Robert Yang [Tue, 25 Feb 2014 09:44:05 +0000 (17:44 +0800)]
dpkg-compiler.m4: remove -Wvla (fix build on CentOS 5.8)
Remove the -Wvla flag from the set of compiler warning flags, since gcc
on old host systems such as CentOS 5.8 doesn't support it, and it
causes a build error for dpkg-native.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Donn Seeley [Tue, 25 Feb 2014 09:44:04 +0000 (17:44 +0800)]
dpkg: fix a link problem for dpkg-native on CentOS 5.8
[ CQID: WIND00392830 ]
CentOS 5.8 provides the kernel support and headers for the
sync_file_range() syscall, but glibc 2.5 doesn't implement the
sync_file_range() syscall stub, so we can't link dpkg-native. Add a
patch that makes dpkg require a glibc version >= 2.6 in order to use
sync_file_range().
Signed-off-by: Donn Seeley <donn.seeley@windriver.com> Signed-off-by: Lei Liu <lei.liu2@windriver.com> Signed-off-by: Randy MacLeod <Randy.MacLeod@windriver.com> Signed-off-by: Jeff Polk <jeff.polk@windriver.com> Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Robert Yang [Tue, 25 Feb 2014 09:44:03 +0000 (17:44 +0800)]
grub: exclude .pc from POTFILES.in (fix build on CentOS 5.8)
Exclude the .pc from po/POTFILES.in since quilt uses "patch --backup",
which will create the backup file under .pc, this may cause unexpected
errors, for example, on CentOS 5.x, if the backup file is null
(newfile), it's mode will be 000, then we will get errors when xgettext
try to read it.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Stefan Stanacar [Tue, 25 Feb 2014 10:46:11 +0000 (12:46 +0200)]
oeqa/targetcontrol: properly get the host ip
For the SimpleRemote target where we need the host ip and
it wasn't set in conf, we tried to determine it automatically.
However ip route output isn't the same for every network, we
need the last field from the first line.
Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Laurentiu Palcu [Tue, 25 Feb 2014 12:36:17 +0000 (14:36 +0200)]
package_manager.py: create index files for all backends in PACKAGE_CLASSES
The previous implementation was checking IMAGE_PKGTYPE and created the
index files just for the backend used to create the image. Apparently,
'bitbake package-index' should attempt to create the index files for all
backends specified in PACKAGE_CLASSES.
[YOCTO #5827]
Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If the output from opkg-query-helper.py is empty, output.split('\n')
would result in a list containing one element which is an empty string
while iterating over each line in the output. An exception is then
thrown by the line:
pkg, pkg_file, pkg_arch = line.split()
with the message:
Exception: ValueError: need more than 0 values to unpack
To avoid this, we add a condition to only split the output if it isn't
empty.
Signed-off-by: Jonathan Liu <net147@gmail.com> Signed-off-by: Saul Wold <sgw@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Mon, 24 Feb 2014 15:10:49 +0000 (15:10 +0000)]
sstate: Drop 'SafeDep' code from setscene validation function
I have a feeling this code exists from the time before we had proper
coverage of one sstate task by another task. At that time it was a
"poor" persons version of that idea, we now have much better
code internal to bitbake which handles this.
Worse, this code actually breaks certain rebuild scenarios,
e.g.:
Richard Purdie [Mon, 24 Feb 2014 16:12:07 +0000 (16:12 +0000)]
package: Drop do_package_write task
The reasons this task was introduced are lost in the mists of time. It
allowed for the a single "package_write" task instead of spelling out
the explicit package backends, however in all but one case we do that
anyway.
As such as might as well give in and delete the task, converting that
single reference into explicit dependencies.
This gives bitbake a bit less work to to when processing the runqueue
since there are less tasks (but more dependencies in some cases).
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Mon, 24 Feb 2014 16:05:46 +0000 (16:05 +0000)]
classes/utility-tasks: make do_listtasks a little more friendly
* Sort the list so it's at least in some form of logical order. I looked
at sorting by dependencies, but that's a topological sort, and given
no such function is shipped as part of the python standard libraries
it would seem excessive to pull one in just for this. In any case, I'm
not sure that for the data we have this would lead to any particularly
pleasing result.
* Show the doc values as defined in documentation.conf (where present)
as a description
Addresses [YOCTO #4856].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Mon, 24 Feb 2014 16:05:45 +0000 (16:05 +0000)]
classes/package_tar: fix conflicts with package_deb / package_ipk
Avoid tar noticing that the directory is changing when
do_package_write_deb or do_package_write_ipk are running at the same
time as do_package_write_tar (because DEBIAN and CONTROL are being added
and removed while tar is running so the directory changes).
Fixes [YOCTO #5652]
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Mon, 24 Feb 2014 16:05:44 +0000 (16:05 +0000)]
classes/sstate: fix taints being undone on execution of sstate tasks
The code here that deletes stamps was also deleting the taint files; so
forcing an sstate task with -f would force it to execute and then
because the taint file was deleted in the process, the next execution
would simply restore the output from sstate again. We need to exclude
the taint files just like we did in bb.build.make_stamp().
Fixes [YOCTO #5805].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Robert Yang [Mon, 24 Feb 2014 09:35:55 +0000 (04:35 -0500)]
lz4: fix CC
It defined the CC to ${TARGET_PREFIX}gcc which lost the --sysroot and was
incorrect, it would cause unexpected errors, we should define it as CC=${CC}.
[YOCTO #5869]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jonathan Liu [Fri, 21 Feb 2014 04:43:38 +0000 (15:43 +1100)]
image_types.bbclass: fix cpio IMAGE_CMD to preserve working directory
The working directory is changed in a subshell when executing cpio to
preserve the working directory for any subsequent commands. This is to
keep the working directory consistent when generating multiple image
types.
Signed-off-by: Jonathan Liu <net147@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Hongxu Jia [Fri, 21 Feb 2014 06:23:08 +0000 (14:23 +0800)]
rootfs.py: tweak _multilib_sanity_test for ipk incremental image generation
The _multilib_sanity_test installs multilib packages in a temporary
root fs, and compare with the current image to figure out duplicated
files that come from different packages.
While incremental image generation enabled and the previous image
was existed, there was an Multilib check error:
...
ERROR: Multilib check error: duplicate files tmp/work/qemux86_64-poky-
linux/core-image-minimal/1.0-r0/multilib/lib32/lib/libc.so.6 tmp/work/
qemux86_64-poky-linux/core-image-minimal/1.0-r0/rootfs/lib/libc.so.6
is not the same
...
The reason is the file in the existing image has been prelinked by
previous image generation and the file in a temporary root fs is not
prelinked, even though both of them came from the same package, the
Multilib check failed.
Hongxu Jia [Fri, 21 Feb 2014 06:21:36 +0000 (14:21 +0800)]
rootfs.py: support BAD_RECOMMENDATIONS for ipk incremental image generation
While incremental image generation enabled and the previous image is
existed, if BAD_RECOMMENDATIONS is changed, the operation on the
existing image is complicated, so remove the old image in this situation.
The same with PACKAGE_EXCLUDE and NO_RECOMMENDATIONS.
Hongxu Jia [Fri, 21 Feb 2014 06:12:52 +0000 (14:12 +0800)]
package_manager.py: support ipk incremental image generation
Add the following three functions to OpkgPM class:
- The 'dummy_install' is used to dummy install pkgs, and returns the log
of output;
- The 'backup_packaging_data' is used to back up the current opkg
database;
- The 'recover_packaging_data' is used to recover the opkg database
which backed up by the previous image creation;
Tweak 'remove' function in OpkgPM class, which the options for remove
with dependencies was incorrect.
Tweak 'handle_bad_recommendations' function in OpkgPM class:
- Fix none value check;
- Add the existance check of opkg status file;
Hongxu Jia [Fri, 21 Feb 2014 06:10:17 +0000 (14:10 +0800)]
manifest.py: add create_full for OpkgManifest class
The function create_full creates the manifest after the package in
initial manifest has been dummy installed. It lists all *to be
installed* packages. There is no real installation, just a test.
Saul Wold [Thu, 20 Feb 2014 20:55:09 +0000 (12:55 -0800)]
runqemu: Ensure ROOTFS path is absolute
There is a problem if a relative path is passed to the kernel for NFS usage
that it will not correctly find it, so ensure that the ROOTFS path is absolute.
Valentin Popa [Thu, 20 Feb 2014 10:09:36 +0000 (12:09 +0200)]
gtk+3: upgrade to 3.10.7
(*) removed patches that were already
commited upstream
(*) added new patch to solve the build issue
(*) changed the package URL because
the old one is unreliable.
Kai Kang [Thu, 20 Feb 2014 02:11:09 +0000 (10:11 +0800)]
update-rc.d.bbclass: fix inhibit check
In update-rc.d.bbclass it checks variable INHIBIT_UPDATERCD_BBCLASS to
inhibit from inheriting this class. But it is wrong logic that when
'sysvinit' is in 'DISTRO_FEATURES', INHIBIT_UPDATERCD_BBCLASS will not
be checked.
Replace 'or' with 'and' to fix it.
Signed-off-by: Kai Kang <kai.kang@windriver.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Dave Lerner [Sat, 15 Feb 2014 15:27:31 +0000 (09:27 -0600)]
valgrind: integration of regression tests to ptest
Modifies valgrind's regression test framework to be compatible
with the yocto PTEST framework as follows:
* existing recipe valgrind*bb adds new methods: do_compile_ptest and
do_install_ptest.
* new file run-ptest adds the wrapper interface to the valgrind
regression test script vg_regtest.
* existing valgrind regression test script 'vg_regtest' changes
to report the status of the valgrind component tests in the
format that PTEST expects, instead of the valgrind formats, but only
when vg_regtest is invoked with an optional --yocto-ptest argument
* four new patches disable building tests that don't compile with
the yocto compiler and default options. See the patches for details.
Signed-off-by: Dave Lerner <dave.lerner@windriver.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Tom Zanussi [Wed, 19 Feb 2014 20:53:37 +0000 (14:53 -0600)]
wic: Make exec_native_command() fail if a command isn't found
Because exec_cmd() return values can in certain cases be non-zero yet
non-fatal, we don't want to automatically make them fatal (though
there should at least be a warning in such cases, which this patch
also does); non-zero return values are definitely fatal however if
they mean that a native command wasn't found, so have
exec_native_cmd() check the return value of exec_cmd() for that case,
and bail out if so.
[YOCTO #5835]
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Changes:
- rename DESCRIPTION with length < 80 to (non present tag) SUMMARY
- drop final point character at the end of SUMMARY string
- remove trailing whitespace of SUMMARY line
Paul Barker [Wed, 19 Feb 2014 15:15:17 +0000 (15:15 +0000)]
opkg: Fix add-exclude.patch
The case statement for ARGS_OPT_ADD_EXCLUDE added to the argument handling
switch statement in opkg was missing a "break;" at the end, so it was falling
through into the handler for ARGS_OPT_NOACTION. Thus when "--add-exclude" was
specified on the command line it was as if "--noaction" was also being
specified. This appears to be the root cause of YP bug 5311.
Tested using the case described by Alexandru Georgescu in YP bug 5311:
Cristian Iorga [Fri, 14 Feb 2014 13:36:27 +0000 (15:36 +0200)]
libvorbis: upgrade to 1.3.4
- Updated website;
- Cleaned up bugtracker web address;
- Removed PR;
- Switched to ${BP} variable;
- removed obsolete_automake_macros.patch,
included in upstream;
Laurentiu Palcu [Tue, 18 Feb 2014 13:49:44 +0000 (15:49 +0200)]
lib/oe/image.py: add image dependency mechanism
This commit adds a dependency mechanism to image creation, so that we can
split the images creation execution in groups, that can be executed in
parallel, having the dependencies satisfied in the same time. The old
code didn't need this since everything was serialized.
Technically, it adds a dependency graph topological sort class that the
main Image class can use to sort out the dependencies.
Images that have dependencies have to declare them using the NEW
IMAGE_TYPEDEP variable, like in the example below:
Laurentiu Palcu [Tue, 18 Feb 2014 13:49:43 +0000 (15:49 +0200)]
image*.bbclass, bootimg.bbclass: add image type dependencies
The following dependencies were manually added in the image creation
code. However, in order to have an image dependency mechanism in place,
use a new variable, IMAGE_TYPEDEP, to declare that an image type depends
on another being already created.
The following dependencies are added by this commit:
elf -> cpio.gz
live -> ext3
vmdk -> ext3
iso -> ext3
hddimg -> ext3
This commit adds also another new variable: IMAGE_TYPES_MASKED. Currently,
masking out certain types from IMAGE_FSTYPES was hardcoded in the image
creation code.
Ross Burton [Tue, 18 Feb 2014 12:13:01 +0000 (12:13 +0000)]
avahi: handle SO_REUSEPORT not being available
Linux < 3.9 doesn't have the SO_REUSEPORT option so instead of failing to start
when built with >=3.9 kernel headers but booted on <3.9 kernels, continue as if
SO_REUSEPORT wasn't available.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Martin Jansa [Tue, 18 Feb 2014 11:05:03 +0000 (12:05 +0100)]
libxkbcommon: Add PACKAGECONFIG for x11
* allow to explicitly disable x11 with --disable-x11, otherwise
do_configure fails for DISTROs without x11 in DISTRO_FEATURES:
| No package 'xcb-xkb' found
| configure:18763: $? = 1
| configure:18777: result: no
| No package 'xcb' found
| No package 'xcb-xkb' found
| configure:18793: error: xkbcommon-x11 requires xcb-xkb >= 1.10 which
was not found. You can disable X11 support with --disable-x11.
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Matthieu Crapet [Mon, 17 Feb 2014 16:28:58 +0000 (17:28 +0100)]
image_types: minor, inline CPIO_TOUCH_INIT()
Since generic compression solution has been introduced
(see revision b7e4ed41ee480f00b7265341e9e2d2c2b9135143),
CPIO_TOUCH_INIT() is only called by IMAGE_CMD_cpio.
diffconfig() is a new task that makes a diff between the
old and new config files and writes to the fragment.cfg result file.
menuconfig() always copy the original config file, so the user
doesn't need to copy it.
Signed-off-by: João Henrique Ferreira de Freitas <joaohf@gmail.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Instead of using 'diff' command between two kernel config files,
the task diffconfig does the job creating the file
$WORKDIR/fragment.cfg that user should review and use.
[YOCTO #3862]
Signed-off-by: João Henrique Ferreira de Freitas <joaohf@gmail.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Steffen Sledz [Tue, 11 Feb 2014 08:59:31 +0000 (09:59 +0100)]
lighttpd: introduce /etc/lighttpd.d subdir for config file includes
Extend default config file by a directive to include config file
fragments from /etc/lighttpd.d. This allows other web application
packages to put their configuration there.
The working directory needs to be changed before the image creation
commands instead of afterwards.
exposed an issue when generating tar images, where it would cd into
rootfs and then create the rootfs tarball and remain there since the
cmd to cd into deploy dir moved up the immediate following cmd like
tar.bz2 or tar.gz were still looking for tarball in current directory
which esentially was not deploy dir but IMAGE_ROOTFS instead
Khem Raj [Sun, 16 Feb 2014 19:48:16 +0000 (19:48 +0000)]
binutils: ld-is-gold should not affect native and crosssdk recipes
gold linker does not support all architectures, currently arm and x86
are best supported, therefore we can not enable this as a full distro
option where we need to support other architectures e.g. ppc, mips
currently, if we enable ld-is-gold distro feature conditionally then it
invalidates native and sdk version of native binutils because configure
option would change. With this patch we limit ld-is-gold feature to
imapact cross binutils and target binutils only. This means that we
can reuse the sdk and native versions across architectures.
Khem Raj [Sat, 15 Feb 2014 22:04:05 +0000 (22:04 +0000)]
icu: Disable the default LDFLAGSICUDT for target compile
By default LDFLAGSICUDT=-nodefaultlibs -nostdlib for Linux
which means DT_NEEDED section for libicu will not be populated
with dependencies when we reset it to be empty then the default
libraries it needs are added to DT_NEEDED section in ELF header
This patch is a workaround as I consider it for now, since it could
be a problem in glibc dynamic loader for arm (especially for hf case)
where its unable to load shared objects which dont have any dependencies
expressed in DT_NEEDED segment.
here is when LDFLAGSICUDT=-nodefaultlibs -nostdlib
btw. ldd reveals the fist one to be static library while the second
one is detected as shared library. This could be a clue into how
elf headers are being interpreted by dynamic loader.
The data seems to be all static in libicudata which could
load it quicker and thats what could be confusing dynamic linker ..may
be
Khem Raj [Sat, 15 Feb 2014 07:19:27 +0000 (23:19 -0800)]
rootfs.py: Check for LDCONFIGDEPEND being empty string
We override LDCONFIGDEPEND to be empty string for uclibc
however the current check is for it being None as a result
the function is still executed but ldconfig-native is not
built as dependency for rootfs when building with uclibc
Fixes errors like below
File:
'/home/kraj/work/angstrom-repo/sources/openembedded-core/meta/lib/oe/rootfs.py',
lineno: 191, function: _run_ldconfig
0187: def _run_ldconfig(self):
0188: if self.d.getVar('LDCONFIGDEPEND', True) is not None:
0189: bb.note("Executing: ldconfig -r" +
self.image_rootfs + "-c new -v")
0190: self._exec_shell_cmd(['ldconfig', '-r',
self.image_rootfs, '-c',
*** 0191: 'new', '-v'])
This change appears to cause more problems than it fixes since the
compression commands usually work in the deploy dir but the archive ones
have always worked in the rootfs dir (which is clear from the tar command
we use).
Koen Kooi [Sat, 15 Feb 2014 15:28:55 +0000 (16:28 +0100)]
bitbake-prserv-tool: make help text show .conf suffix
'export' will accept any output filename, but 'import' needs a '.conf'
suffix to work. Otherwise you'll get:
koen@beast:/build/v2013.12$ bitbake-prserv-tool import x.txt
ERROR: Traceback (most recent call last):
File "/build/v2013.12/sources/bitbake/lib/bb/cookerdata.py", line 162,
in wrapped
return func(fn, *args)
File "/build/v2013.12/sources/bitbake/lib/bb/cookerdata.py", line 172,
in parse_config_file
return bb.parse.handle(fn, data, include)
File "/build/v2013.12/sources/bitbake/lib/bb/parse/__init__.py", line
100, in handle
raise ParseError("not a BitBake file", fn)
ParseError: ParseError in x.txt: not a BitBake file
ERROR: Unable to parse x.txt: ParseError in x.txt: not a BitBake file
Importing from file x.txt failed!
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Scott Garman [Sat, 15 Feb 2014 19:04:47 +0000 (11:04 -0800)]
runqemu: add ability to skip using an existing tap device
Support the sitauation where a user could have another VM running
which uses tap devices. To prevent runqemu from trying to use the
same tap device, runqemu will skip using a tap device if it finds
a filename tapX.skip within its lock directory.
This fixes [YOCTO #5815]
Signed-off-by: Scott Garman <scott.a.garman@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Khem Raj [Sun, 16 Feb 2014 01:45:25 +0000 (01:45 +0000)]
openssl: Fix build on uclibc
we need to map OS string correctly to include linux-uclibcspe
which is what we use with ppc+spe on uclibc, additionally move
gnuspe triplet mapping to same code as well
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ming Liu [Mon, 17 Feb 2014 08:48:15 +0000 (16:48 +0800)]
rpm: fix a endian incompatible error in generating tag
A flaw was found in the way rpm generating arbitrary tags, which leads to a
incorrect query result, this issue is introduced by a incompatible endianess
when the generating process is executed on different architectures.
This patch resolves it by taking the byte order that host uses.
Signed-off-by: Ming Liu <ming.liu@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We don't need to depend on do_package anymore after the split out of
do_packagedata (which happened a while back in OE-Core commit 6107ee294afde395e39d084c33e8e94013c625a9). We do need dependencies on
do_packagedata though since the SDK construction needs to make use of
the pkgdata files.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Mon, 17 Feb 2014 14:22:29 +0000 (14:22 +0000)]
classes/populate_sdk_base: fix race condition with do_rootfs
do_rootfs has ${S} in cleandirs, and during do_populate_sdk we call
exec_func() several times, which by default uses ${B} as the working
directory. If do_populate_sysroot and do_rootfs race against eachother,
the directory may not exist at the exact instant that the setup code
for do_populate_sdk tries to cd into it. We don't actually use ${B}
for do_populate_sysroot so we can set it to something else just for that
task to avoid the race.
NOTE: because this task name contains an underscore, the override will
not work; the BitBake patch that changes these to hyphens for the
task override is required for this patch to work (but won't break things
without it.)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Mon, 17 Feb 2014 14:22:28 +0000 (14:22 +0000)]
eglibc-locale: add eglibc-gconv to PACKAGES_DYNAMIC
If you try to add eglibc-gconv to an image, you'll get a NoProvider
error because nothing at parse time states that it provides this (it's
added to PACKAGES in populate_packages). This problem manifests itself
in Hob for example when customising an image.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Valentin Popa [Thu, 13 Feb 2014 13:45:20 +0000 (15:45 +0200)]
weston: upgrade to 1.4.0
(*) added fbdev backend in case the user will
want to use weston even if drm fails.
(*) removed tablet-shell flag because in 1.4
tablet-shell has been removed from src.
(*) tested on qemux86 and genericx86 with
core-image-weston and core-image-sato.
Josua Mayer [Fri, 14 Feb 2014 15:02:24 +0000 (16:02 +0100)]
package.bbclass: do_split_packages should always return something.
So far, when do_split_packages is passed a non-existing folder, it will return nothing.
While building Mesa with PACKAGECONFIG="egl gles" for qemux86, do_package threw a nice exception on a line reading "pkgs += do_split_packages":
"Exception: TypeError: 'NoneType' object is not iterable"
I think a function should always return something, and in this specific case, returning an empty list should be right.
This patch simply fixes the return statement to do just that.
Paul Eggleton [Tue, 11 Feb 2014 18:00:15 +0000 (18:00 +0000)]
classes/cpan-base: fix signatures changing when perl is in sysroot
The previous fix for this same issue (OE-Core rev f31f6a70ec24e8c9515d69c5092e15effc5e7d4d) was not sufficient - we are
setting the PERLVERSION variable from the get_perl_version function, but
we're setting it using immediate expansion; thus the value is going into
the signature and is still different between the time the recipe is
cached on an empty TMPDIR and after perl is in the sysroot and we run
bitbake -S perf. We could remove the immediate expansion, but that would
mean the get_perl_version function would be called more often, so just set
vardepvalue on the PERLVERSION variable to fix the problem instead.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Saul Wold <sgw@linux.intel.com>
Richard Purdie [Sun, 16 Feb 2014 16:29:25 +0000 (16:29 +0000)]
lib/oe/package_manager: Make empty package directories non-fatal
Its perfectly reasonable to call -c package-index with empty package
directories. This ensures the builds don't fail in such cases,
resolving autobuilder failures.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
