Updating the SRCREV to import the following kern-tools patch:
kgit-meta: always clear series file on branch transitions
This was triggered by the patch optimization changes, that no longer
run do_patch if a leaf/final branch is not being processed.
Without this change, invalid patches, or already applied patches in
an existig series file will be re-used which leads to missing files,
or patch errors.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ricardo Neri [Fri, 6 Feb 2015 20:59:58 +0000 (12:59 -0800)]
runqemu: Add option for BIOS binary filename
If we intend to use a custom BIOS binary file, it would be good to be
able to specify the name of the binary via the -bios option of qemu. The
name of the custom binary could be different from the default name used
by qemu. For instance, it was bios.bin in qemu 1.5.0 and changed to
bios-256k.bin for 2.1.0.
Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Tomas Novotny [Mon, 9 Feb 2015 14:14:45 +0000 (15:14 +0100)]
qmake_base.bbclass: fix hardcoded paths in a generated qt.conf
The qmake_base class generates internal qt.conf for build with hardcoded
qt4 target paths. This is ok for qt4-x11 as this is the default
$QT_DIR_NAME, but it is wrong for qt4-embedded, because qtopia is its
default directory.
Signed-off-by: Tomas Novotny <tomas.novotny@tbs-biometrics.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ricardo Neri [Fri, 6 Feb 2015 20:59:57 +0000 (12:59 -0800)]
rootfs.py: ignore "NOTE:" when catching warnings
The script looks for any warn|Warn pattern in the log. However,
it may happen that the warn|Warn pattern appears as part of the name
of the objects described in the NOTE lines of the log. Thus, to avoid
false positives, ignore the line if it contains the word "NOTE:"
Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Rob Woolley [Fri, 30 Jan 2015 21:55:09 +0000 (16:55 -0500)]
dhcp-client: Ignore partial checksums
dhclient will fail to get an IP address if run inside a guest when traffic is
flowing over a virtual network interface. The user will see the error
message:
5 bad udp checksums in 5 packets
No DHCPOFFERS received.
Unable to obtain a lease on first try. Exiting.
Failed to bring up eth0.
This is because Linux only uses partial checksums for packets that go over
virtual network interfaces and dhclient does not like this.
An extra parameter is added to decode_udp_ip_header() in dhclient to indicate
whether or not dhclient should ignore partial checksums. This is used
when the TP_STATUS_CSUMNOTREADY bit is set by the guest kernel.
This fix has been included in Fedora and Ubuntu, however it has not yet been
accepted by ISC upstream. Likely because it is specific to behaviour in Linux
and other UNIX variants do not seem to be affected.
The patch was imported from the dhcp source RPM in Fedora 21
(http://pkgs.fedoraproject.org/cgit/dhcp.git/tree/dhcp-xen-checksum.patch?h=f21)
Originally contributed to fedora-cvs-commit by David Cantrell on Jan 30 2007
(https://www.redhat.com/archives/fedora-cvs-commits/2007-January/msg01442.html)
Submitted to dhcp-bugs@isc.org - [ISC-Bugs #22806] - by Michael S. Tsirkin
(http://comments.gmane.org/gmane.comp.emulators.kvm.devel/65236)
(https://lists.isc.org/pipermail/dhcp-hackers/2010-April/001835.html)
Upstream-Status: Submitted [dhcp-bugs@isc.org] Signed-off-by: Rob Woolley <rob.woolley@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Richard Purdie [Fri, 13 Feb 2015 13:05:59 +0000 (13:05 +0000)]
gcc-target: Don't install target gcc libdir files
Installing /usr/lib/gcc/* means we'd have two copies, one from gcc-cross
and one from here. These can confuse gcc cross where includes use #include_next
and builds track file dependencies (e.g. perl and its makedepends code).
For determinism we don't install this to the sysroot, ever and rely on the
copy from gcc-cross.
[YOCTO #7287]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Thu, 12 Feb 2015 14:50:11 +0000 (14:50 +0000)]
autotools/siteinfo: Avoid races over siteinfo files
If a siteinfo enabled tasks re-executes at the wrong moment whilst something else is
in do_configure, the _config files can be removed which upsets autoconf and
causes build failures.
Use the same approach as we do for dealing with the aclocal files. We already
parse the manifests so look out any *_config files and if so, copy them, then
reference the copy from siteinfo instead. This has the advantage of also being
more deterministic.
[YOCTO #7101]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Cut and paste the pieces of build-aux/options-parser inline into the main
ltmain.sh code. This removes a performance degradation caused by the
repeated calls to func_quote_for_eval, the mechanism funclib uses
to construct the functions used for option parsing.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bruce Ashfield [Mon, 26 Jan 2015 18:19:10 +0000 (13:19 -0500)]
linux-yocto: basic octeon III support
Importing the following commits to add basic octeon III support to the
3.14 linux-yocto kernel:
1e0bbd1dd68e MIPS:OCTEON: More OCTEONIII support f51a5843ee2d MIPS: Octeon: CVMSEG LM loads may cause dcache parity errors 58bcba842781 MIPS: Octeon: Implement the core-16057 workaround ba5a219685e8 MIPS Override assembler ISA for kernel FPU instruction. 7759a0511965 MIPS donot build fast TLB refill handler with 32-bit kernels. 3f51e46ff641 MIPS Add minimal support for OCTEON3 to c-r4k.c bbde024d903a MIPS Add function get ebase cpunum 9f18310679cb MIPS OCTEON Enable use of FPU d24496ee8fc8 MIPS OCTEON Add OCTEON3 to get cpu type
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Robert Yang [Tue, 27 Jan 2015 06:53:55 +0000 (22:53 -0800)]
m4-native 1.4.9: remove it
Its do_compile would be failed once our autoconf-naitve is built:
| /path/to/m4-native/1.4.9-r2/m4-1.4.9/missing: line 54: aclocal-1.10: command not found
| WARNING: `aclocal-1.10' is missing on your system. You should only need it if
| you modified `acinclude.m4' or `configure.ac'. You might want
| to install the `Automake' and `Perl' packages. Grab them from
Also:
| /path/to/m4-native/1.4.9-r2/m4-1.4.9/missing: line 54: automake-1.10: command not found
We don't need keep a native GPLv2 recipe, so remove it.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Robert Yang [Sat, 31 Jan 2015 08:54:07 +0000 (00:54 -0800)]
coreutils 6.9: disable native
ERROR: The recipe coreutils-native is trying to install files into a shared area when those files already exist. Those files and their manifest location are:
tmp/sysroots/x86_64-linux/bin/kill
Matched in manifest-x86_64-util-linux-native.populate_sysroot
We have 8.23, so disable it rather than fix.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Robert Yang [Tue, 27 Jan 2015 08:18:55 +0000 (00:18 -0800)]
gzip 1.3.12: a few fixes
* Let gzip_1.3.12.bb and gzip_1.6.bb use gzip.inc to remove duplicated
code.
* Fixed for gzip_1.3.12.bb:
WARNING: QA Issue: gzip requires /bin/bash, but no providers in its RDEPENDS [file-rdeps]
* Remove native extend from gzip_1.3.12.bb, keep gzip_1.6.bb's.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
The patch 2.5.9 doesn't support --disable-xattr, there are already
PACKAGECONFIG and PACKAGECONFIG[attr] in patch_2.7.1.bb, so remove them from
patch.inc to fix the problem.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Robert Yang [Tue, 27 Jan 2015 06:19:33 +0000 (22:19 -0800)]
matchbox-desktop: remove 0.9.1 and 2.0 version
For 0.9.1:
ERROR: matchbox-wm: Recipe file does not have license file information (LIC_FILES_CHKSUM)
ERROR: Licensing Error: LIC_FILES_CHKSUM does not match, please fix
For 2.0:
| /path/to/matchbox-desktop-2.0/configure: line 2488: syntax error near unexpected token `build_old_libs,'
| /path/to/matchbox-desktop-2.0/configure: line 2488: ` _LT_DECL(build_old_libs, enable_static, 0,'
And there is a matchbox-desktop_git.bb which is "GPLv2+ & LGPLv2+", so remove them.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Robert Yang [Tue, 27 Jan 2015 06:05:43 +0000 (22:05 -0800)]
matchbox-wm 1.2: remove it
It doesn't work:
ERROR: matchbox-wm: Recipe file does not have license file information (LIC_FILES_CHKSUM)
ERROR: Licensing Error: LIC_FILES_CHKSUM does not match, please fix
And there is a matchbox-wm_git.bb, they have the same GPLv2+ license.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Ash Charles [Thu, 29 Jan 2015 21:34:40 +0000 (13:34 -0800)]
busybox: Include 'mesg' utility
The 'dot.profile' supplied by base-files calls 'mesg'. For simple
images using busybox e.g. core-image-minimal, this utility should be
available to avoid errors on login:
-sh: mesg: command not found
Provide mesg with busybox.
Signed-off-by: Ash Charles <ashcharles@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Use kB instead of MB for the partition size to get a better granularity.
This is needed on some SoC (i.mx, omap) where it is necessary to create
partitions as small as 64kB.
Keep the backward compatibility by assuming MB when no unit is provided.
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com> Tested-by: Maciej Borzecki <maciej.borzecki@open-rnd.pl> Acked-by: Tom Zanussi <tom.zanussi@linux.intel.com> Tested-by: Tom Zanussi <tom.zanussi@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Dan McGregor [Wed, 4 Feb 2015 16:09:12 +0000 (10:09 -0600)]
terminal.py: add tmux new window option
Add a new terminal type that makes a new window in the running tmux
session instead of splitting the window. 80x25 is not enough to run
menuconfig inside a split window, so add the option to create a new
window instead.
Use the new window option by default when the split window would be
less than 19 lines high.
Signed-off-by: Dan McGregor <dan.mcgregor@usask.ca> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ross Burton [Wed, 4 Feb 2015 21:31:28 +0000 (21:31 +0000)]
Revert "bitbake.conf: don't remove WARN_QA and ERROR_QA from hashes"
It turns out that changing WARN_QA and ERROR_QA results in do_configure's QA
postfunc re-executing, so changing a QA test results in a complete rebuild.
This is just too much and the lesser evil of needing to do a full rebuild to
verify changed QA flags is preferable to an enforced full rebuild.
Pengyu Ma [Tue, 3 Feb 2015 02:11:08 +0000 (10:11 +0800)]
mesa: Upgrade to 10.4.3
flags for EGL and GBM Gallium had been removed in 10.4,
so remove these flags, files or packages.
Signed-off-by: Pengyu Ma <pengyu.ma@windriver.com> Tested-by: Nicolas Dechesne <nicolas.dechesne@linaro.org> Signed-off-by: Ross Burton <ross.burton@intel.com>
Maciej Borzecki [Wed, 4 Feb 2015 10:53:07 +0000 (11:53 +0100)]
wic: bail out on kickstart parse errors
Kickstart file errors are not emitted by the parser. In case errors, the
current code will continue to execute possibly creating a faulty or
incomplete image. The patch enables the parser to emit all encountered
errors. The error message is shown for the user and further execution is
ceased.
Signed-off-by: Maciej Borzecki <maciej.borzecki@open-rnd.pl> Signed-off-by: Maciek Borzecki <maciek.borzecki@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Chen Qi [Wed, 4 Feb 2015 07:11:32 +0000 (15:11 +0800)]
image.bbclass: don't let do_rootfs depend on BUILDNAME
BUILDNAME is set by cooker as a string of current time. Letting do_rootfs
task depend on this variable gets us no benefit. Besides, letting do_rootfs
task depend on this variable will cause us trouble when executing
`bitbake -S none core-image-minimal'. With current code, this command
gives us error complaining about the different bashhash of do_rootfs task.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Paul Eggleton [Tue, 3 Feb 2015 15:28:52 +0000 (15:28 +0000)]
classes/license: tweak license format messages
Strictly speaking not all of these characters are operators, so reword
the message to describe them as separators. Also use the standard
"recipename: message" format.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Timo Mueller [Tue, 3 Feb 2015 12:21:28 +0000 (13:21 +0100)]
apr-util: Allow use in SDK
This adds 'nativesdk' in the BBCLASSEXTEND.
Building apr-util for the SDK requires that some paths in the rules.mk
file are fixed to point to the actual enviroment. These include pointing
the apr configuration variables to the actual apr builddir and
correcting the name of the libtool.
Signed-off-by: Timo Mueller <timo.mueller@bmw-carit.de> Signed-off-by: Lukas Bulwahn <lukas.bulwahn@oss.bmw-carit.de> Signed-off-by: Ross Burton <ross.burton@intel.com>
Maciej Borzecki [Fri, 9 Jan 2015 13:24:02 +0000 (14:24 +0100)]
wic: fix extended/logical partition layout
The patch fixes an issue in laying out extended and logical partitions
by wic. The original code produced disk images in which the size 3rd
partition as described in MBR was incorrect. Depending on the type of
file system used for that partition and size of the partition, it would
be impossible to mount the partition correctly. For instance, kickstart
file in which the 3rd partition had size of 1GB and used ext4 fs, would
result in an image with an umountable partition. The root cause is
reservation of sectors for EBR through stealing of last sector from the
last primary partition.
Signed-off-by: Maciej Borzecki <maciej.borzecki@open-rnd.pl> Signed-off-by: Maciek Borzecki <maciek.borzecki@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Chunrong Guo [Thu, 22 Jan 2015 06:27:49 +0000 (14:27 +0800)]
uboot-config.bbclass: Allow multiple U-Boot config for machine
This adds support to build multiple U-Boot configs for a machine; this
is useful when we have support for different media boots which require
different U-Boot configuration (e.g: eMMC and NAND).
Below there's an usage example:
,----[ i.MX6Q SABRE AUTO based example ]
| UBOOT_CONFIG ??= "sd eimnor nand spinor"
| UBOOT_CONFIG[sd] = "mx6qsabreauto_config,sdcard"
| UBOOT_CONFIG[eimnor] = "mx6qsabreauto_eimnor_config"
| UBOOT_CONFIG[nand] = "mx6qsabreauto_nand_config,ubifs"
| UBOOT_CONFIG[spinor] = "mx6qsabreauto_spinor_config"
`----
Signed-off-by: Chunrong Guo <B40290@freescale.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
The libtool would be very slow if run "automake/autoconf --version", for
example, when compile xz-native (make -j1, only compile, no confiure or
install):
before patched: 19s
after patched: 11s
Use plain text to instead of running them.
NOTE: it is still a little slower than libtool 2.4.2 when compile
xz-native because of other parts:
make -j1: about 2s slower
make -j32: about 0.4s slower
If we run to do_install:
(PARALLEL_MAKE = "-j32")
libtool 2.4.2:
$ bitbake xz-native -cinstall && bitbake xz-native -ccleansstate && time bitbake xz-native -cinstall
real 0m21.092s
user 0m28.292s
sys 0m3.932s
libtool 2.4.5:
$ bitbake xz-native -cinstall && bitbake xz-native -ccleansstate && time bitbake xz-native -cinstall
real 0m21.380s
user 0m31.140s
sys 0m4.068s
About 0.3s, slightly different.
But when using /bin/bash as CONFIG_SHELL, the new libtool would be much slower:
real 0m23.106s
user 0m44.044s
sys 0m4.280s
About 2s slower, for the big package like cairo, it is more slower (about 6s),
unfortunately, /bin/bash is most default CONFIG_SHELL for the recipes since
configure checks first check bash.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Barker [Sun, 21 Dec 2014 13:26:27 +0000 (13:26 +0000)]
opkg-keyrings: New recipe
This recipe wraps package and package feed verification keys into a package,
making the management and deployment of verification keys much easier. Comments
on how to select keys for inclusion in this package are provided in the recipe
file.
Signed-off-by: Paul Barker <paul@paulbarker.me.uk>
Paul Barker [Sun, 21 Dec 2014 13:06:12 +0000 (13:06 +0000)]
opkg: Backport improvements to opkg-key
The improvements to opkg-key made for v0.3.0 are backported. The rest of v0.3.0
needs further testing and bugfixing prior to release, but the improvements to
opkg-key have been tested and work well.
Signed-off-by: Paul Barker <paul@paulbarker.me.uk>
Paul Barker [Fri, 19 Dec 2014 17:35:29 +0000 (17:35 +0000)]
opkg-collateral: Simplify source files
Instead of having the source files 'opkg.conf.comments', 'dest' and 'src', we
simply have one 'opkg.conf' file containing all the necessary configuration.
This does mean that replacing 'src' in a bbappend to define distro-specific
package feeds is no longer supported, but these feed configurations should be
packaged separately from the rest of the opkg configuration anyway so that
things can be updated independently. The best recipe to use for managing feed
configs is currently distro-feed-config in the meta-oe layer.
Signed-off-by: Paul Barker <paul@paulbarker.me.uk>
Ross Burton [Wed, 21 Jan 2015 17:53:31 +0000 (17:53 +0000)]
bitbake.conf: don't remove WARN_QA and ERROR_QA from hashes
Changing WARN_QA and ERROR_QA should cause do_package_qa to re-execute, so
removing them from the sstate hashes is harmful.
They were added back when sanity testing was part of packaging and this was the
lesser evil, compared to changing sanity tests causing a re-package of
everything.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Ross Burton [Mon, 2 Feb 2015 11:03:28 +0000 (11:03 +0000)]
native: remove PN from DEPENDS automatically
If a recipe (say, wayland) has DEPENDS=wayland-native and BBCLASSEXTEND=native,
when built as wayland-native it has DEPENDS=wayland-native which results in a
circular dependency.
Typically this is resolved by having explicit DEPENDS_class-native statements
but as this is duplication which can lead to inconsistent dependencies,
automatically remove the circular dependency.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Khem Raj [Thu, 29 Jan 2015 20:44:20 +0000 (12:44 -0800)]
systemd: Backports fixes to 216
Fix systemd-timesyncd assertion
when networkd is disabled then we now do not
create /run/systemd/netif/links but timesyncd needs it. So lets
manually create this file when networkd is disabled so timesyncd
can still function
When enabling systemd-timesyncd we need systemd-timesync user
Backport patches to enable timesyncd when resolved and networkd
are disabled
replace the resolv.conf symlinink patch with a proper backport
Change-Id: I53f1a53eec4e4a4dbdfb7e8cd155d544ee5d81ec Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Forces native RDEPENDS of xmlto-native to be built and installed
since those files are required when creating yocto-docs
if not found, they are downloaded from upstream everytime
significant time difference is achieved.
[YOCTO #7148]
Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Chen Qi [Thu, 29 Jan 2015 02:22:14 +0000 (10:22 +0800)]
package_manager.py: fix rootfs failure with multilib enabled
With the current code, if we use debian package backend and enable
multilib support, the do_rootfs process would always fail with error
messages like below.
E: Unable to locate package packagegroup-core-boot
This patch fixes the above problem.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
terminal.py: fixes launching multiple windows of gnome-terminal
When resolving a patch, a new process of gnome-terminal
is created for every patch to be resolved, it "waits"
for the previous one to end, instead of launching
multiple windows at the same time.
[YOCTO #7254]
Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Chen Qi [Thu, 29 Jan 2015 05:44:18 +0000 (13:44 +0800)]
runqemu-internal: get rid of annoying but harmless error message
If tap0 interface has been created by other users on the system, we would
get error messages when using runqemu. Error messages are like below.
Acquiring lockfile for tap0...
touch: cannot touch ‘/tmp/qemu-tap-locks/tap0.lock’: Permission denied
/buildarea2/chenqi/poky/scripts/runqemu-internal: line 139: /tmp/qemu-tap-locks/tap0.lock: Permission denied
flock: 8: Bad file descriptor
The system can still boot up because runqemu would try to create a new tap
interface. So the error message above is harmless, yet somewhat annoying.
This patch fixes the above problem. With this patch, the output would be as
follows.
Acquiring lockfile for tap0...
Acquiring lockfile for /tmp/qemu-tap-locks/tap0.lock failed
Setting up tap interface under sudo
Acquiring lockfile for tap1...
Running qemu-system-arm...
Signed-off-by: Chen Qi <Qi.Chen@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Dan McGregor [Tue, 27 Jan 2015 17:21:12 +0000 (11:21 -0600)]
image.bbclass: make kernel depmod data optional
This allows an image to skip the creation of kernel depmod
data. It is useful for creating an image that will run as a
container image inside a host with no knowledge of the parent's
kernel.
Signed-off-by: Dan McGregor <dan.mcgregor@usask.ca> Signed-off-by: Ross Burton <ross.burton@intel.com>
