Dominic Sacré [Wed, 25 May 2016 11:13:44 +0000 (13:13 +0200)]
dropbear: Remove incorrect SFTPSERVER_PATH from CFLAGS
Openssh now installs the sftp-server binary as /usr/libexec/sftp-server,
whereas the dropbear recipe assumes a different path.
Dropbear uses the correct path by default, so it's no longer necessary
to override SFTPSERVER_PATH via CFLAGS.
This fixes SFTP access to systems using dropbear as the SSH server.
Signed-off-by: Dominic Sacré <dominic.sacre@gmx.de> Signed-off-by: Ross Burton <ross.burton@intel.com>
Jussi Kukkonen [Wed, 25 May 2016 10:10:39 +0000 (13:10 +0300)]
docbook-xsl-stylesheets: Upgrade 1.78.1 -> 1.79.1
* Use $PV at appropriate places in do_install
* Install some new files, avoid installing 25M of java archives
* License checksum change is just copyright years changing
* Fix the URI delegation file (that prevents fetches from
http://docbook.sourceforge.net/ during build) to refer to a
non-versioned directory
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
This dependency was floating, which results in non-deterministic builds. Add
a configure argument and associated PACKAGECONFIG to fix this.
The libxml dep is only needed when bind was built with a dep on libxml due to
its httpstats feature. So, when you enable the httpstats config in bind, and
want to build dhcp, be sure to also enable the bind-httpstats config in dhcp.
Signed-off-by: Christopher Larson <chris_larson@mentor.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Marek Vasut [Tue, 24 May 2016 13:06:48 +0000 (15:06 +0200)]
gcc: Backport nios2 gcc ICE fix
Backport a patch from mainline gcc 5 branch to fix ICE triggered
when cross-compiling libdrm for nios2 architecture. The same ICE
is fixed in gcc6 already.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Khem Raj <raj.khem@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Brendan Le Foll [Tue, 24 May 2016 11:12:16 +0000 (12:12 +0100)]
npm: add target_arch flag to npm
npm takes a target_arch flag which needs to be set to do some gyp compilations
correctly. It also doesn't use the same mapping as OE for target arch so a
small function is required to make the mapping work. Function is taken from
meta-nodejs
Signed-off-by: Brendan Le Foll <brendan.le.foll@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bruce Ashfield [Tue, 24 May 2016 15:53:59 +0000 (11:53 -0400)]
linux-yocto/4.4: beaglebone: build in the usb controller drivers
Merging the following meta data change:
[
In the current codes, we build the drivers for usb controller as
modules. But for some image types, such as minimal or
full-cmdline, these driver modules are not installed to the rootfs by
default. This makes the using of the usb pretty inconvenience. So
make them all builtin.
Reported-and-suggested-by: hiims <h@101.org.il> Signed-off-by: Kevin Hao <kexin.hao@windriver.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
]
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This fixes a couple issues introduced by the previous patch. There were
a couple spots where HOST_CFLAGS needed changing to CFLAGS. This fixes builds
with security_flags.
Signed-off-by: Christopher Larson <chris_larson@mentor.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Otavio Salvador [Mon, 23 May 2016 20:45:27 +0000 (17:45 -0300)]
openssl: Ensure SSL certificates are stored on sysconfdir
Debian and other generic distributions has moved the certificates for
sysconfdir (/etc/ssl) and made the libdir content to link for it.
This provides several advantages specially for read-only
rootfs. Another benefit is that it ensures foreign implementations
(e.g: BoringSSL, from Chromium, when running with OpenSSL backend for
the certificates) to find the content correctly.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Randy Witt [Mon, 23 May 2016 21:57:00 +0000 (14:57 -0700)]
sstate.bbclass: Don't create symlinks, download to the correct location
Previously the sstate was all downloaded to the same directory and then
symlinks were added in the directories that pointed to the siginfo and
sstate in the parent directory.
This change makes it so that now the files are just downloaded to the
correct location without the need for symlinks.
Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 24 May 2016 04:18:51 +0000 (16:18 +1200)]
devtool: upgrade: tweak conflict handling
Make a couple of changes to the rebase operation:
1) Only wrap the actual rebase command in try...except since a failure
in any of the other commands should be an error, not a warning
2) If it's a conflict (which unfortunately we can only tell by checking
for the keyword "conflict" since git doesn't return error codes based
on the type of error) then print a message clarifying that the user
needs to resolve the issue themselves to finish the upgrade.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 24 May 2016 04:18:50 +0000 (16:18 +1200)]
devtool: upgrade: handle upgrading recipes with a versioned inc file
The gdb recipe in OE-Core has an inc file with the version in it;
since the inc file is pulled in with a "require ${PV}.inc", when
upgrading the recipe we need to also rename the inc file it will fail to
parse and the upgrade itself will fail.
Fixes [YOCTO #9574].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Mon, 23 May 2016 09:32:23 +0000 (10:32 +0100)]
useradd: Fix infinite build loop
http://git.openembedded.org/openembedded-core-contrib/commit/?id=642c6cf0b6a0371de476513162bd0cefa9c438b3
introduces a problem if the USERADD_PARAM variable has trailing
whitespace as the code infinitely loops causing build hangs.
Add a similar sed expression to $remaining to avoid this.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Martin Jansa [Fri, 29 Apr 2016 17:46:09 +0000 (19:46 +0200)]
linux-yocto-rt, core-image-rt*: Explicitly skip when PREFERRED_PROVIDER_virtual/kernel isn't set to linux-yocto-rt
* just like linux-yocto-dev is doing
* fixes following errors in world builds:
ERROR: Nothing PROVIDES 'linux-yocto-rt' (but /home/jenkins/oe/world/shr-core/openembedded-core/meta/recipes-rt/images/core-image-rt-sdk.bb DEPENDS on or otherwise requires it)
ERROR: linux-yocto-rt was skipped: PREFERRED_PROVIDER_virtual/kernel set to linux-yocto, not linux-yocto-rt
ERROR: linux-yocto-rt was skipped: PREFERRED_PROVIDER_virtual/kernel set to linux-yocto, not linux-yocto-rt
ERROR: Required build target 'core-image-rt-sdk' has no buildable providers.
Missing or unbuildable dependency chain was: ['core-image-rt-sdk', 'linux-yocto-rt']
ERROR: Nothing PROVIDES 'linux-yocto-rt' (but /home/jenkins/oe/world/shr-core/openembedded-core/meta/recipes-rt/images/core-image-rt.bb DEPENDS on or otherwise requires it)
ERROR: linux-yocto-rt was skipped: PREFERRED_PROVIDER_virtual/kernel set to linux-yocto, not linux-yocto-rt
ERROR: linux-yocto-rt was skipped: PREFERRED_PROVIDER_virtual/kernel set to linux-yocto, not linux-yocto-rt
ERROR: Required build target 'core-image-rt' has no buildable providers.
Missing or unbuildable dependency chain was: ['core-image-rt', 'linux-yocto-rt']
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Patrick Ohly [Wed, 18 May 2016 16:24:16 +0000 (19:24 +0300)]
image.bbclass: additional output in create_symlinks
When a symlink does not get created, it is useful for debugging to log
what would have been created and why it was skipped.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com> Signed-off-by: Ed Bartosh <eduard.bartosh@intel.com> Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Patrick Ohly [Wed, 18 May 2016 16:24:14 +0000 (19:24 +0300)]
image.bbclass: support duplicate compression types
When a derived distro adds a certain type, say zip, to
COMPRESSIONTYPES and later OE-core does the same, we end up with the
type being listed twice, and that would have undesired effects
(commands generated twice).
So to support such loosely coupled extension, we de-duplicated the
list of types first.
Alternatively, such a situation could also be treated as error. But that
seems unnecessary because typically commands for the same type will also
do the same thing.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com> Signed-off-by: Ed Bartosh <eduard.bartosh@intel.com> Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Dai Caiyun [Sat, 27 Feb 2016 23:05:53 +0000 (02:05 +0300)]
coreutils: Fix rootfs creation errors
1) error: file /usr/share/man/man1/su.1 from install of shadow-doc-4.2.1
conflicts with file from package coreutils-doc-6.9-r5
2) error: file /usr/share/man/man1/hostname.1 from install of net-tools-doc
conflicts with file from package coreutils-doc-6.9-r5
3) error: file /usr/share/man/man1/hostname.1 from install of net-tools-doc
conflicts with file from package coreutils-doc-8.25
Signed-off-by: Dai Caiyun <daicy.fnst@cn.fujitsu.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Dai Caiyun [Sat, 27 Feb 2016 23:03:45 +0000 (02:03 +0300)]
net-tools: Fix rootfs creation errors
1) error: file /usr/share/man/man1/hostname.1 from install of net-tools-doc
conflicts with file from package coreutils-doc-6.9-r5
2) error: file /usr/share/man/man1/hostname.1 from install of net-tools-doc
conflicts with file from package coreutils-doc-8.25
Signed-off-by: Dai Caiyun <daicy.fnst@cn.fujitsu.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
ghostscript: Update URL_SRI considered as 'old release'
The Ghostcript project started to place their tarballs in two places
starting at 9.19 as explained in [1]. 9.18 version is considered old,
so including the 'old-gs-releases' in the URL.
[1] http://downloads.ghostscript.com/public/
[YOCTO #9573]
Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Roy Li [Tue, 10 May 2016 02:09:33 +0000 (10:09 +0800)]
systemd: re-enable mount propagation for udevd
With MountFlags=slave, those mounts then become private to the systemd-udevd
namespace and are no longer accessible from outside the namespace, which is
not expected
Signed-off-by: Roy Li <rongqing.li@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Richard Purdie [Sat, 21 May 2016 11:30:11 +0000 (12:30 +0100)]
bitbake.conf/toolchain-shar-extract: Use en_US.UTF-8 as locale
Under python 3, if we spawn python processes, we need to have a UTF-8 locale,
else python's file access methods will use ascii. You can't change that mode
once the interpreter is started so we have to ensure a locale is set. Ideally
we'd use C.UTF-8 since OE already forces the C locale but not all distros support
that and we need to set something so en_US.UTF-8 seems as standard we we can get.
Also set this into the environment used when installing SDKs since
python can be run and we need to ensure we use a standardised locale
which is available from things like buildtools-tarball.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Fri, 20 May 2016 10:21:18 +0000 (22:21 +1200)]
devtool: sdk-update: drop support for local updates
Having two code paths here makes maintenance difficult, and it doesn't
seem likely that you would use the local case in real usage anyway, so
drop the local support entirely.
This should allow us to resolve [YOCTO #9301].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sat, 21 May 2016 10:55:36 +0000 (11:55 +0100)]
nativesdk-glibc: Extend relocation support to locales
Currently locales are not found in a relocated buildtools-tarball
such as that used in eSDK. This breaks bitbake when used under python3.
This patch adds enough relocation magic to nativesdk-glibc so that
the binary locales can be found even in a relocated buildtools-tarball
and bitbake works successfully under python3. The eSDK also works
correctly after this change too.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sat, 21 May 2016 10:53:08 +0000 (11:53 +0100)]
glibc-locale: Enable binary locale generation for nativesdk-glibc
python3 has much stricter locale requirements than previous versions.
If a locale isn't present, python3 reverts to ascii mode under which
bit
bake can't operate.
We therefore need working binary locales in things like uninative-tarball and buildtools-tarball. This patch enables binary locales for nativesdk-glibc.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Andrew Bradford [Fri, 20 May 2016 19:04:05 +0000 (15:04 -0400)]
lttng-modules: Upgrade to 2.7.3 release
Linux 4.5 and later cause lttng-modules versions prior to 2.7.2 to fail
to compile due to Linux vmscan changes. See lttng-modules git commit d0d2908478bdc8c36faaeae6fcb687052cb5f93b on lttng-modules branch
stable-2.7: "Fix: update vmscan instrumentation for kernel 4.5".
Signed-off-by: Andrew Bradford <andrew.bradford@kodakalaris.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We weren't consistent in the HOST_ (aka BUILD_) and non-HOST_ flags, so we
were using BUILD_CPPFLAGS to compile target stuff, for example. Sort that out,
and make sure we obey LDFLAGS.
Signed-off-by: Christopher Larson <chris_larson@mentor.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Otavio Salvador [Wed, 18 May 2016 16:40:48 +0000 (13:40 -0300)]
core-image-weston: If X11 is enabled, add XWayland support
When the distribution has X11 and Wayland backends, we enable XWayland support
and include matchbox-terminal as a test application for the XWayland backend.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Ross Burton <ross.burton@intel.com>
Tom Hochstein [Wed, 18 May 2016 16:40:47 +0000 (13:40 -0300)]
weston-init: Use weston-launch when starting weston as the first windowing system
When weston is started as the first windowing system (i.e. not under X
nor under another Wayland server), it should be done with the command
weston-launch to set up proper privileged access to devices.
Signed-off-by: Tom Hochstein <tom.hochstein@nxp.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Ross Burton <ross.burton@intel.com>
Otavio Salvador [Wed, 18 May 2016 16:40:44 +0000 (13:40 -0300)]
weston-init: Error out if loading a nested instance
The Weston nested instance support is not implemented. This errors out
displaying an informative error message so someone insterested on it
may look at implement this later.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Ross Burton <ross.burton@intel.com>
Tom Hochstein [Wed, 18 May 2016 16:40:41 +0000 (13:40 -0300)]
weston: Provide a default version of weston-launch that doesn't require PAM
weston-launch requires PAM for starting weston as a non-root user.
Since starting weston as root is a valid use case by itself, we check
the distro for 'pam' and build weston-launch with or without
non-root-user support.
Signed-off-by: Tom Hochstein <tom.hochstein@nxp.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Ross Burton <ross.burton@intel.com>
Tom Hochstein [Wed, 18 May 2016 16:40:36 +0000 (13:40 -0300)]
weston: weston-launch: Handle invalid command line options
Exit the program if an unrecognized command line option is found.
Signed-off-by: Tom Hochstein <tom.hochstein@nxp.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Ross Burton <ross.burton@intel.com>
However, that broke the generation of the opkg-based images, that lacks
openssl. The error is a python stack trace, showing that shutil.copytree
tries to copy a non-existing directory.
This relates to [YOCTO #9040].
Signed-off-by: Anders Darander <anders@chargestorm.se> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
bavery [Thu, 19 May 2016 22:04:39 +0000 (15:04 -0700)]
gdb: Backport patch to changes with AVX and MPX
The current MPX target descriptions assume that MPX is always combined
with AVX, however that's not correct. We can have machines with MPX
and without AVX; or machines with AVX and without MPX.
This patch adds new target descriptions for machines that support
both MPX and AVX, as duplicates of the existing MPX descriptions.
The following commit will remove AVX from the MPX-only descriptions.
This commit is backported from 7.12
Signed-off-by: bavery <brian.avery@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Peter Seebach [Wed, 18 May 2016 21:30:47 +0000 (16:30 -0500)]
acl.inc, run-ptest: improve ptest functionality on limited rootfs
ACL's ptest has a handful of failure modes which can be triggered by
a restrictive or small system. First, the ptest requires that daemon
be in the bin group, which run-ptest attempts to do using gpasswd,
but gpasswd is part of shadow, and oe-core removes shadow when it
doesn't think shadow will be needed. Even if, say, a package has
RDEPENDS on it. Whoops. So we manually sed the group file. This
will probably work.
Second, the filesystem used for the test has to support ACLs,
so we create a dummy ext3 filesystem and use that.
Third, the root/permissions test relies on the assumption that
"mkdir d" produces a directory which non-root users can access,
but in a secure product which defaults to umask 077, this doesn't
work. (That fix has been separately reported to upstream acl
through their bug report form.)
(This may prevent the test from running without mkfs.ext3, but it
allows the test to run on targets where root doesn't have ACL
support. Tradeoffs, tradeoffs everywhere.)
Signed-off-by: Peter Seebach <peter.seebach@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Herve Jourdain [Wed, 18 May 2016 12:33:37 +0000 (20:33 +0800)]
mesa: fix mesa_populate_packages() when dri is disabled
When compiling mesa, if dri is disabled in PACKAGECONFIG, or if the list of DRI
drivers is empty, it will cause populate_package to fail, because it can't find
- rightfully - the directory for the DRI drivers. This patch checks that the
directory indeed exists before trying to get a list of the files in it
[ use oe.path.join instead of + - RB ]
Signed-off-by: Herve Jourdain <herve.jourdain@neuf.fr> Signed-off-by: Ross Burton <ross.burton@intel.com>
Paul Eggleton [Mon, 9 May 2016 04:43:31 +0000 (16:43 +1200)]
classes/kernel: check OLDEST_KERNEL at configure time
If the kernel being built is older than OLDEST_KERNEL and we're building
with glibc, then the C library we're building is probably not going to
be compatible with the kernel and we should warn the user. (This is
easier to do here rather than when building glibc, because we don't
necessarily have the information we need to determine the kernel version
there, whereas we do here.)
Fixes [YOCTO #8653].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
* Blacklist SSTATE_DIR, DL_DIR and TMPDIR by default to avoid problems
such as the one mentioned in [YOCTO #9605].
* Blacklist BB_NUMBER_PARSE_THREADS since we already blacklist
BB_NUMBER_THREADS and they may be set separately.
Fixes [YOCTO #9605].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Marek Vasut [Sun, 8 May 2016 11:21:53 +0000 (13:21 +0200)]
kernel: fitimage: Repair misuse of shell test command
The kernel fitImage must be amended with signature if and only if
UBOOT_SIGN_ENABLE = 1 . In the current case, the UBOOT_SIGN_ENABLE
could be either 0 (default) or 1 , which test -n always correctly
interprets as non-empty string, thus always true. This does not
match the logic above though, so replace the test with check which
passes only for UBOOT_SIGN_ENABLE = 1 .
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Yannick Gicquel <yannick.gicquel@iot.bzh> Cc: Richard Purdie <richard.purdie@linuxfoundation.org> Signed-off-by: Ross Burton <ross.burton@intel.com>
Upstream in 5.33 btmgmt was moved from experimental to common READLINE section,
in commit e4f0c5582f1fe3451d5588243adba9de1ed68b80, but this was never updated
in the recipe.
Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org> Signed-off-by: Ross Burton <ross.burton@intel.com>
