Currently the overrides are being applied backwards. This means something which is
platform specific is overriding something which is machine specific which
is clearly not intended.
This patch corrects the ordering to match the normal expected behaviour of
OVERRIDES.
Secondly, all overrides are being searched for each path in turn. What should
really happen is that we should look for the highest priority override (e.g. distro
or machine) in each layer, then move on to platform/tune (e.g. armv7a) and
then to arch (e.g. arm). This patch therefore also reverses the for loops
to achieve this behaviour and give the result the user would expect.
(From OE-Core rev: 92cbf7eeea553bfa24c7081473fa8bc4ebc1f552)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Having a SRC_URI called ${PN} is asking for trouble. When extending FILESPATH,
alsa-state can be treated as a directory and copied over the contents of ${WORKDIR}
which is invariably not what the user wants.
Avoid this by renaming the SRC_URI to something else and only call
it alsa-state at install time.
(From OE-Core rev: 04c73333e4b539de96f096ca2954b2313175edc4)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Before checking if the image is a valid value, we need to check if the
list of items is big enough or we will raise an exception.
Reported-by: Lauren Post <lauren.post@freescale.com>
(From OE-Core rev: b833837cac377e7c1c3ff18531b152340b669329)
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
tar version 1.27 returns:
tar: --same-order option cannot be used with -c
with the commandlines we have been using. We can remove the -s option (which
is --same-order) to remove the error.
(From OE-Core rev: 69c26e795c117aabfaf313abbfd10e70ede633d9)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
tar version 1.27 returns:
tar: --same-order option cannot be used with -c
with the commandlines we have been using. We can remove the -s option (which
is --same-order) to remove the error.
(From OE-Core rev: 3d5a6d0a480a0fa98260a3b3ffc71b8d9e3e58af)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
when x11 is not defined in DISTRO_FEATURES, there are no x11 headers
so True is not defined leading to :
| ui/sdl.c:62:8: warning: type defaults to 'int' in declaration of 'doing_grabs' [-Wimplicit-int]
| static doing_grabs = True;
| ^
| ui/sdl.c:62:22: error: 'True' undeclared here (not in a function)
| static doing_grabs = True;
| ^
(From OE-Core rev: d3d5ccd3275196fb9e4cb8f49d3c087b97c20ed6)
Signed-off-by: Eric Bénard <eric@eukrea.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* it's autodetected and configure doesn't allow to explicitly disable it
(From OE-Core rev: 983446c4f4274b9fd482b2d6b711428f4483c02f)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* dbus-glib was missing in DEPENDS in order to enable dbus backend
but because dbus was in DEPENDS lets enable this PACKAGECONFIG by
default
(From OE-Core rev: c832c457ff40f0f65ab28c17b78b88cb79823db1)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
At the start of install, python swizzles the makefiles around, leading
to chunks of do_compile running again. These race against the install
target, leading to errors if pieces are being recompiled whilst others
are being installed.
For now, workaround this by running the compile target with the new
makefile, then running install ensuring a parallel make race doesn't
happen.
(From OE-Core rev: 72938a8631cfe5be5ac88ad67f2db595e2487e86)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The postinst files were being generated using purely the name of the package,
this unfortunately meant the run order would be based on the name of the
package and not the order in which it was installed on the filesystem.
If package A requires package Z to be fully installed, this causes a problem.
Note:
rpm - as the rpm based install proceeds the order is defined and captured.
so the problem is resolved there.
ipk - this unfortunately does not appear to solve the problem for ipk, as
the status file is not ordered in any appreciable way. This does not
cause any regressions however and sets the stage for a proper fix.
deb - this -may- fix the deb install. Early testing indicates at least some
ordering to the status file. But it's unclear if it completely resolves
the issue.
(From OE-Core rev: b5bafccb89f45d7cdd515b4ba45e0152ca7922de)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[YOCTO #5313]
When performing an attemptonly install, we should skip an errors and
continue to install everything else. However, there is a case where two
packages can conflict, and cause a hard failure.
This workaround, ignores this and allows the image to be constructed.
Note: Some items in the failed transaction may not get installed.
To fix this properly we need to find the issue in smart, and make it ignore
or at least attempt to resolve these kinds of conflicts.
(From OE-Core rev: 4228005689e31ebcafcf0969e80fcc021d1ae063)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In case of QEMU machines, a global public DNS is provisioned,
as the network is not configured via DHCP.
Google's public global DNS server 8.8.8.8 is used.
Partial fix for [YOCTO #4587] (qemu machines case).
(From OE-Core rev: 05ebf5bbacc58315f1b33ffb80148d568f88855d)
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When these functions change, the package should rebuild but currently it
does not. We need to add the dependencies manually as the dependency
code can't track dynamically created variables.
(From OE-Core rev: 4003218f774c38bb5de0d95a43153f8b8d7fc4ce)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When these functions change, the package should rebuild but currently it
does not. We need to add the dependencies manually as the dependency
code can't track dynamically created variables.
(From OE-Core rev: c5d7100a358244085a697a23790676df5eb4afa3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
OE-Core commit 75a14923da has moved
run-postinsts script execution from S98 to S99 in rcS.d. run-postinsts.service
should check for this script and run it on first boot rather than
S98run-postinsts, which is for opkg/dpkg.
(From OE-Core rev: fe039170236080291c0220476a5809774f82ee5c)
Signed-off-by: Muhammad Shakeel <muhammad_shakeel@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Make kmod-ptest able to compile with separated source and
build dir. Since kmod test files contain kernel modules for
many different architectures, strip and arch gets confused
and throws errors.
(From OE-Core rev: 61cb45869caaed6578a217effaa72d247395d078)
Signed-off-by: Tudor Florea <tudor.florea@enea.com>
Signed-off-by: Björn Stenberg <bjst@enea.com>
Signed-off-by: Maxin B. John <maxin.john@enea.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixes the following QA error:
ERROR: QA Issue: nativesdk-dbus: Files/directories were installed but not shipped
/run
/run/dbus
(From OE-Core rev: 796b7510853e71f158ad18dcea4cd1a10c7ef294)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We have a number (in fact most) BSPs that require INPUT_EVDEV for basic
functionality. The size increase is minimal, so we'll add it to the
standard configuration for all platforms.
(From OE-Core rev: 965affabbf36fa5b5e466d331479c2295269242e)
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
bb.utils.copyfile is for a specific purpose and more complicated than needed
here, so just use shutil.copyfile.
(From OE-Core rev: dc81df215cc94c279991df35125d94770a1bc3d2)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some packages may return error while running opkg-cl configure, during
first boot. This will fail 'ExecStart' and 'ExecStartPost' will not run.
Without 'ExecStartPost' opkg-configure service will continue to run on
successive boot attempts. 'ExecStartPost' should disable this service
after first boot irrespective of 'ExecStart' status.
(From OE-Core rev: cd6041071ddf76693cda7632379ceddd1d21a7fb)
Signed-off-by: Muhammad Shakeel <muhammad_shakeel@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The cross-canadian compilers are now build once per architecture but were being
installed into tune specific locations which is incorrect. This adjusts things
so they are make TARGET_ARCH specific. We gain the tune specific parts from the
target sysroot which remains tune specific, the compiler and tools are independent
ot that.
binutils/gcc require sysroot options but since we reset at runtime, these shouldn't
have dependencies in the sstate checksums. They are therefore also excluded.
With these patches, switching machines does not result in a rebuild of *-cross-canadian
and the compiler is correctly located and referenced in the target images.
(From OE-Core rev: f58acab6414fe96d9e07ebbe86b348d2ac2bed5f)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
MACHINE=X bitbake nativesdk-pixman; MACHINE=Y bitbake nativesdk-pixman
where X is an armv7a machine and Y is not results in pixman rebuilding
due to the neon option. This is incorrect. The ultimate fix will be
not to apply target system overrides in the nativesdk case. Until
we do that we can at least work around the issue as done by this patch.
(From OE-Core rev: 2881747ec270bffc039d10198399e2686407b521)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We only care about the end result in this case, not the specific inputs
that went into determining the gcc option. This change updates the code
to reflect that.
(From OE-Core rev: 83055511dab1e2aeefe84262ebe52595848c844b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If you build nativesdk for machine A, then change to B you will see sstate
manifest warnings for the packagedata files. The stamps are machine
specific and should not be, ditto for native.
This patch copies the populate-sysroot extra stamp entries to avoid
these warnings.
(From OE-Core rev: a64de25d6006ec6dd777d8f3820a48244dfbf62b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Instead of looping until the generated signature is non-zero, just
return 'ffffffff' if it's zero. This avoids an infinite loop if the
generated signature is always zero.
(From OE-Core rev: 34c74ba45c04efc2dd998545bf7a31d7b06a0ee0)
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>
Commit 609d5a9ab9e58bb1c2bcc2145399fbc8b701b85a added additional
initramfs functionality and created an additional task on kernel.bbclass
Adding this task was missed on the linux-dummy recipe, which causes task
dependency issues due to image.bbclass depending on
"virtual/kernel:do_bundle_initramfs". This change adds a dummy task
which resolves the dependency issue.
(From OE-Core rev: c0a8c5c07e0dd6f0ae302e9a4dcf7973e73e68e1)
Signed-off-by: Nathan Rossi <nathan.rossi@xilinx.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This ensures that makedevs will not cause image creation failures
when it encounters a pipe (fifo) that exists from a previous image.
This handles mode changes and it will correctly fail for dangling
symlinks.
[YOCTO #5288]
(From OE-Core rev: 3a4b0e7973bef43f16058137e64600e2f890b117)
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
dpkg uses xz as a compressor and the binary image is needed on the system for
dpkg to work correctly.
[YOCTO #1881]
(From OE-Core rev: 8f6bc04c9b563f8659ce0e053072deca02da38a5)
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There is a potential issue with the fastop code in pseudo since a process may
exit and allow some other function to run before the server has processed
the commands run by the process. Issues have been see with unpredictable
file permissions.
To avoid this, we ping the server before exitting which guarantees it has
processed the current command queue.
The patch was written by peter.seebach@windriver.com
[YOCTO #5132]
(From OE-Core rev: a8a1f12c51ffdca011db194894fd7d14c119fb09)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There was a bug in handling of the mxcsr register since cpu flags
were not getting updated after fxrstor operations. This small tweak
fixes that.
[YOCTO #5248]
(From OE-Core rev: 5dc43cdc08e6698afa16ba79f3506a1555bb3710)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Buffer used for copying a "%c" character was getting
out of scope when it was required by the sprintf operation.
[YOCTO #5272]
(From OE-Core rev: c7de71813c8f47438f44749136877442cf73d536)
Signed-off-by: Marius Avram <marius.avram@intel.com>
Signed-off-by: Irina Patru <irina.patru@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
exit 0 was done if $D != NULL, if one or more
shlibsign executions fails.
(From OE-Core rev: 5dc3eb72c4b9b68ab13310383a90fe7779bf92a7)
Signed-off-by: David Nyström <david.nystrom@enea.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
A zero MBR disk signature is generally seen as no signature and
another partitioning program might install a new signature.
(From OE-Core rev: b6cafb1fcd6c168f8f4a4d2d5c74f3b425f156f3)
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>
Activating the INITRAMFS_TASK can cause circular dependencies, but
that is up to the end user to resolve in recipes. The INITRAMFS_TASK
should also trigger immediate linking of the cpio task in the first
compile pass. This was a subtle regression introduced by: 609d5a9ab
(kernel.bbclass, image.bbclass: Implement kernel INITRAMFS dependency
and bundling).
This patch restores the previous behavior and only affects the
INITRAMFS_TASK which is not set by any of the default build profiles
in oe-core.
Reviewed-by: Andrea Adami <andrea.adami@gmail.com>
(From OE-Core rev: 36faac868e086e9c23537b107cdd973d7fd980bd)
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently opkg uses a script to configure packages during first time boot.
This script is present in rcS.d and when 'sysvinit' is disabled this
script doesn't execute. For systemd only distros this newly added service
will run the opkg configure during first boot only.
(From OE-Core rev: fdcfcea5b8eae1769a217f8efea9c02f037f63fa)
Signed-off-by: Muhammad Shakeel <muhammad_shakeel@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
