glib-2.0: Fix localedir expectations for uclibc

DATADIRNAME is used by many applications using autotools
to install locale data, we get a wrong value for uclibc
systems since it does not recognise it as proper linux
systems and start putting locale info in /usr/lib instead
of /usr/share

(From OE-Core rev: 7897ab5f54d2dc21795b2a53b82b9c911157f0ca)

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

gnutls: Link in pthread explicitly for uclibc

Otherwise we end up with missing symbols from libpthread

(From OE-Core rev: 9f09353a1548d392a3f71e800be8e1b849960584)

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

uclibc: Remove iconv.h

We get it from libiconv on uclibc anyway no need to
transplant uclibc one

(From OE-Core rev: 9da1e04b1543d0a08fdd679a72c9cf6455811ab9)

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

uclibc: Implement clock_adjtime()

systemd needs it
Also update to latest master

(From OE-Core rev: ced5bc5537ccff89145dae1220a40ab8fd4783f5)

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

bitbake: toaster: added file types to the Outputs column in the build page

The file types are displayed in the Outputs column in the build page.
The file types are derived from the target image filenames.

[YOCTO #5947]

(Bitbake rev: 37ae4e94d6991d4f05b0236b525e29797ed6e49c)

Signed-off-by: Farrell Wymore <farrell.wymore@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Sort layers in alphabetical order

In the Summary tab of the Configuration page, make
sure that the table of layers is sorted by layer name
in ascending alphabetical order.

(Bitbake rev: f339555df40307420ce80a4ef8cba1a4d284d380)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Remove the data-toggle attribute

Help icons with the .get-help class do not need the
data-toggle attribute. That attribute initialises the
Bootstrap tooltips, but they are already initialised
in main.js for any element with the .get-help class.

(Bitbake rev: 886a87de2d36382f40830c0d28fd2f55871d0993)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Don't show clear search button if text input field is empty

The clear search button should only show when the search text input
field is populated. If it is empty (as it happens when a filter
returns no results) the clear search button should not display.

(Bitbake rev: 6cc4aaf08b51bb6cd8d1813c33989a788bd06990)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Fix Recipe sorting in tasks table

In the tasks table and the other tables derived from it
(Time, CPU usage and Disk I/O) sorting by Recipe was not
working correctly. This change fixes the problem by
specifying use of the recipe name to sort.

(Bitbake rev: fa24c857e6f5b3c56f89a2a85dea6ad271f82c01)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Not using task_color tag for execution heading

The modifications to the task_color tag in commit
23a7c338d387ac2ba13a7a1114a4abc75228c960 broke the styling
of failed tasks in the tasks.html template. Undo the
changes to the task_color tag and use an if statement
instead to set the .muted class when the execution
heading says "Not executed".

(Bitbake rev: e3b3205674f606b927f1bf568202a592ca6453c9)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Presentation fixes for task.html

Tidying up the presentation in the task.html template.
The changes include:

* Correct the markup for the help tooltips
* Fix the help content for the outcome heading
* Make sure <dt> tags do not show for empty log
file, time, cpu and disk I/O values
* Eliminate an extra <dl> for tasks with sstate attempts
* Add <strong> tag to the sstate restored alert
* Replace the .alert-info class with the .muted class
for the no dependencies messages
* Make sure the Executed heading does not inherit
the .red class for failed tasks
* Format time and cpu values to make sure they only
show 2 decimal digits

(Bitbake rev: bbc22958ab37dcd44c03420a7b8f842a1f4e51b1)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Shorten dependency labels

Changing "Task depends on" to "Dependencies" and
"Task reverse dependencies" to "Reverse dependencies".
This matches the labeling to recipe and package
information, with the additional advantage of
making the labels shorter.

(Bitbake rev: 6147339ac2f93a9e60eed8975ac1784caa06bb0b)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Give an extra space to the caret

In the "Collapse variable value" button in configvars.html,
the caret was too close to the button label. Giving it
a bit of extra space.

(Bitbake rev: 5e45f2bb4b0d9b56b5f4734cf40d3a267654ccb9)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Add no search results page

Adding the no search results page to the built packages
(bpackage.html), variables (configvars.html), recipes
(recipes.html) and tasks (tasks.html) tables.

The change copies the code from the build.html template
into the other 4 templates. There is probably a smarter way
of doing this, though.

(Bitbake rev: 3feff77091b3926c10ca8f30890c9049633ccb66)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Fixing "Show all builds" link

Replacing the placeholder attribute with the value attribute
in the no search results page broke the "Show all builds" link.
This change applies the inline javacript used for the clear search
button for the "Show all builds" link, which fixes the problem.

(Bitbake rev: a58b88aebfde54ea97740f27335582fac195dfdc)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Amend help text in package_built_detail.html

Matching the help text in the tabs to the design specification.

(Bitbake rev: cc478b3b0befc0c8fce93d896074f8d07eaac362)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Small tweaks to the no results page

Correctly align the Search button to the text input field,
add .btn class to the clear search button, replace the
placeholder attribute with the value attribute so that you
can edit your search query, remove the clear search button
from the tab index so that you don't clear the search by
mistake and edit the margins of the .no-results class.

(Bitbake rev: 4ada4307623c1e27b589831d359dc6351e6e3ad5)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Conflicts:
bitbake/lib/toaster/toastergui/templates/build.html
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Replace hyphens with underscores in package name

Package details pages were using hyphens to separate package
name from package version. Changing them to underscores.

(Bitbake rev: c0820f93e0387d2b6ead9e2fa7205d49001d27e6)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Format package size in recipe.html

Package sizes in the recipe details page (recipe.html)
were displaying in bytes. Apply the filtered_filesizeformat
project tag to show the package size in a more human
readable format.

(Bitbake rev: 018db5cf683755a7a41b0ef491e130809e1ff003)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Make Order column part of the minimum table

The Order column in the tasks table should be always shown and
be part of the minimum table (i.e. its checkbox should be
disabled in the Edit columns menu). Changing views.py to
make sure this is the case.

(Bitbake rev: a928e0cfb82dac0eb920913e3f9c5b45e3dcdf2f)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Move <tbody> outside for statement

In the package details pages, the <tbody> tags where inside
the for statements, which caused multiple <tbody> tags to
be generated inside a single table.

To make sure only one <tbody> tag exists per table, moving
the <tbody> tag outside the for statement.

(Bitbake rev: 0c111b24e9f86130bc43c1327a6d12026e92cdf2)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: improve recipe matching for native tasks

This patch improves the recipe matching algorithm for
for matching recipes for native tasks.

(Bitbake rev: c350e4924abab8688c539608fd7f3af687d7265a)

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

bitbake: toaster: fix target file inode type information

This patches fixes the inode type saved when writing the
target file list information.

(Bitbake rev: 9f34a1c5e94d73cdba1def7059c60211514e054c)

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

bitbake: bitbake: cooker: some IMAGE_FEATURES not recognized

Fixes an issue in hob which happened when the local.conf file was
modified externally by appending "eclipse-debug" to the IMAGE_FEATURES
variable. The reason of the problem is that some IMAGE_FEATURES are
not available in the image.bbclass file and they are declared in the
core-image.bbclass. Now a default hob image will inherit core-image.

[YOCTO #5711]

(Bitbake rev: 81413d94f40f58d790d7a7dc4259108f9c5d4fc0)

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

bitbake: toaster: add errors and warnings listing

Add the errors and warnings listing to the build
dashboard page.

(Bitbake rev: 2004c0fd1dd049d26279ee10a18e69e6852247fa)

Signed-off-by: Amit Kumar Chaudhary <amit@floatingpondtech.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Add counters to filters

Added count to filters in Build, Tasks and Config variables.

[YOCTO #5821]

(Bitbake rev: afd73f8ee47d0fcf600fa919f2d2bc1b7d2eeadf)

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

bitbake: toaster: add license manifest path to database

Based on image_name, the target is obtained, and the path
is added to the database.

[YOCTO #5649]

(Bitbake rev: 911b5191133956c30d53f57629c115db196b9ac8)

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

bitbake: toaster: populate target image file table

Using ImageFileSize Metadata event, the image output file and its
size are populated into target_image_file table.

[YOCTO #5189]
[YOCTO #5228]

(Bitbake rev: a0b06d362b9aa08fda293489467af343c6ca6de4)

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

eglibc: Update SRC_URI to point to final download location

We have been using interim location for the src_uri
now the tarballs are placed in usual location so update
the src_uri accordingly

(From OE-Core rev: 58aae40bbf8edd9c5ca70b2094eeb724391f1993)

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

wic: Use pseudo for ext mkfs command

The switchover from populate-exfs.sh to mke2fs forgot to preserve
pseudo, add it back.

(From OE-Core rev: 032309928f931b32cf63a0ebf174de029ba17401)

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

wic: Use mke2fs to create the ext2/3/4 image

OE-core commit f3a95ca6886b55e5819b068bdbd2cceb882d91a6 removed the
populate-extfs.sh. So mke2fs should be used to create ext2/3/4 image.

(From OE-Core rev: 4dea928eef325364922b1e0cbabc2a581356e804)

Signed-off-by: João Henrique Ferreira de Freitas <joaohf@gmail.com>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

Create gstreamer-1.0 package groups

This recipe is based on gst-meta-base

(From OE-Core rev: f5b16a1879351eac015a541150dec547f971bb31)

Signed-off-by: Sébastien Mennetrier <s.mennetrier@innotis.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gnomebase: remove trailing whitespace

It seems that the bitbake parser is incredibly forgiving of trailing whitespace
in multi-line variables, but delete it to be sure.

(From OE-Core rev: 0c404fa1a0372c56f700eb93900abd83175199ef)

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

oe-setup-builddir: small rework

Changes:
- drop useless subshell creation in test:
if ! (test -r "$BUILDDIR/conf/local.conf"); then$
- replace "source" builtin by "." (bashsism)
- fix indentation 4 spaces (drop some tabs too)
- fix return => exit  (return is not allowed in main)
- drop "sed -i" (doesn't exist in BSD sed)
- for homogeneity, always use [ ] (instead of test)
- replace old [ "x" = "x$VAR" ] by [ -z "$VAR" ]

(From OE-Core rev: 3a116577446f02bda0ef4e035360293ff73c9eef)

Signed-off-by: Matthieu Crapet <Matthieu.Crapet@ingenico.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: runqueue: Improve sigchld handler

The sigchld handler was reaping any processes and this was leading to
confusion with any other process handling code that could be active.
This patch:

a) Ensures we only read any process results for the worker processes
   we want to monitor
b) Ensures we pass the event to any other sigchld handler if
   it isn't an event we're interested in so the functions are properly
   chained.

Together this should resolve some of the reports of unknown processes
people have been reporting.

(Bitbake rev: fe8baaa2f533db7a1b7203476c675588923d8d45)

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

populate_sdk_deb: Fix meta-toolchain-sdk with amd64 ARCH

Map SDK_ARCH x86_64 to DEB_SDK_ARCH amd64

Without this patch meta-toolchain-gmae fails to do_populate_sdk

| The following packages have unmet dependencies:
|  nativesdk-packagegroup-sdk-host : Depends: nativesdk-autoconf but it
is not installable
|                                    Depends: nativesdk-libtool but it
is not installable
|                                    Depends: nativesdk-shadow but it is
not installable
|                                    Depends: nativesdk-unfs-server but
it is not installable
|                                    Depends: nativesdk-makedevs but it
is not installable
|                                    Depends: nativesdk-automake but it
is not installable
|                                    Depends: nativesdk-qemu but it is
not installable
|                                    Depends: nativesdk-pkgconfig but it
is not installable
|                                    Depends: nativesdk-pseudo but it is
not installable
|                                    Depends: nativesdk-qemu-helper but
it is not installable
|                                    Depends: nativesdk-opkg but it is
not installable
|  packagegroup-cross-canadian-qt5022 : Depends:
gdb-cross-canadian-x86-64 but it is not installable
|                                       Depends:
binutils-cross-canadian-x86-64 but it is not installable
|                                       Depends:
gcc-cross-canadian-x86-64 but it is not installable
|                                       Depends: meta-environment-qt5022
but it is not installable

(From OE-Core rev: a3ccbdffc04d16ae56699314dbd7ee8b6de75267)

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

lsbtest: fix comparison bashism

== is a bashism use = instead.

(From OE-Core rev: c90d1047c41148cbd57f26b5a34563346602a71b)

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

btrfs: patch to allow for relative paths

This seems to address the btrfs image failure

[YOCTO #5146]

(From OE-Core rev: 9b22b32863f9110522362706a28036cc945201d5)

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

bind: Update to 9.9.5

Remove CVE patches that are in bind
Updated COPYRIGHT includes date changes the NetBSD Copyright
Modifies the Base BSD License to 3-Clause (removes advertising clause)w
Add patch to disable running tests on host
Add python-core to RDEPENDS for dnssec-checkds and dnssec-coverage and fix path to python

(From OE-Core rev: 041576d6d63ad807ca405dcea9eeecf1c9ccd7fe)

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

dhcp: Update to 4.3.0

Rebase fix-external-bind patch
LICENSE file has date update to 2014

(From OE-Core rev: 58fb2f8eac69bc6ae5bcba8227d161888af5a230)

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

util-linux: Update to 2.24.1

rebase native patch to remove sundisklabel as it's deprectated upstream
Update LIC_FILES_CHECKSUM for modified text, no License Changes
remove deprecated elvtune flag
Rebase the fix-configure patch due to change in configure.ac

(From OE-Core rev: f1faa0dcc6103fe720142b154ffbe3970f44d957)

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

populate_sdk_*.bbclass: remove old rootfs_list_installed_depends()

We're using the python routines now.

(From OE-Core rev: 04fa624394a6400803425fd61a79ce9552a79f99)

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

buildhistory.bbclass: Fix dependency files creation

Call the new python routines.

[YOCTO #5904]

(From OE-Core rev: 570baf4921a5b34cc97834e670b143073d136dea)

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

bitbake: bitbake-worker: Ensure children have default sigterm handler

The children of the worker should have the default SIGTERM handler,
else they'll try and do cleanup which should only happen in the
parent leading to all kinds of bizarre build failures.

(Bitbake rev: a53c8d1f846d94082aa459996c4114f10970b8ef)

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

bitbake: bitbake: fix typo in variable name

When passing -t with an incorrect server type, the error message was using a
variable that doesn't exist.

(Bitbake rev: 98b991287df06cd89955c1d0591fce3b5d4403d1)

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

bitbake: toaster: fix 'show all' labels for tasks and variables

Fix the 'Show all' button label to indicate 'variables' and 'tasks'
instead of the internal page label for the variables page and the
Disk I/O, CPU Usage, and Time pages.

[YOCTO #5915]

(Bitbake rev: e84866b7aa5a1b272c81fde548f3c7828b036bb9)

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

bitbake: toaster: add 'bblayers.conf' files to local config filter

In the variables table, add to the local configuration
filter the 'bblayers.conf' in addition to the 'local.conf'
files.

[YOCTO #5912]

(Bitbake rev: 3c41481fd00056712348d305d0246d84e9a2ff0a)

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

bitbake: toaster: add empty outcome support to task page

Add the new "empty" outcome value to the task detail page,
with help text, help notice, plus sorting for empty in the
tasks page.

[YOCTO #5917]

(Bitbake rev: 874b627b96809c274023b5fa330dfbdd14676d19)

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

bitbake: toaster: Fix help text for task details outcome

Changes made to models.py and task.html to make sure the
help text for the task outcome is displayed.

[YOCTO #5917]

(Bitbake rev: f3001af0b73f103a5d37af180e8c4c54a6249e97)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: erase checks for stop command

We remove the checks for the stop command, since we assume
the user really means it wants to shutdown the system when
he issues the command.

[YOCTO #5376]

(Bitbake rev: 5d6594de8876be8ce4df56b846fee0fc687a6261)

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

bitbake: toaster: add commands to list and delete builds

We add Django commands for the manage.py to manage the database
content.

The two commands added are:
* buildslist - produces a list of current builds
* builddelete - deletes a build and all associated data from the database

(Bitbake rev: e9a8c32512bb270cda3dee4a3ed5fd22204c24bc)

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

bitbake: toaster: write files-in-image to the database

Adding code to write files-in-image data from the metadata
event to the database.

(Bitbake rev: f3a69ef7cc536a4b879d60936199a1002584c4f4)

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

bitbake: toaster: update database schema

This is a schema update to account for

* relaxing constraints in Target_Image_File table
* modifying permission filed to string value
* adding license_manifest_path field
* adding image_size field

(Bitbake rev: f8488845ee9670e16a63c331dd2e97fdb314929c)

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

classes: Add gummiboot class

Adds a gummiboot class similar to grub-efi class and makes the necessary
changes so it can be used for live/hddimg images as well.

One can set EFI_PROVIDER = "gummiboot" in local.conf to use gummiboot instead of grub-efi.
Gummiboot requires some kernel options that are not enabled by default, so one has to build
with KERNEL_FEATURES_append = " cfg/efi-ext".

The install scripts have been updated too, keeping the old behaviour around,
but accounting for the new boot loader config files (if they exist).
It can be argued that the installer and bootimg are a bit wierd and not necessarily correct,
but I wanted to have the exact same behviour with gummiboot.
With the default EFI_PROVIDER = "grub-efi" nothing changes, everthing should be just as before.

I've tested live boot, install and normal boot on:
    - FRI2
    - genericx86-64 on NUC
with:
  EFI_PROVIDER = "gummiboot"
  KERNEL_FEATURES_append = " cfg/efi-ext"
in local.conf.

(From OE-Core rev: b457e40fc69cc6503dc566f16495f03606e5333b)

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

bootimg/grub-efi.bbclass: allow using a different class for EFI images

Abstract away some names so one can select using EFI_PROVIDER a different
class than grub-efi for populating live images, basically allowing the use
of a different bootloader than grub-efi.

(From OE-Core rev: 14e5de3b8f4d5902d0ac683ff45fb878e88b40ef)

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

recipes-bsp: Add gummiboot recipe

gummiboot is a simple UEFI boot manager.

Recipe imported from meta-intel with these changes:
 - drop PR and update configure options
 - upgraded to latest version

A couple of notes:
 - If you wish you can install the gummiboot package on the target and
use 'gummiboot install' to add the payload to the ESP (see gummiboot --help,
just make sure the ESP partition has the boot flag on as gummiboot won't accept it
otherwise). However the point of this recipe is to be used by bootimg.bbclass and
generate images with gummiboot instead of grub-efi.
 - You need a kernel which has CONFIG_EFI_STUB=y at least
 - The default linux-yocto kernel config does not enable that, easiest way
to enable is to build with KERNEL_FEATURES_append = " cfg/efi-ext" in local.conf

(From OE-Core rev: 8f95d0598c31fff76e2a58fecb6c96197121a044)

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

recipes-bsp: add gnu-efi recipe

gnu-efi is required by gummiboot.
Imported from meta-intel, PR dropped, no new version available.

(From OE-Core rev: fdb65c569db2b55023f9d1d72959e4e7d187a1d8)

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

mtd-utils: Update version to include fixes after 1.5.0

This includes fixes on master branch not in the 1.5.0 release

(From OE-Core rev: 51a2478b5542b2a1315c98d3f807535b9f0c46d2)

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

e2fsprogs: fix cross compilation problem

When compiling e2fsprogs for qemumips64, we got compilation errors.
The root cause is that parse-types.sh script would fail for mips64.

However, the type size checking in parse-types.sh doesn't make much
sense in case of cross compilation. This is because that the typedef
statements in asm_types.h are for the target machine while the generated
binary in parse-types.sh script is executed on build machine.

This patch fixes such problem.

(From OE-Core rev: 45877518e2d277c63d5a561408fbfcd4cb5a2a76)

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

.templateconf: New file for customized template defaults

This file will allow easy customization of the build tools,
in particular the default setting of TEMPLATECONF, which
should reduce the need to fiddle with scripts in the future.

(From meta-yocto rev: 75ba1055872b74735c674cca2e8d54c2d4a545a0)

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

scripts/oe-setup-builddir: Keep track of TEMPLATECONF setting

Keeping track of the TEMPLATECONF variable in the build
tree will let this script produce the same output when
listing 'conf-notes.txt' every time the script is run,
regardless of whether or not TEMPLATECONF has been
provided by the user.

Note that the default value for TEMPLATECONF now comes from
an easily customizable file $OEROOT/.templateconf

(From OE-Core rev: 4474357faf2cbca250e99630d749776fbc80d44b)

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

oe-init-build-env: Improve script sourcing detection.

This script is only useful when sourced into a shell.
These changes improve the detection of this operation,
no matter how the script is referenced.

(From OE-Core rev: cac863e958a0c8fe0f8a84dc194273c699f0c40f)

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

oeqa/utils/qemurunner: search for login string in the entire boot log

If some service is to spammy we might miss the login, so search
in the entire log instead of just the last few lines.
Use qemu in regex too (to avoid a login string from some service).

(From OE-Core rev: 1117d25fdfabbf1afdc7030e05efc6a32df871e3)

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

oeqa/targetcontrol: fix loading a controller using a class name

This was wrong and if one would do TEST_TARGET = "SimpleRemoteTarget"
instead of TEST_TARGET = "simpleremote" it would complain
that there is no such controller when there is.

(From OE-Core rev: 47d2049d13ab71e0310e9eedaf307d6c3e530b44)

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

oeqa/targetcontrol: allow a custom port for SimpleRemoteTarget

We had the ability to use a custom port for the ssh connection,
but we weren't using it.

(From OE-Core rev: c1f2a3c41969df0b7f08cf314b2cb7c9a6030092)

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

diffutils: enable ptest support

Install diffutils test suite and run it as ptest.

(From OE-Core rev: 14cc60db64975ad7cb8ae7d50e0233e9d59713df)

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

xf86-video-intel: add recipe for 2.99.910, remove the git one

Intel graphics stack releases >= 2013Q3 need
xf86-video-intel >= 2.99.902. However, keep the stable release around
too, in case people need it.

The git recipe is not really used. Remove, since it has missing
PACKAGECONFIG, license checksums and so on.

(From OE-Core rev: f707b6d81d2548e1bc8effdf267d1e40cc2cb806)

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

classes/sstate: don't overwrite buildhistory vardepsexclude

This was stomping over the value added in buildhistory.bbclass, leading
to task signatures *still* changing when buildhistory was added to
INHERIT.

Fixes [YOCTO #5897].

(From OE-Core rev: f7a0aa412a921a41d301be072cedcda2662ffbd7)

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>

dmidecode: add powerpc to compatible host

(From OE-Core rev: e69d9eaf4f81bfcf5b5aa73dd29784548f1fa166)

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

libsdl: bump PR

* another PRINC removed

(From OE-Core rev: 85db931054caa4dfb7bfca596558dcff4de286fc)

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

kernel-yocto: replace '=' with '+='

The dependency to CCACHE_DIR was moved to ccache.bbclass in
commit 2acf8da4f13c175ea818b9514677b7059de1e3e2:
[ ccache: Separate out into its own class ]

then the '=' should be replaced by '+=', otherwise, it will overwrite
the original ${CCACHE_DIR} in dirs.

Signen-off-by: Ming Liu <ming.liu@windriver.com>
(From OE-Core rev: e8b90254747651670031e6b2b8a702732124ecac)

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

ptest: Don't enable ptest for nativesdk and cross-canadian packages

The ptest is useful in context of deployment to the target hence
we don't need ptest for nativesdk and cross-canadian packages

(From OE-Core rev: c0c96df9f6d571b6d1e26013822375942003e3b9)

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

eglibc: Fix build for FSL ppc/fpu

The patch needs to adapt to upstream changes
should fix errors on e500mc+ based SOCs

This should fix [YOCTO #5871]

(From OE-Core rev: 6235cc3ccf98dce15ffe3313cf4e6cdf9c922bce)

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

ltp: Don't link against libfl

We have already defined yywrap function in scan.l file. After this, we no
longer need to link against libfl.
Since we are using --as-needed by default in linker, which means we end
up with errors like

libfl.so: undefined reference to `yylex'

(From OE-Core rev: db8b59b4ffee73c6e80029083793531d7fd408c4)

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

package_deb: Map TARGET_ARCH x86_64 to DPKG_ARCH amd64

Without this patch packages are generated as x86_64. Which cannot be
installed by default.

root@qt5022:~# dpkg -i alsa-utils_1.0.27.2-r0_x86-64.deb
dpkg: error processing alsa-utils_1.0.27.2-r0_x86-64.deb (--install):
 package architecture (x86-64) does not match system (amd64)
Errors were encountered while processing:
 alsa-utils_1.0.27.2-r0_x86-64.deb

(From OE-Core rev: a08eacc6d821d6946b23a99bca5abf785875b1cf)

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

oeqa/utils/qemurunner: use the right kill call so we don't send SIGTERM to bitbake-worker

After recent bitbake-worker changes [1] this was killing the bitbake-worker,
which isn't what we want and kill(-pid) was the wrong call anyway.

runqemu.pid is the right PGID as the process was started with preexec_fn=setpgrp
(so no need to do os.getpgid(self.runqemu.pid))

[1] http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=15688798520896690561824b2fdc227c8a365c82

(From OE-Core rev: f4bf21b40ec6fa07c4a7bb450e3cfae12128dc2c)

Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: runqueue: Use SIGCHLD instead of polling waitpid

Instead of a significant number of calls to waitpid, register a SIGCHLD
handler instead.

(Bitbake rev: 76029d08ad56a0a264ff9738a0336971a455b7f5)

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

bitbake: server/process: Optimise latency when finishing idle functions

When idle functions finish, its likely we have some other work
to do, so don't sleep in the select call but instead, skip it.
This removes small amounts of latency in common commands.

(Bitbake rev: 069d6538f83b607cb46c6fe21bf6c596e8b99242)

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

bitbake: server/process: Drop unnecessary exit delay

When the server exits, we no longer appear to need this delay. This
is likely due to improvements in the various exit codepaths. There
is therefore no longer any point in taking the latency hit.

(Bitbake rev: 8e75ee29ae07e13f23525c5c6045fbf6cdbe7675)

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

bitbake: server/process: Use a pipe for quit events instead of Event()

Its not possible to notice the change of status of an Event() in
the select call we sleep in. It would be possible in python 3.3 but
for now use a pipe instead. This removes small latency when bitbake
commands finish since the system doesn't sit in the select call.

(Debugging these kind of issues is apparent by setting a long sleep
for the select call)

(Bitbake rev: def28239b0f0d5f1cf13214b263114a5328538b7)

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

bitbake: knotty: Remove latency when exiting

There is no point in waiting 0.25s for when we should be processing
the shutdown. This simply reordering removes latency from the
bitbake command.

(Bitbake rev: f147b41bcaf9d05b5ba3a70100f1ca799979aee7)

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

bitbake: knotty: Add missing continue statement for runQueueExitWait event

The continue statement was missing for this event and the event was then
listed in the "known safe to ignore list". Clean this up.

(Bitbake rev: c4ee342300bf905e6e3bef581c61b86289461536)

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

bitbake: providers/runqueue/taskdata: Optimise logger.debug calls

A run of "bitbake bash -c unpack" when the task has already been
completed resulted in about 9000 calls to logger.debug(). With this
patch which comments out some noisy/less usefull logging and moves
other logging calls outside loops, this number is reduced to 1000
calls. This results in cleaner logs and gives a small but
measurable 0.15s speedup. The log size dropped from 900kb to 160kb.

(Bitbake rev: d2677f084fe1d8846db77d89ef5e6ffb18dc171a)

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

bitbake: cooker: Handle SIGTERM more gracefully

If the cooker receives a SIGTERM it currently hangs using 100% CPU,

This patch adds in an intercept for the event and puts the cooker into
shutdown mode allowing it to exit cleanly/safely and avoiding the hang.

(Bitbake rev: 00c22434123739b0819b31d7b1d353901a3e12da)

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

bitbake: server/process: Deal more gracefully with SIGTERM

Currently a SIGTERM to the UI process causes the UI simply to lock up.

By setting an exit flag, the waitEvent can raise a SIGINT, allowing the
UI to break out the event loop and exit. Currently this is results in a
traceback but that is more desirable than a hanging process.

(Bitbake rev: 0d12041eceeae6bba2034b04913bb13abd67bd15)

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

bitbake: server/process: Use the setFeatures command on the server instead of a manger

The use of a manager in the process server causes some issues since it remains
around for the lifetime of the server even though its only used during
initialisation and the system doesn't respond well to SIGTERM events
to the extra process (and two threads) the implementation involves.

Switching to a dedicated command simplifies the server process structure.

(Bitbake rev: 74532a7cf8ccea8b85f1cda5d5bc23d2f3c72a08)

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

bitbake: cooker/command: Add setFeatures command

Add a command to allow backends to set particular 'features' on the
cooker (server).

(Bitbake rev: f547d6ec6cfd677d71fa96dd3c69823c00dc6c69)

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

bitbake: runqueue.py: Gracefully handle a missing worker process

If the worker has already gone missing (e.g. SIGTERM), we should
gracefully handle the write failures at exit time rather than throwing
ugly tracebacks.

(Bitbake rev: 1b1672e1ceff17fc4ff8eb7aa46f59fce3593873)

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

bitbake: runqueue.py: Handle worker disappearing gracefully

If the worker (or fakeworker) process disappears for some reason, the
system doesn't currently even notice. To fix this, we call waitpid
periodically, looking for exit events of our children. If these
occur, we can gracefully shutdown the server.

(Bitbake rev: ee28ddadaa7ef91e4d4b7d22fc267382aaad6d01)

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

bitbake: bitbake-worker: Gracefully handle SIGTERM

Currently if bitbake-worker handles a SIGTERM, it leaves the child
processes to complete or hang. It shouldn't do this so hook the SIGTERM
event and gracefully shutdown any children.

(Bitbake rev: 551406f3f9ee94de09d2da6e16fea054c6dbfdb7)

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

bitbake: knotty: Deal with exceptions not resetting terminal configuration

When an exception occurred, the terminal parameters (such as echo)
may not be reset correctly. This change ensures they do get
atexit time in all cases, avoiding the terminal corruption issues
that could sometimes occur.

(Bitbake rev: e1d89166f2dfe46412ff9a5610dd57b0cef74fe3)

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

bitbake: user-manual: Review edits applied throughout from Paul Eggleton.

Applied review edits from Paul Eggleton spanning the entire book.
Small areas affected in all chapters except the "Hello World"
Appendix.

Noteworthy changes in the variables chapter where I added two new
variables: BBINCLUDELOGS_LINES and BB_CONSOLELOG.

(Bitbake rev: 2d4b09be70c6df0c1605f7e291149c682999cf50)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: user-manual: Review edits applied from Paul Eggleton.

Review of the entire manual by Paul.  I have implemented his
suggestions throughout.

(Bitbake rev: 5cd310d1df194cd171691a4bcfb98024e2bc66b8)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: user-manual: removed SVK references - all types.

(Bitbake rev: fbbd4cd54753a5d91b5c342d60d007bd2b172570)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: user-manual-ref-variables.xml: Edits to BB_GENERATE_MIRROR_TARBALLS variable

Added some clarification around who would like to set this
variable and why.

(Bitbake rev: 97acce37528ed11165e37996c7076de46c349b0a)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: user-manual-metadata.xml: Added BB_ORIGENV example.

(Bitbake rev: d4328e2df355252870c3bf01031538a77795cedb)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: user-manual-metadata.xml: Grammar fix.

(Bitbake rev: 71e314b86957ecc01d77262d0a000f6f31db2bcc)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: user-manual-execution.xml: Review edits for "Execution" chapter.

Applied various changes from Richard Purdie's review of the
"Execution" chapter.

(Bitbake rev: 714c0e93bc2d1ae6227991a1c6e4c88504c1870c)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: user-manual: Review edits from Richard (second draft)

Applied the comprehensive set of review comments from Richard
Purdie.  All files affected.  One major point here was that the
"BitBake Command" chapter was eliminated.  This information was
folded into various areas of the book.  Consequently, the bits
including the file for make had to be updated.

(Bitbake rev: 8ec38c6b456a92a0e0b9b04c2793a5b148be5027)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: user-manual: Added "Hello World" Appendix.

I took Bill's chapter and made it into an appendix.  I did some
re-writing to make it not so much like a getting-started feel,
although it still leans way that way for an appendix.  The content
is not complete.

Had to add in a line to the user-manual.xml file so that the
new appendix would be part of the book.

Had to use a different form of the command in the
user-manual-cusomization.xsl file in order to not through a bunch
of errors for an unrecognized parameter value.  I commented out
the existing one.

(Bitbake rev: 80e9306c288ca2ab42585f99fb0f396253cb8253)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>