scripts/runqemu: show an error if /dev/net/tun is unusable

If /dev/net/tun is either not present or is not writable by the
user, then show an appropriate error message. (QEMU needs access to this
device in order to enable networking; it may be missing if it is not
enabled or loaded into the kernel, and some distributions such as CentOS
5.x set restrictive permissions upon it.)

(From OE-Core rev: a00b94900d437828f25debce1c30ffcc0bbf29e9)

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

qemu: for native, do not fail if kvm is unavailable

When building qemu-native, if the linux kvm header is unavailable (as
it is on CentOS 5.x 32-bit) then do not pass the --enable-kvm switch to
the configure script, thus avoiding failed do_configure.

(From OE-Core rev: 8c21c71f005b601f58925e9912f2cf44127e291d)

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

icecc-create-env-native_0.1.bb: drop 'PRIORITY = "optional"' as it is the default

(From OE-Core rev: 49e629e78ba914c3372a9969ad84eff20ef8f009)

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

grub_1.99.bb: drop 'PRIORITY = "optional"' as it is the default

(From OE-Core rev: 4279fdea7297223a488e93c702d83cf2e99f52f3)

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

grub-efi-native_1.99.bb: drop 'PRIORITY = "optional"' as it is the default

(From OE-Core rev: 85add8f186ce939642b99b4a547cdfa8cda3464e)

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

gdk-pixbuf: Ensure the binaries can be relocated

There are paths hardcoded into the binaries provided by this recipe. This
patch adds the neccessary environment options to ensure they can be relocated
successfully avoding build failures.

(From OE-Core rev: b9dfccc018f32a47fc045f35d5f53d7269d791ed)

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

python: bump PR due to arm OVERRIDES change

(From OE-Core rev: 86a815fe1b2e447e264cb40f6f65e3343d6b6188)

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

xf86-video-omapfb: bump PR due to arm OVERRIDES change

(From OE-Core rev: d421f5e6d32f747596fd31cb6e90f87eee806e5c)

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

qt4: use armv6 override instead of armv6-vfp

(From OE-Core rev: e093d550fa86d2c2d46b0df760d8a967b3366389)

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

directfb-examples: add package directfb-examples to OE-core

Add this package for checking whether directfb run rightly or not.
It also is an useful tool to test directfb.

[YOCTO #1674]

(From OE-Core rev: 5d5c8c581346dc0297be2bb1531f7569154b309a)

Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

dosfstools: Add patch to disable fat32 autoselection and behave as 2.10

It appears msdos image population and fat32 images are incompatible.
This reverts to the 2.10 behaviour of defaulting to fat16 instead of
using fat32 for large images, allowing image generation to work
correctly. This is a workaround and a proper fix is really needed.

(From OE-Core rev: c2de8d41236cf1293db9e6c69d69e8d14f55ffd1)

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

boost : Do not use icu when building boost.

Without this, boost needs to build icu to create images using
boost regex. RP inidicated he would rather disable icu, than add an
extra dependency to the recipe.

(From OE-Core rev: f60417055f869acb871be1f01c6900fdf685d71a)

Signed-off-by: Philip Balister <philip@balister.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

u-boot: renamed dir from uboot to the more appropriate u-boot

recipe dirs normally have the name of the (main) package and the official name is u-boot

(From OE-Core rev: e9899d52ade2181bd97dcf79bec64650e8b0f718)

Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

qt4.inc : Add uic, moc, rcc, and lrelease to tools package.

These binaries are needed in the target tools package so we can build
qt apps on the target. You also need to install qmake from meta-oe.

Tested by building gnuradio on the USRP e100 with qt support.

(This effectively reverses OE-core rev
69eeb3d2276e5b10d084b47d308ecfc8daf8b467, however test builds for
qemux86 and qemumips produced no packaging warnings such as those
described in that commit, and the architecture of the executables is
correct.)

(From OE-Core rev: c96db08915a554fb5e4bb2c360b919c8392b32c6)

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

kernel.bbclass: Allow do_compile_kernelmodules to use PARALLEL_MAKE

Without this we don't take advantage of any configured multiple CPU
cores which seems a shame.

(From OE-Core rev: 10b354c6ce7bac3b4cce5e6a649d4fd3ceca235b)

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

conf/machine/include/arm add extra MACHINEOVERRIDES like x86 does

* motivated by this NAK
  http://patchwork.openembedded.org/patch/15777/
  and today's discussion on #yocto I hope it's worth it to send this RFC

(From OE-Core rev: e3e1fef27345e2ea923b76b1e6bcb9cd5572cec6)

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

connman: improve packaging

The scripts being included on connman-scripts was the test utility and
thus we are moving them with the rest of test utilities in
'connman-tests'. A runtime dependency on python-dbus were missing as
well as those scripts uses it.

As no dhclient, resolvconf and udhcpc plugins are being generated the
respective dependency map entries were also removed.

(From OE-Core rev: 6f9594cddefe74ce7381e4fa2a68cb8a542d6d2e)

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

task-core-gtk-directfb.bb: Add task list to run gtk over directfb

Add task-core-gtk-directfb.bb to OE core for running gtk over directfb.

[YOCTO #1674]

(From OE-Core rev: ad2737079a78e1bfa0edce5bd8f4daf61ee791e4)

Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gtk: add demos to the configuation of gtk+

Add demos for checking whether gtk+ run over directfb successfully or not.

[YOCTO #1674]

(From OE-Core rev: 05af060591f3aec72c0d39a133d44db09fd8450a)

Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gtk.inc: add feature based on directfb

gtk run over x11 at current OE-core. If gtk want to run over directfb, then \
the configuration related to x11 should be disabled and directfb should be enabled.

[YOCTO #1674]

(From OE-Core rev: c95e859672599e43e156ac12dc1d919e0dd34510)

Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pango: Disable x11 when pango doesn't run over x11

pango run over x11 at current OE-core. If pango want to run over non-x11, then \
the configuration related to x11 should be disabled.

[YOCTO #1674]

(From OE-Core rev: e20bb10573dd67fd60b5b7f944eb0cbec2332a04)

Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cairo: add directfb DISTRO_FEATURE

cairo run over x11 at current OE-core. If cairo want to run over directfb, then \
the configuration related to x11 should be disabled and directfb should be enabled.

[YOCTO #1674]

(From OE-Core rev: dd5b44034014ba1962cfca1bf7430f8c8dc5243c)

Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

ed: remove unsupported option

(From OE-Core rev: 65fa6a9039043a59a236d28aacce3c424a734158)

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

bash: Add SRC_URI Checksums for GPLv2

(From OE-Core rev: ae8c1f2aacd0ed2625757f57b575962db1386e4c)

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

gmp: Add SRC_URI Checksums for GPLv2

(From OE-Core rev: 9597a66e797852a44bbcf29e722eb410fd3fd386)

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

sed: Add SRC_URI Checksums for GPLv2

(From OE-Core rev: e60ccfc4e30787f00b76124491cc53fe8494af47)

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

gzip: Add SRC_URI Checksums for GPLv2

(From OE-Core rev: b8b8b0383e806cc1d3a6f8114338112310f940f2)

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

gawk: Add SRC_URI Checksums for GPLv2

(From OE-Core rev: ede29b86b14ccea9d265a0cc84057ca3062dc9f7)

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

findutils: Add SRC_URI Checksums for GPLv2

(From OE-Core rev: 70d190b0d09a691bbc2191c93c9de75bb0437017)

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

ed: Add SRC_URI Checksums for GPLv2

(From OE-Core rev: c30c89c8dc449cf7642565f2e35c7b1a922fcc33)

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

readline: Add SRC_URI Checksums for GPLv2

(From OE-Core rev: 652afb4125597c6dbafb33c54e3300cd88461576)

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

coreutils: Add SRC_URI Checksums for GPLv2

(From OE-Core rev: 16c526219b4d2852eec534f251f1a69e8a79c4a2)

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

eglibc: Generate cross testing script during install

EGLIBC has long had cross testing capability, so use
that there are few steps needed on setting up the target
as well as the build host. This patch tries to generate
a script with all the needed environment on build host system

To use this script one needs to setup the target as described
in the testing intructions of eglibc and then this script
can directly interface to the target and run the tests from
eglibc's build directory. To run this script

bash <script> username@testing_hostname/ip

The output can be dumped into a log file for later viewing
and checked for "Error " to see the failed tests

Removes INC_PR its not used anywhere

(From OE-Core rev: 9244a3b0565fd52d7973fda42a4f706b16240316)

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

task-core-nfs: Add nfs-utils-client to RDEPENDS

include nfs client in tast-core-nfs as well
this is handy when we need it on target e.g.
testing eglibc needs it and we can just do
EXTRA_IMAGE_FEATURES += "nfs-server"

(From OE-Core rev: c09c6e61991b30c99d68fa6c2e5639d490a65e89)

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

qemu-0.15.1: Bump recipes from 0.15.0 to 0.15.1

No other changes (except checksum updates) then git mv were needed

(From OE-Core rev: 48b56cf4912d2ec324d18391865940b675a72c1e)

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

license: PR bump failure

This addtask for do_populate_lic should be reverted back
to before do_package as before do_compile causes failures
when recipes are updated.

[YOCTO #1819]

(From OE-Core rev: 513c620815fcf8e7bcc9ac38a916850ad6173a16)

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

scripts/send-pull-request: fix typo

s/No cover letter will be send./No cover letter will be sent./

(From OE-Core rev: a82e5abf399f3ced7dcb4d5c543fdea084ef6c63)

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

x86 tune: fix TUNE_PKGARCH definition for proper PACKAGE_ARCH

rpmbuild can not handle the PACKAGE_ARCH of these kinds:
x86_64-x32, core2-64, core2-64-x32

With these kinds of PACKAGE_ARCH the --target parameter of rpmbuild
becomes like: core2-64-x32-poky-linux-gnux32 ; And rpmbuild extracts
%_target (arch) wrongly as core2 generating these kinds of rpms with
incorrect filenames: zip-3.0-r0.core2.rpm

So this commit fixes the issue by making PACKAGE_ARCH like this:
x86_64_x32, core2_64, core2_64_x32
Now --target parameter of rpmbuild becomes like:
core2_64_x32-poky-linux-gnux32 ; And rpmbuild extracts %_target (arch)
correctly as core2_64_x32 generating these kinds of rpms with correct
filenames: zip-3.0-r0.core2_64_x32.rpm

(From OE-Core rev: 1a599cc822ad517f9ba70ceb0e39c5572d37a5a6)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

openssl-1.0.0e: Update x32 Configure

Make linux-x32 as close to linux-x86_64 as possible:

1. Add -mx32 -DMD32_REG_T=int.
2. Changed to -O3.
3. Remove -pipe -g -feliminate-unused-debug-types.
4. Remove -DHAVE_CRYPTODEV -DUSE_CRYPTODEV_DIGESTS.
5. Add :::x32 for multilib.

Signed-Off-By: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-Off-By: H.J. Lu <hjl.tools@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libacpi: Use the cross strip instead of host strip

to avoid this build error on x32:

| strip: Unable to recognise the format of the input file `test-libacpi'
| make: *** [test-libacpi] Error 1
| ERROR: oe_runmake failed

(From OE-Core rev: 4d7882a57210c88dd9e2b8619b2bd0fa4bdbf9f8)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

siteinfo: create a new siteinfo for x32-linux

X32 has different sizes for many types:

1. off_t and ino_t are 8 byte, instead of 4 byte.
2. pthread_mutex_t is 32byte, instead of 24 byte.

(From OE-Core rev: 061d8015d6f15825bf9c4245671313bec35cefbf)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pax: fix compilation with x32 toolchain

off_t is 8byte for x32.  We need to check both _FILE_OFFSET_BITS and
size of off_t to see if file offset is 64bit.  This patch adds
AC_CHECK_SIZEOF(off_t) and checks SIZEOF_OFF_T == 8.

(From OE-Core rev: 4c2ec631bf2b9581233a470c415f4d5cfe59a2a4)

Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mesa: Properly create share library for x32

Also pass -mx32 parameter to gcc to create create share library
for x32.

Signed-Off-By: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-Off-By: H.J. Lu <hjl.tools@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

findutils: Fix compilation for x32 toolchain

Work around gnulib time_t assumption in findutils for x32

time_t is 64bit and long int is 32bit on x32.  But gnulib used in
findutils assumes time_t values fit into long int.  Such assumption is
invalid for x32 and should be removed.

This patch is a workaround to compile gnulib for x32.

(From OE-Core rev: 8a80a3510c65419a6fcd9e38ad3ef4234778b921)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-Off-By: H.J. Lu <hjl.tools@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libacpi: Fix libdir for x32

libacpi doesn't use base_libdir for install.  This patch adds
base_libdir support so that x32 libraries are installed properly.

(From OE-Core rev: 98a363ef0202511cf32f8408011c4ad178173af7)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pulseaudio: fix compilation with x32 toolchain

This commit makes assembly syntax compatible with x32 toolchain
to avoid these x32 gcc errors:
| pulsecore/svolume_mmx.c: Assembler messages:
| pulsecore/svolume_mmx.c:107: Error: `(%esi,%rdi,4)' is not a valid
base/index expression
| pulsecore/svolume_mmx.c:135: Error: `(%esi,%rdi,4)' is not a valid
base/index expression
| pulsecore/svolume_mmx.c:161: Error: `(%esi,%rdi,4)' is not a valid
base/index expression
| pulsecore/svolume_mmx.c:162: Error: `8(%esi,%rdi,4)' is not a valid
base/index expression
| pulsecore/svolume_mmx.c:180: Error: `(%esi,%rdi,4)' is not a valid
base/index expression
| pulsecore/svolume_mmx.c:210: Error: `(%esi,%rdi,4)' is not a valid
base/index expression
| pulsecore/svolume_mmx.c:244: Error: `(%esi,%rdi,4)' is not a valid
base/index expression
| pulsecore/svolume_mmx.c:245: Error: `8(%esi,%rdi,4)' is not a valid
base/index expression
| make[3]: *** [libpulsecore_1.1_la-svolume_mmx.lo] Error 1

Orignally these assembly lines are written for x86_64 ABI, now they are
also compatible with
X32 ABI.

(From OE-Core rev: ccf01c858218ae67d609ced3a05d9a93a6ffc5d4)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mdadm: Make custom CC definition conditional

By hardcoding CC's definition in the Makefile, all the gcc parameters
set by tune settings are lost. Causing compile failure with x32 toolchain

As the bitbake defined CC is good, there is no need to redfine CC in the
make file, hence making the CC definition in the Makefile conditional.

This fixes bug: [YOCTO #1414]

(From OE-Core rev: c877ec8ff1dfbf0123f94711f44cea043815a9c1)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gst-fluendo.inc: remove unneccessary hack

This fixes bug: [YOCTO #1403]

the custom definition of CC was causing build isuses with x32 toolchain.
And also I found out that the hack is not neccessary anymore. the
affected gst-fluendo-mpegdemux recipe builds fine without the CC hack.

(From OE-Core rev: 81cbedaaddd589e03fe2f6cb789907b1e555a9d6)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

dosfstools: update native to 2.11

This unify recipes for target and native builds and also drops the the
already merged patches.

(From OE-Core rev: 3a401ddce55e185c8ccfdc43c1440fd77daff9ae)

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

lzop-1.03: add recipe

This is needed by some kernels when CONFIG_KERNEL_LZO=y (specifically, given
the current defconfig, this affects linux-omap4 2.6 in the meta-ti layer).

(From OE-Core rev: 9b3991cf6a8bb751e45c1abfc48d65d952c048b8)

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

task-self-hosted: Move tools to host-tools task

Moved kernel modules for iptable and tun, along with
dhcp-client iptables, mc, screen to host-tools as they
are not sdk.

(From OE-Core rev: 24498e983ab8455ef0dda67efa8c16c1197f3597)

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

task-self-hosted: Add host-tools task

Added tools to create and manage ext* partitions
e2fs*, parted

(From OE-Core rev: 4723c5d93de0a1d9e2b7b592f4784211984cadc1)

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

task-self-hosted: Added perl-pod (slight reorder)

(From OE-Core rev: cee52dbf46578ff087a8a0786f8daba22414bfc9)

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

task-self-hosted: Reorder by Alpha

No additions/subtractions

(From OE-Core rev: f66ae52a6f7b723738fade94bab3424862e71443)

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

texinfo: fix compile failure due target makedoc binary being used

Need to have the texinfo-native build and install a host sysroot makedoc
binary and then patch the target build to use this binary. This requires
that we don't ASSUME_PROVIDED texinfo-native any longer since we need to
install this makedoc tool which is not part of the normal distrubtion.

[YOCTO #1664]

(From OE-Core rev: 9fa98de54a73465f06484ba863eccf1e07cc1e2a)

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

bitbake.conf: remove texinfo-native from ASSUME_PROVIDED

We need to build texinfo-native to get and install the makedoc tool

[YOCTO #1664]

(From OE-Core rev: 8899f4840787ef043d952f8ea2ce5d78e5cc41ab)

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

ncurses: libcurses.so should link to libncurses.so

The current libcurses (backward compat) links to libncurses.so.5.9,
this causes a problem since this library also needs libtinfo, the
libncurses.so, uses a "ld script" to include both, libcurses.so
should point to libncurses.so (the script) not the library.

(From OE-Core rev: 6e3a0c0eb6938f3f04fb42784fe1744f30762d3b)

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

connman: Default to enabling ethernet interfaces

This fixes various qemu image failures where connman has changed its
default and would deconfigure network interfaces causing qemu images
and runtime image testing to fail.

(From OE-Core rev: c9a420be15d54a8494cb45d6c932b86895e94e13)

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

scripts/runqemu-ifup: Ensure netmask is set correctly

Without this the command will add a route for the subnet 192.168.7.0 which
means multiple qemu instances can't operate correctly since all but the last
one will be masked out.

(From OE-Core rev: 9e00d6b343120496ec0dd72240c7b04e0a8b7eaa)

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

license/copyleft: exception for ast parse

As the standard for LICENSE format is not well defined, we
may well run into issues where LICENSE cannot be parsed via
ast. In cases like this, we need to warn and continue.

(From OE-Core rev: 07f94fa1d661d38b775ce2d99b84e5610c369392)

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

license.bbclass: remove deb exclusion

This commit requires 4a13766c7b223d as it removes the exclusion
of deb package types from license manifest creation.

(From OE-Core rev: 5328587ecaa14a54ad7bced6f7ead218471dbf86)

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

license.bbclass: remove bashism.

bashism in create_license_manifest removed.

(From OE-Core rev: 035a569d5241be9a217dbdbd6650ee04c4c4d67f)

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

classes/rootfs_deb: implement query functions for buildhistory

Implement the functions required for buildhistory to be able to query
installed packages, get dependencies etc. for deb-based images.

(From OE-Core rev: 58fbb430040c9cce9f2c5f1515a4453dd49032cc)

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

connman: avoid checking for linux/if_alg.h header

Disable building of 'alg-test' and checking for if_alg.h header.

(From OE-Core rev: db692fd5706eab9bfd94c115591d1411cf8e196e)

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

connman: disable ntp support

OE-Core does not provide ntpd at this moment and as there's no known
user of this plugin at this moment we are disabling it.

[YOCTO #1817]

(From OE-Core rev: c67d7c9215bd8af2b5dcc4a72b6edf00157724d5)

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

linux-yocto: update rt branches to 3.0.12 + rt30

Updating the SRCREVs to v3.0.12 and the rt30 patch.

(From OE-Core rev: cb42d82afc4b8fda723c4ef2db91880b69f09ae6)

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

linux-yocto: update META SRCREV for EFI changes

Adding EFI configuration from the following meta branch commit:

    Add EFI scc and cfg files

    Basic EFI support only requires CONFIG_EFI=y, this is sufficient for
    some boards, and desirable for small configs. This is done with efi.scc.

    Additional support for CONFIG_EFI_VARS, CONFIG_EFI_PARTITION, and CONFIG_FB_EFI
    is provided via efi-ext.scc (extended) as this pulls in the block layer,
    framebuffer support, and virtual terminals.

    I'd like EFI_VARS to be part of the base config, but I have received
    reports of it failing in some situations. Keeping it separate ensures
    basic boot can work with the fragments as defined.

(From OE-Core rev: d590f6d12ba7d650ba1a0dd9d5211fb3b2c50fb1)

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

linux-yocto: v3.0.12 + EFI fixes

Updating SRCREVs to pickup the v3.0.12 -stable updates and to get
three EFI fixes from Darren Hart:

 x86, efi: Convert efi_phys_get_time() args to physical addresses
 x86, efi: Make efi_call_phys_prelog() CONFIG_RELOCATABLE-aware
 x86/rtc: Don't recursively acquire rtc_lock

(From OE-Core rev: f87f62f1e7de88b482a9d1877deae42939aedbe0)

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

connman: update to 0.78

Updates the package to 0.78 bringing changes done on meta-oe and extra
cleanups as:

* xtables_version.patch: removed as it has been merged upstream

* use sed to fix dbus access policy

* remove build depends on dhclient as connman has an internal client
  now

* make wifi and bluetooth building optional

* add ppp in depends that is used by some plugins

* add ntp in depends and enable ntp plugin

* package tools

* enable fake plugin

This uses PR as "r1" since it was previously available on meta-oe.

(From OE-Core rev: 4215b2fb04b3ed61e8e1cf0847639cf6fe31d337)

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Acked-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

connman: move files from connman-0.75 to connman dir

This is being done in preparation to get newer version of connman into
OE-Core.

(From OE-Core rev: 154b76a5103a09ab3acd26c428c9ebb21f99fe8f)

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

connman: drop git version

This was clearly unused as dbusperms.patch cannot be fetched by this
recipe and noone noticed the build failures caused by it.

(From OE-Core rev: 1383b18041f2fb988651f7b53ed126e01f5e2797)

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

scripts/send-pull-request: allow sending of patches without a cover letter

(From OE-Core rev: 326ef69d1296cd35a86d1eaee0741e6fc956af52)

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

gtk.inc: ship gtk-demo to independent package

gtk-demo can test gtk over directfb

[YOCTO #1674]

(From OE-Core rev: 4a13766c7b223d82e8cf682db999a135d2b8412c)

Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gconf: Disable dbus-x11 when x11 isn't in DISTRO_FEATURES

If x11 isn't in DISTRO_FEATURES, then an error information "no providers ..." will \
arise. I modified this bb file to disable "dbus-x11" when no x11 in DISTRO_FEATURES.

[YOCTO #1674]

(From OE-Core rev: da2f6a60fe6b06ebffe92c3d99564711071981c9)

Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

qemu-config: Disable dbus-x11 when no x11 in DISTRO_FEATURES

dbus-x11 could be not needed by an image without x11.
So I modified this bb file to enable dbus-x11 when x11 is in DISTRO_FEATURES and disable \
dbus-x11 without x11 feature.

[YOCTO #1674]

(From OE-Core rev: f78cb9fbc90f6e40cd20c59ba544389662d87659)

Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package.bbclass: Adding license collection

emite_package_data needs to have LICENSE, as pkgdata is as an
audit of licenses collected during do_populate_license.

(From OE-Core rev: 60f0b866b4b7c4aa337a2ff764455741a73665da)

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

OECore license fixes: meta/*

This is a quick audit of only the most obviously wrong licenses
found within OECore. These fixes fall into four areas:

- LICENSE field had incorrect format so that the parser choked
- LICENSE field has a license with no version
- LICENSE field was actually incorrect
- LICENSE field has an imaginary license that didn't exist

This fixes most of the LICENSE warnings thrown, along with my prior
commit adding additional licenses to common-licenses and additional
SPDXLICENSEMAP entries.

HOWEVER..... there is much to be done on the license front.
For a list of recipes with licenses that need obvious fixing see:

https://wiki.yoctoproject.org/wiki/License_Audit

That said, I would suggest another license audit as I've found
enough inconsistencies. A good suggestion is when in doubt, look at
how openSuse or Gentoo or Debian license the package.

(From OE-Core rev: 3083dd70b3a9fa01fcc3cf00373b05502505996e)

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

license: manifest and license path

This is a fairly large commit, so I want to step through some of
what this accomplishes:

1. Additional licenses

I've added/modified/corrected some of the common licenses
within OE-core in prep. for a major license audit. Most of these
are in SPDX format. A few, there is no OSI equivalent.

2. Additional SPDX mappings

I've added some additional SPDX mappings to account for removing
some duplicate licenses. (ie GPL-2 and GPL-2.0 were the same)
I've also remapped a few things to more accurately reflect what
they should be pointing at.

Note: Artistic/LGPL/GPL/MPL. Quite a few LICENSE fields list these
licenses. They make no sense and need auditing. In a future commit
I have some fixes to particularly egregious LICENSE fields, but
a full audit should be done. I've listed to obvious candidates at:

https://wiki.yoctoproject.org/wiki/License_Audit

3. License manifest

We now have a license manifest generation that occurs in rootfs for
everything BUT .deb. This requires the changes Paul Eggleton has
done to rootfs_* particularly the list_installed_packages function.
The manifest is accurate during a parallel bitbake now (Weee!) and
is prime for my planned SPDX format manifest during the next period.

4. License manifest on image.

We also want the ability to add licenses to the image. This
functionality is also in base-files and will be stripped out in my
next commit. The manifest is not added by default and is a two var
setting in license.conf:

If I want *just* the manifest on the image (small) then I set:
COPY_LIC_MANIFEST = "1"

This copies the manifest to:

/usr/share/common-licenses/license.manifest

If I want the actual PKG license text on the image (much larger)
I need to set both both COPY_LIC_MANIFEST and COPY_LIC_DIRS in
license.conf. This will create:

/usr/share/common-licenses/(package name)/(licenses in LIC_SRC_URI)

Word of warning. This can be larger than wanted depending on image
and is probably ripe for linking licenses, but I ran out of time this
week to get that done.

5. Custom License search path.

We now have the ability to add licenses to the build without touching
common-licenses. This is set via license.conf:

LICENSE_PATH += "/path/to/custom/licenses"

You want to make sure the license is unique. license.bbclass picks the
first license it finds.

(From OE-Core rev: 558b5043e1d5a36caff137093fd04abcf025af1c)

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

gettext-minimal-native: Fix License

Only .m4 macros are included in this recipe so the correct license is FSF-Unlimited
as specified in those marcros, not the main gettext license. This patch corrects
that mistake.

(From OE-Core rev: 25dc5af38230fdf5891597c07c4f96c633001006)

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

Add copyleft compliance class

Deploys sources for recipes for compliance with copyleft-style licenses
Defaults to using symlinks, as it's a quick operation, and one can easily
follow the links when making use of the files (e.g. tar with the -h arg).

By default, includes all GPL and LGPL, and excludes CLOSED and Proprietary.

(From OE-Core rev: f0c36abba71c261f33c9906f122dd6cd9ed19221)

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

oe.license: add license flattening code

This flattens a license tree by selecting one side of each OR operation
(chosen via the user supplied function).

(From OE-Core rev: 6984961314c8ba2aceab9acabb658f96ed249fef)

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

license: split license parsing into oe.license

In addition to moving this functionality to oe.license, makes the string
preparation more picky before passing it off to the ast compilation. This
ensures that LICENSE entries like 'GPL/BSD' are seen as invalid (due to the
presence of the unsupported '/').

(From OE-Core rev: 20d4068045c76e9dc2aff0c152dd02d6a109c9dd)

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

oe.test_types: move into an oe.tests package

(From OE-Core rev: 36cc35b4cbb91049a63daa7c915f538047db0f76)

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

sysstat: don't run populate-volatile.sh update in do_rootfs, first boot or without populate-volatile.sh

(From OE-Core rev: 6bf82c26f95359c08611bc9a44accf1593b2a850)

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

wget: enable https and openssl

(From OE-Core rev: 852c4b8dcc99aa93e45680c0e42917d7dd202d55)

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

self-hosted-image: remove extra space since we use additional disk instead

(From OE-Core rev: 9c82bad3c341d1ef54537f0c023998747c125381)

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

initrdscripts: add sleep to avoid kernel messages before install message

As suggested by Darren Hart

[YOCTO #725]

(From OE-Core rev: ed76654c613d38095f085640acb6591b9739a60e)

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

grub: Create /etc/grub.d

(From OE-Core rev: 5f1b7ef11cf08e9a21040f16c908ada22230837b)

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

initramfs-framework: provides a modular initramfs

Provides the API and modules for a modular initramfs. The currently
included modules are:

 * initramfs-module-debug adds support to dynamic debugging of
   initramfs using bootparams

 * initramfs-module-udev: enables udev usage

 * initramfs-module-mdev: enables mdev usage

 * initramfs-module-e2fs: adds support for ext4, ext3 and ext2
   filesystems

(From OE-Core rev: 7b69ad2167a1f0e57db82817b98a0cbcb70a0dd3)

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

qt4: various tidy-ups

* -reduce-relocations is already specified in qt4.inc, no need to add it
  in qt-${PV}.inc
* Add QT_GLFLAGS to QT_CONFIG_FLAGS in qt4.inc rather than spreading it
  through other inc files.
* Add "-xmlpatterns -no-rpath -qt3support -silent" to QT_CONFIG_FLAGS in
  qt4.inc rather than qt-${PV}.inc; these have been supported since at
  least 4.5.2 and therefore shouldn't be version-specific.
* Move "-no-fast -silent -no-rpath" to EXTRA_OECONF in qt4-native.inc
  rather than the versioned recipes for the same reason
* Remove redundant setting of LICENSE in qt4-embedded.inc

(From OE-Core rev: c9f714062c1100cedbcb2c16d16656e3f5442133)

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

qt4: fix debug symbol stripping

Adds a patch imported from Debian to not strip the executables as part
of the Qt build process. This fixes the warnings shown at packaging
stage as well as the content of the debug symbol packages.

(From OE-Core rev: c2f6dc09a7663f69308558553be9420df3a81b6b)

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

bitbake.conf/image.bbclass: Ensure images use the correct passwd/group files

We need pseudo to use the rootfs passwd/group files belonging to the
rootfs when building images. This patch ensures that we use the rootfs
files instead of those in the sysroot which can lead to incorrect file
ownership issues.

(From OE-Core rev: c4da803ef78322b758380eb0af0dcb73cae6553c)

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

distro_tracking_fields.inc: Update dropbear,iproute2,net-tools tracking information.

(From OE-Core rev: 743472ea6871a1cf0ac0f024e0b645b2594ac01c)

Signed-off-by: Mei Lei <lei.mei@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

dropbear: Upgrade from 0.53.1 to 2011.54

This new version added ALLOW_BLANK_PASSWORD option. So change the allow-nopw.patch content to enable this function.

(From OE-Core rev: e876096fcbb42039d568a7acbc506e4099e9a443)

Signed-off-by: Mei Lei <lei.mei@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

dpkg: Fix preinst execution

Add a patch to fix exeuction of pre/post install scripts. See the patch
header for more details.

(From OE-Core rev: dc433b556f5b44a6d818a81df6be484ad59daebd)

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

image_types.bbclass: disable cleanmarkers for jffs2 images

* Add '-n' to suppress cleanmarker nodes since jffs2
* is usually used for NAND flash and the cleanmarkers are created in
* the OOB area by flash_eraseall -j
*
* From man pages: -n, --no-cleanmarkers
* Do not write cleanmarker nodes to the beginning of each erase block.
* This option can be useful for creating JFFS2 images for use on NAND flash,
* and for creating images which are to be used on a variety of hardware with differing eraseblock sizes.

(From OE-Core rev: 70a276509f0f006fcc269296afc3dcc88d2825e1)

Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

image_types.bbclass: implement jffs2 summary images (sum.jffs2)

* Building the jffs2 filesystem to include summary information speeds up
* the mount process considerably at the cost of increased size.
* The rate of speedup is generally higher on NAND chips and on the chips
* where the erase block size is large.

(From OE-Core rev: 2cb6a0d7aaa1ba712a899ee77a4c720da4e50214)

Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

rootfs_ipk: implement FEED_DEPLOYDIR_BASE_URI support

this feature was present in OE-classic, is documented in OE-core
but its implementation is actually missing.
The present implementation is directly copied from OE-classic
and tested on an arm926 machine.

(From OE-Core rev: 63b619fd7a3d0ca55a965f0a5b99f6e9c8584ea4)

Signed-off-by: Eric Bénard <eric@eukrea.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

matchbox-session-sato: respect VIRTUAL-RUNTIME_initscripts settings

(From OE-Core rev: 493746f1d7f1410284880203c1ac462d61ab5299)

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

subversion: update 1.7.0 to 1.7.2 and force libtool recreation

* subversion-1.7.* had libtool-2.4, oe-core now has 2.4.2 and it was
  failing:
  x86_64-linux-libtool: Version mismatch error.  This is libtool 2.4.2, but the
  x86_64-linux-libtool: definition of this LT_INIT comes from libtool 2.4.
  x86_64-linux-libtool: You should recreate aclocal.m4 with macros from libtool 2.4.2
  x86_64-linux-libtool: and run autoconf again.

(From OE-Core rev: aa9d0de4225fe482ddbf1486f8018bc87419e228)

Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>