Robert Yang [Sat, 28 Dec 2013 03:29:24 +0000 (11:29 +0800)]
tremor: remove unrecognised option
Remove the --disable-rpath, it seems taht there is never such an option
in tremor's configure, and it uses our libtool, so we don't have to
worry about the rpath.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Saul Wold [Fri, 20 Dec 2013 18:47:26 +0000 (10:47 -0800)]
openssl: use PACKAGECONFIG to disable perl bits
Adding perl to the RDEPENDS caused a performance hit to the overall build time since this was
the only package that depended on perl. The openssl-misc package is not installed by default
so use a PACKAGECONFIG which can be overridden to allow the perl scripts along with perl to
be installed.
Signed-off-by: Saul Wold <sgw@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Alexandru DAMIAN [Wed, 18 Dec 2013 19:24:47 +0000 (19:24 +0000)]
toaster.bbclass: fix crash on buildstats collection
Toaster needs buildstats to be enabled in order to
collect task statistics. The toaster.bbclass didn't
have a guard to stop task data collection if the buildstats
weren't enabled, leading to a crash.
This patch verifies that the task variables are defined before
trying to use them.
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Cristiana Voicu [Fri, 20 Dec 2013 09:43:49 +0000 (11:43 +0200)]
valgrind: upgrade to 3.9.0
Changed patches directory to "valgrind";
Removed 4 patches because they were merged upstream;
Updated 2 patches because now it has configure.ac instead of .in;
Changed license md5 for 2 files because of a small change for copyright
year
Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Fri, 20 Dec 2013 14:55:09 +0000 (14:55 +0000)]
libav: add libpostproc to PROVIDES (for 0.8.x version only)
There is a separate libpostproc recipe in meta-oe for use with 9.x and
later versions of libav for those few that need libpostproc; however if
you just add meta-oe and try to build libpostproc without selecting the
libav 9.x version recipe, you'll be building the libpostproc recipe
together with libav 0.8.x, which provides its own libpostproc; this
leads to confusing errors at packaging time. In order to flag up that
these conflict more appropriately, add libpostproc to PROVIDES
explicitly so that you at least get a multiple providers error at the
start of the build.
Fixes [YOCTO #5335].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 20 Dec 2013 17:16:25 +0000 (17:16 +0000)]
terminal: Exclude BB_ORIGENV from the signatures
devshell was printing a traceback when exiting due to the use of dump_sigs()
being called on the task. This is turn was since this function referenced
BB_ORIGENV. We might as well globally exclude this for now since its a
data store object and cannot be pickled, not would it make sense to do so.
[YOCTO #5683]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 20 Dec 2013 12:08:56 +0000 (12:08 +0000)]
sstate/sstatesig: Add populate_lic to list of arch invariant sstate tasks
Like fetch, unpack and patch, populate_lic doesn't vary between different
archs so we should mark it as such. This means better sstate cache reuse
with fewer duplicate files as well as less confusing sstate debugging.
sstatesig also needs to account for the fact BPN is used for sstate files
in these cases.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 20 Dec 2013 12:06:41 +0000 (12:06 +0000)]
bitbake.conf: Exclude WORKDIR changes from sstate checksums
The layout of stamp files ensures that changes to WORKDIR mean recipes get rebuilt correctly.
Since WORKDIR usually contains MULTIMACH_TARGET_SYS and that depends on tune variables,
including WORKDIR in sstate checksums adds a lot of noise to the system for what amounts to
no gain.
On the other hand, removing it reduces noise, reduces the size of the siginfo files and
reduces the amount of processing bitbake has to do. It therefore seems like dropping it
from the checksums is an all around win.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 20 Dec 2013 12:06:02 +0000 (12:06 +0000)]
sstate: Convert to use ':' as a filename sperator and use SSTATE_SWSPEC globally
Currently the code has problems differentiating between "gcc-cross" and "gcc-cross-initial"
sstate files. We could add in a ton of special casing but tests show this isn't scaling
well. Using a more unique separator resolves the issue.
The choice of which separator to use is a hard one. We need something which isn't commonly
used in PN, PV, PR, *_OS and *_ARCH which rules out '-', '_' and it needs to work ok with
webservers/http which makes ';' and '%' harder.
The change also sets SSTATE_SWSPEC globally since writing out differently named siginfo
files for the fetch/unpack/patch tasks is a waste of diskspace, the hashes match for
all PN in the majority of cases and if they don't, its not a big issue as the hash is
different. This makes the results from sstate debugging more understandable.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Valentin Popa [Thu, 19 Dec 2013 14:02:57 +0000 (16:02 +0200)]
runqemu: Allow user to set -vga option with qemuparams
At the moment, the user cannot to set -vga other then vmware
(because "vmware" is set by default); and the first argument
in qemuparams has higher precedence.
Stefan Stanacar [Wed, 18 Dec 2013 19:02:05 +0000 (21:02 +0200)]
lib/oeqa/runtime: rewrite the systemd test module
They are basically the same tests but:
- they look cleaner, using one single method / assert
- output from unittest will be cleaner (and includes a verbose status when needed)
- they are better grouped and use a real, active, enabled service
(machineid will be dropped and hostnamed was a static service)
Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Cristian Iorga [Thu, 19 Dec 2013 09:34:52 +0000 (11:34 +0200)]
libtirpc: upgrade to 0.2.4
obsolete_automake_macros.patch removed,
no longer needed.
Disabled GSS APIs, as those APIs are
not present in our distro.
Patches directory is now generic,
as patches are not created for a specific version.
Richard Purdie [Thu, 19 Dec 2013 09:40:10 +0000 (09:40 +0000)]
sstatesig: Correctly handle matches spanning stamps and sstatedir
By resetting filesdates at this point, we lose matches from stamps which
may not have been in sstatedir. When we don't have hashes specicifed,
its better to return all matches and have the caller decide which are
relavent and which are not since this function has no ability to
decide. There will almost always be one match from stamps we need
to keep and refer to.
Richard Purdie [Thu, 19 Dec 2013 09:39:35 +0000 (09:39 +0000)]
sstatesig: Update for the removal of sstate-name
We've dropped sstate-name so we can remove this code. The fallback was
incorrect since we use taskname without the do_ prefix so this patch
updates to account for that too.
Richard Purdie [Thu, 19 Dec 2013 09:38:18 +0000 (09:38 +0000)]
sstatesig: Ensure we return all matches for find_sigdata
When the hashes to find isn't specified we need to return matches from both
the sstate cache and the local stamps directory regardless of how many we've
found so far. If we don't do this, we can miss stamps and the comparison is
less accurate/incorrect.
Richard Purdie [Wed, 18 Dec 2013 17:12:23 +0000 (17:12 +0000)]
ptest: Fix dependency/ordering issue
We've seen occasional issues on the autobuilder where files appear during
do_populate_sysroot and this is due to the ptest installation happening
in parallel. This fixes the order to be deterministic.
Ross Burton [Wed, 18 Dec 2013 16:20:07 +0000 (16:20 +0000)]
systemd: delete files that end up in /var/volatile
${localstatedir}/log/journal is created by tmpfiles.d on boot, and the README
just causes warnings that /var/volatile isn't empty when mounting on top of it.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 18 Dec 2013 16:20:06 +0000 (16:20 +0000)]
systemd: pass --root in prerm if running on host
If a systemd image is built without a package manager then packages will be
removed from an image during rootfs generation, but without passing --root the
systemctl will look on the *host* system.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 18 Dec 2013 16:20:02 +0000 (16:20 +0000)]
useradd.bbclass: add dependency on base-files
Packages that use useradd.bbclass should have a dependency on base-files so that
the /etc/skel directory is populated. Without this dependency base-files may or
may not be installed when the postinst runs, and the skel content may or may not
be copied.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 18 Dec 2013 13:56:02 +0000 (13:56 +0000)]
sstate: Get rid of crazy name mapping
When originally developed, it was thought a task may have more than one associated
sstate archive. The way the code has grown that idea is now not possible or needed.
We can therefore assume one sstate archive per task and drop the crazy name
mapping code. Simpler is better in this case.
The downside is that various sstate archives will change name so this forces a cache
rebuild. Given the other sstate changes going in at this time, this isn't really
a bad thing as things would rebuild anyway.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 18 Dec 2013 13:55:08 +0000 (13:55 +0000)]
sstate/gcc: Fix shared workdir handling for siginfo files
For a shared workdir, any one of the fetch/unpack/patch tasks may run yet the
PN and architecture fields in SSTATE_PKGSPEC may differ. This makes looking up
the appropriate siginfo file near impossible.
I've tried several different ways of resolving this and this is the neatest
solution I could find, its still rather ugly. I believe the usefulness of
better sstate debugging outweighs the ugliness of the code.
This patch also changes the sstate_checkhashes() code to look for siginfo
files rather than the actual sstate packages themselves. This means the
function can be used in other contexts to find info files for tasks that
may not have sstate data. It is assumed that sstate mirrors will have both
files available. This is done to allow bitbake to query whether tasks have
matching signatures in sstate directories or not.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 18 Dec 2013 13:49:23 +0000 (13:49 +0000)]
sstate: Write out siginfo files for non-sstate tasks
Currently siginfo files are only written for sstate tasks. In order to be truly
debuggable, its helpful to have the siginfo for intermediate tasks. This
adds that functionality so the extra siginfo files are written out too.
This will be used to add better sstate debugging in future changes.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
mykhani [Tue, 17 Dec 2013 08:28:35 +0000 (13:28 +0500)]
openssl.inc: Install c_rehash utility with openssl
c_rehash utility is not being installed with openssl.It conveniently
generates hash and symbolic links based on it for CA certificates
stored locally for SSL based server authentication
Signed-off-by: Yasir-Khan <yasir_khan@mentor.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Otavio Salvador [Tue, 17 Dec 2013 01:44:37 +0000 (23:44 -0200)]
pointercal-xinput: Mark as SIGGEN_EXCLUDERECIPES_ABISAFE
When including xinput-calibrator, in commit "xinput-calibrator: move
it from meta-oe to oe-core" the pointercal-xinput has not been added
to the SIGGEN_EXCLUDERECIPES_ABISAFE. This changes adds it to the
meta/conf/layer.conf's file list.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Mon, 16 Dec 2013 10:57:49 +0000 (10:57 +0000)]
scripts/contrib: Add graph-tool
A simple script I put together for getting the paths from one node to
another in a dot graph. This is useful for example in working out why
a particular recipe is getting built in conjunction with dot graph files
produced by bitbake -g.
Zhong Hongbo [Tue, 19 Nov 2013 09:19:00 +0000 (17:19 +0800)]
bluez4: use legacy pygobject instead of gobject-introspection
pygobject-3.0 use GObject from gi.repository instead of
gobject modules. Since oe-core use pygobject-2.x, change
all of "from gi.repository import GObject" to
"import gobject" for bluez4.
Kai Kang [Wed, 11 Dec 2013 09:05:53 +0000 (17:05 +0800)]
LSB_Test.sh: update rpm platform file when enable multilib
When enable multilib for x86_64, lsb rpm packages for archs x86_64 and
i486 will be installed. But i486 target packages fail to be installed
that no such pattern in rpm platform file. Update platform file to
enable install i486 rpm packages as well.
Signed-off-by: Kai Kang <kai.kang@windriver.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
commands install_initd and remove_initd should under directory /usr/lib/lsb/.
lsb recipe creates these commands under directory ${libdir} which may
expand to /usr/lib64 when multilib is enabled on qemux86-64. That will
cause LSB command check for install_initd and remove_initd fail. So
correct it.
Signed-off-by: Kai Kang <kai.kang@windriver.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Ross Burton [Tue, 10 Dec 2013 19:38:26 +0000 (19:38 +0000)]
insane: handle recursive configures when checking for unknown configure options
Some recipes have configure scripts that recursively call other configure
scripts (e.g. dropbear). These multiple-line matches were not being handled
correctly, so iterate over every matching line instead of assuming only one line
was found.
[ YOCTO #5646 ]
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As libdrm enables modules automatically, based on what it needs / can build,
it's unlikely we'll *ever* get a situation where all the modules are emitted,
so libdrm-drivers, which depends on all of them, ends up being useless and
non-installable. Make it use rrecommends, so it only pulls in what we have.
Signed-off-by: Christopher Larson <kergoth@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sun, 15 Dec 2013 16:30:29 +0000 (16:30 +0000)]
qemu/local.conf.sample: Add sdl PACKAGECONFIG
Currently the SDL configuration option for qemu floats. This is confusing to new users
and makes the build non-determinstic. This patch adds a PACKAGECONFIG option, defaulting
to off and adds documentation to local.conf.sample leaving it on by default since this
is the configuration our quick start assumes.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sun, 15 Dec 2013 16:33:47 +0000 (16:33 +0000)]
nativesdk: Set LIBCOVERRIDE to a value
nativesdk was just unsetting LIBCOVERRIDE however that causes some build failures
for xorg-libs which used a libc override. This adds in a mechanism to force
nativesdk to glibc and give the option of allowing another selection like
uclibc if anyone ever does the work to make it operational.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Chen Qi [Wed, 11 Dec 2013 05:41:36 +0000 (13:41 +0800)]
image.bbclass: fix for zap_root_password
Previously, this function replaces the root password with '*' if
'debug-tweaks' is not in IMAGE_FEATURES. It not only zaps empty root
password, but also zaps non-empty root password. That means, if the
user uses a bbappend file for base-passwd to set the root password, he
would not be able to login as root; if the user uses 'EXTRA_USERS_PARAMS'
to set the root password, he would still not be able to login as root.
What we really want from this function is to disallow empty root password
if 'debug-tweaks' is not in IMAGE_FEATURES. This function should not remove
non-empty root password because that password is usually deliberately set
by the user.
This patch renames zap_root_password to zap_empty_root_password to
better reflect the intent of this function. It also modifies the code
to make this function work correctly.
Chen Qi [Sat, 16 Nov 2013 07:27:47 +0000 (15:27 +0800)]
shadow-native: allow for setting password in clear text
Allow user to set password in clear text. This is convenient when
we're building out an image.
This feature is mainly used by useradd.bbclass and extrausers.bbclass.
This patch adds a new option '-P' to useradd, usermod, groupadd and groupmod
commands provided by shadow-native. The shadow package on target and in SDK
will not be affected.
Chen Qi [Thu, 12 Dec 2013 08:38:04 +0000 (16:38 +0800)]
init-ifupdown: adjust priority of networking to fix system freeze problem
On mpc8315e, a system freeze is encountered at system boot time if
connman and init-ifupdown are installed.
The error message before the freeze is:
"ip: RTNETLINK answers: File exists"
This problem is introduced by the following commit. dc80eea sysvinit: fix problem in switching runlevels
Part of the above commit is to make the networking init script run
at runlevel 2 3 4 5 instead of runlevel S. However, after the change,
networking is run after connman. And this causes the problem stated
above.
Make networking run first when entering runlevel 2 3 4 5, so that we
don't break things.
Martin Jansa [Fri, 13 Dec 2013 02:37:18 +0000 (03:37 +0100)]
ltp: Use patch generated without -M
* in my builds both versions worked, but Saul reported that it fails to
apply with:
Applying patch
0001-Rename-runtests_noltp.sh-script-so-have-unique-name.patch
patch: **** Only garbage was found in the patch input.
Now I've see the same issue on different builder (with Ubuntu 12.04).
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Cristian Iorga [Wed, 11 Dec 2013 17:24:01 +0000 (19:24 +0200)]
bluez5: upgrade to 5.12
- Configure options no longer available were removed.
- The content of patch directory was generic, so the name
of directory is now generic.
- Switched to xz format for tarball
(40% decrease in download size).
oe-selftest: Add track_for_cleanup method to be used in cleanup tasks
Added a track_for_cleanup(path) method that removes the given path in the
tearDown method. This mechanism can be used to make sure a file or directory
we created will be removed at the end of a test, regardless of what happens.
Add a new method that can be used by the tester to add a command to the
executed in the tearDown stage of the test. This mechanism can be used to
make sure certain test-specific cleanup tasks are done in the case of a
test failure.
Chen Qi [Wed, 11 Dec 2013 09:06:29 +0000 (17:06 +0800)]
libgcrypt: fix compile error with '-O2' in sysroot path
Characters like '-O2' or '-Ofast' will be replaced by '-O1' when
compiling cipher. If we are cross compiling libgcrypt and sysroot
contains such characters, we would get compile errors because the
sysroot path has been modified.
Fix this by adding whitespaces before and after the original matching
pattern in the sed command.
Koen Kooi [Wed, 11 Dec 2013 08:50:07 +0000 (09:50 +0100)]
distutils bbclass: delete site.py*
Multiple recipes package this generated file and will clash during installation with:
| Collected errors:
| * check_data_file_clashes: Package python-nose wants to install file /build/v2013.06/build/tmp-angstrom_v2013_06-eglibc/work/beaglebone-angstrom-linux-gnueabi/cloud9-gnome-image/1.0-r0/rootfs/usr/lib/python2.7/site-packages/site.pyc
| But that file is already provided by package * python-setuptools
| * check_data_file_clashes: Package python-nose wants to install file /build/v2013.06/build/tmp-angstrom_v2013_06-eglibc/work/beaglebone-angstrom-linux-gnueabi/cloud9-gnome-image/1.0-r0/rootfs/usr/lib/python2.7/site-packages/site.py
| But that file is already provided by package * python-setuptools
Bruce Ashfield [Thu, 5 Dec 2013 19:31:55 +0000 (14:31 -0500)]
linux-yocto-dev: allow static SRCREVs via include
The linux-yocto dev recipe is intended to track the latest revisions of the
kernel tree by default. To control revision churn, and integrate into a
regular build schedule having the ability to specify a set of SRCREVs without
modifying the recipe itself is desired.
So we introduce an optional include file, and variables that control whether
or not the static SRCREVs are used: USE_MACHINE_AUTOREV and USE_META_AUTOREV,
to add this flexibility.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Ross Burton [Wed, 11 Dec 2013 14:04:34 +0000 (14:04 +0000)]
site: add xorg_cv_malloc0_returns_null
util-macros 1.18 will look in the autoconf cache for this value so instead of
using overrides in xorg-lib-common, use the per-libc site cache.
Alert readers may notice that the site file claims that glibc returns NULL from
malloc(), when the previous change (e628c8aba0189de30de2833882b9999ff3b6547a)
claimed that it didn't. The previous change was incorrect, whilst malloc(0)
returns a valid pointer, realloc(p,0) does in fact return NULL, so the Xlib
wrapper functions are needed.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
