd366751)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

ossp-uuid: LICENSE type corrections.

The LICENSE for ossp-uuid is MIT. As well, the LIC_FILES_CHKSUM
was missing the license text within uuid_md5.c

(From OE-Core rev: 99974820c8b12b1c1b05ab8a96f1b25b8e3888da)

Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

at: atd remove useless --make-pidfile option.

For start-stop-daemon, --make-pidfile is used when starting a
program that does not create its own pid file.
atd would create its own /var/run/atd.pid, so remove this option.

(From OE-Core rev: f10d236cda704cd91e185f8dc9c3f52461e2dad1)

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

sysvinit-inittab: fix hang issue of series consoles check

The kernel boot process hangs when /proc/consoles doesn't exists, therefore
check the existence of /proc/consoles before executing pkg_postinst script.

Following is the log when /proc/consoles doesn't exist:
        Running postinst /etc/rpm-postinsts/102...
        cat: /proc/consoles: No such file or directory
        cat: /proc/consoles: No such file or directory
        cat: /proc/consoles: No such file or directory
        INIT: Entering runlevel: 5
        Starting OpenBSD Secure Shell server: sshd
          generating ssh RSA key...
          generating ssh ECDSA key...
          generating ssh DSA key...
        done.
        Starting network benchmark server: netserver.
        Starting system log daemon...0
        Starting kernel log daemon...0
        Stopping Bootlog daemon: bootlogd.
        INIT: no more processes left in this runlevel

(From OE-Core rev: 390e7f1f0b1b21d3c0787a6272583d5829561f95)

Signed-off-by: Zhenhua Luo <b19537@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_ipk: Remove spurious '-i' in grep command for log_check

ipk_log_check uses a case-sensitive grep (which is correct) when deciding
whether there were any errors or not.  But if it decides that there were, it
then uses a case-insensitive grep to display them.  This results in a large
amount of irrelevant and confusing output which makes it hard to see the real
errors amongst the noise.

Suppress this by removing the unwanted -i.

(From OE-Core rev: 57dcacbd6f35ae2d6b505f044bbefad35da66959)

Signed-off-by: Phil Blundell <pb@pbcl.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gcc-configure-cross: factor out --enable-threads argument into ${GCCTHREADS}

This allows BSPs for architectures with no thread support to set (for
example) "GCCTHREADS=no" without having to override all the other configure
parameters.

(From OE-Core rev: 6bb0d37529a82b953d374f2d76c2412d7cee587b)

Signed-off-by: Phil Blundell <pb@pbcl.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sstate: Remove master manifest usage

This was added to allow detection of duplicate files being installed by sstate.
There is a much simpler way, just check if the file already exists. This
effectively uses the kernel VFS as the cache which is much more efficient.

This resolves a significant performance bottleneck (lock contention on a
single file) when running builds that are just being generated from sstate
cache files.

(From OE-Core rev: 603daf343ad3f18c8adb799e3625ae2a18d94f56)

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

packagedata/multilib: Fix search patch for multilib builds

The current multilib search path code for packagedata is flawed since it
doesn't correctly handle changes in the TARGET_VENDOR/TARGET_OS that
multilib may make. This patch enhances the code to correctly build the
search paths so multilib packagedata is found correctly.

(From OE-Core rev: f50c5d36b2da9b36d56d95a7d89404509a1a3e9b)

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

sstate: Fix SSTATE_DUPWHITELIST variable usage

We need to split this variable before using it. Otherwise a single "/"
character in the list whitelists every overlapping sysroot file which
was not the intention making the whole thing useless.

We'll start seeing warnings about overlapping files now this is working
correctly after this patch.

(From OE-Core rev: 9e31c748327e92b809330f4ad7b6aaecb2edf559)

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

bitbake: fetch2/git: Don't use deprecated API

(Bitbake rev: 8e650b3307b60cfe8e7439ea6891c3a85f785af9)

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

bitbake: hob/packageselectionpage:cancel button should redirect to Image conf screen

Cancel button brings you to Image configuration page.

[YOCTO #3105]
(Bitbake rev: 88cd9586f0f6a413c1a6800b3e57444f453afb73)

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob/builddetailspage: change tooltips and remove a dialog from "Build stopped" message

When you stop a build, a "Build stopped" message appears. I have changed 2 tooltips and
also eliminate the alert that comes up when you click 'Edit packages'.

[YOCTO #3160]
(Bitbake rev: f5a21da2faf7ede56cf211b96dffd8aaa4b485b8)

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

kexec-tools: admit mips as a COMPATIBLE_HOST

(From OE-Core rev: 372dc3cf95373225d512160a2ec3e16bf3dc5b8f)

Signed-off-by: Phil Blundell <pb@pbcl.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto/3.4: update beagleboard configs

Updating to pickup a couple of configuration tweaks that were done
as part of a bump to the 3.4 linux-yocto tree for the beagleboard.

In particular, this updates DVI and USB settings to enable the
perhipherals out of the box.

(From OE-Core rev: 3c562b8360edf52aea6280849fa33e8a0f34742b)

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

linux-yocto/3.4: update kver to v3.4.11

Updating the machine and meta SRCREVs to the latest 3.4.x -stable
release.

See git whatchanged v3.4.10..v3.4.11 for the full changelog.

(From OE-Core rev: b164d9776802784a2fee01f2ab267dab4c2beafe)

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

Add and use 'localedir' variable

This avoids the hardcoding of ${libdir}/locale which is all over the place,
and will facilitate use of ${exec_prefix}/lib/locale instead of
${libdir}/locale.

This doesn't actually change any output at this time. Verified this with
buildhistory against the packages produced from core-image-base.

(From OE-Core rev: b744f4cc2912334b8493a89525fd02af8e9b8edf)

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

gcc-common.inc: Consider multilib when renaming libgcc for debian'ness

When doing multilib builds rpm does not find libgcc1 for lib32
multilib because its not honoring the debian renaming scheme for
libgcc-multilib. Lets add MLPREFIX to fix it.

(From OE-Core rev: 9327ca868667b15f29af3123611d6f56b4249a63)

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

rpm: Add base-files as RDEPENDS

This solves a problem when installing rpm using the ipk pkg-management
system where /var/cache was conflicting with the existing /var/cache from
base-files.

(From OE-Core rev: 917f57cbb0906996661eebc6656c2c083ef979e9)

Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libcap: Fix erroneous path in last commit

The directory used as SBINDIR should be ${sbindir} not ${base_sbindir}.
Reported by Martin Jansa <martin.jansa@gmail.com>

(From OE-Core rev: 2c6725c1f8427a0920c2810d649010f98b7869eb)

Signed-off-by: Phil Blundell <pb@pbcl.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake: Unbuffer stdout for log files

It is possible to lose critical log data when python exits in an
unorderly fashion via segmentation fault or certain types of crashes.
This is because the buffer characteristics are inherited from the top
level stdout, which should be set to unbuffered, for the purpose of
all the forked children.

This pushes the buffering to the OS, instead of having python managing
the buffers in its stream handler class.

This change is also to provide the ability to tail logs written from
processes in "real time" because they would be written in an orderly
fashion depending upon the OS characteristics for the file I/O.

(Bitbake rev: c6a367bc3224adafca698a4ffc5414ad83842c16)

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: event.py, knotty.py, ncurses.py, runningbuild.py: Add support for LogExecTTY event

The LogExecTTY even is intended to provide the ability to spawn a task
on a the controlling tty, if a tty is availble.  When a controlling
tty is not availble the previous behavior is preserved where a warning
is issued about the action an end user must execute.

All the available UI's were tested against the new event type.

This feature is primarily intended for hooking up a screen client
session automatically on the controlling tty to allow for a more
streamlined end user experience when using a pure command line driven
environment.  The changes that send the LogExecTTY event are in the
oe-core side.

(Bitbake rev: cffe80d82a46aaf52ff4a7b6409435754043553f)

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: progress.py: Fix traceback when running goggle ui

The following traceback appears when running the following command after the
devshell is exited.

bitbake -u goggle -c devshell busybox

-- traceback --
Traceback (most recent call last):
  File "/work/bitbake/lib/bb/ui/goggle.py", line 35, in event_handle_idle_func
    build.handle_event (event, pbar)
  File "/work/bitbake/lib/bb/ui/crumbs/runningbuild.py", line 299, in handle_event
    pbar.set_text(event.msg)
AttributeError: 'ProgressBar' object has no attribute 'set_text'

(Bitbake rev: a6cc53cdb3c34fc8fd01bbc5ce0008429dc6785c)

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: runqueue: Add --no-setscene to skip all setscene tasks

Mainly intended for the purpose of debugging or forcing builds
from source, the --no-setscene will prevent any setscene
tasks from running.

(Bitbake rev: 440e479f3e248482c38c149643403c6907ac7034)

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

terminal: Send LogExecTTY event to spawn screen

Bitbake has the ability to request to run a command
and if it is not possible fall back to emitting a
log message.  This can be used to start a screen
client automatically on the controling tty if
the UI has an interactive tty.

(From OE-Core rev: 39193bdce698b6339c3d7643eb3c1fcd2246fd56)

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

terminal: pass data store all the way through to terminal class

Passing the data store will be needed for firing a custom event
for the screen class.

(From OE-Core rev: 5ccff8d44626bfd3d1af2a7f81f0567997277809)

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: fetch2/git: Add missing mkdir

bitbake-selftest is failing due to directories not being created. This adds in an
appropriate mkdir so the tests can complete. Presumably in general OE use, something
else is ensuring the parent directory is created.

(Bitbake rev: 1270a07713e2a6c6e6fadcc61b785aebc99ae17b)

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

bitbake: fetch2: improve error output for checksum failures

* Don't print the full exception in the initial warning - if we later
  succeed in fetching the file from a mirror, we won't usually need the
  details (which are in the fetch log if they are needed); otherwise the
  full error will be printed when the fetch operation fails. Also adjust
  the conditional block so that we don't print another warning just
  mentioning we're going to try mirrors.
* Call logger.error() so that with knotty the full log is not printed
* Provide an explanation around the lines we print for easily updating
  the checksums in the recipe. We don't want users to be just blindly
  updating the recipe in case of a transient failure or deliberately
  altered remote file.

(Bitbake rev: 2793413106c925b06783beb7413aa87cbcf246c3)

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

bitbake: fetch2: make fetch failure errors more readable

Most of the time we don't need to see the fetch command; the fetch log
includes the command as a debug message in any case, so omit it. Also
adjust the way command output is printed (we don't need stderr/stdout
labelled, and print "no output" instead of "output:\nNone" when there is
no output.

(Bitbake rev: a75505a52e4da918222100221f79e8a658f90446)

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

bitbake: lib/bb/runqueue.py: fix exceptions with -k and failed targets

If a target dependency is marked as failed and yet we are continuing on
because -k has been specified, don't try to access the dependency's data
in taskData.build_targets since it will have been removed. This fixes
"IndexError: list index out of range" errors in this situation.

Also, do not print the "unhandled exception" message when SystemExit is
raised since we will have reported the actual error already in this
case (e.g. when -k has been specified and some targets failed).

Fixes [YOCTO #3133].

(Bitbake rev: 70eebc184eb1ab3678be87bed019b5beadecdc89)

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

bitbake: hob: fix Gtk-WARNINGs due to invalid markup on Back button

You can't use markup characters (e.g '<' or '>') in the labels for many
widgets - you must use the appropriate entities instead.

(Bitbake rev: 54a16ac999d4a2c4c3f8a4531e8c6fabc39a4147)

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

bitbake: hob: remove confirmation dialog on close

This is not necessary for modern applications - instead we just need to
check if we're in the middle of a build and if so, do the same thing as
pressing the "Stop" button.

(Bitbake rev: a79eb5d918239db1dade8134743e6142a4854930)

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

bitbake: hob: allow configuring default machine using HOB_MACHINE

Allow specifying HOB_MACHINE in local.conf to set the initially
selected machine. With this set, Hob will select the specified machine
and then jump straight into parsing recipes. If you do wish to change
the selected machine with HOB_MACHINE set you still can - you just need
to stop the parsing process first.

Fixes [YOCTO #3148].

(Bitbake rev: c3b623dc7d546a1ededdb532dcbcba4a6230bc65)

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

bitbake: hob: don't show error dialog for errors during building

During building we already report errors in a special tab and
indicate when the build has failed; bringing up a dialog was a
regression introduced in bitbake revision
5bab81b124087d63d6eb62a861e1241714fcd483.

Fixes [YOCTO #3151].

(Bitbake rev: cf0a67d62f631aa48d1afc3fbdd0f73995b1c401)

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