Now that image generation consists of separate tasks, we can improve
the dependencies in these classes. We can also incorporate the tasks
within do_image_complete instead of do_build so we have a specific task
for when the deploy directory is populated with images.
This fixes various expectations about do_image_completed and the way
I'd tried to use it to fix some sanity test problems.
(From OE-Core rev: 9543c4a9ce0ac7d9ced66fa14e48a1aa7401011e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The wic command can be used externally but for this to work, the wic
environment file needs to be present. Therefore write this out
universally, it runs in parallel with other image construction so
any performance implications are negligible.
(From OE-Core rev: b2576f2eab10e4c5dd86449312b417a269cc578e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When we add buildhistory, we expect do_rootfs to rerun, but depending
on IMAGE_FSTYPES, the number of tasks which would execute after
do_rootfs varies (e.g. live would add do_bootimg and we recently
added do_image).
Therefore limit the test to -c rootfs and then we're clear that only
one task should re-run.
(From OE-Core rev: 9e9a33281e6008c3b699b5bbb738921829dbc501)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently, iso and hddimg links don't respect RM_OLD_IMAGE. This
updates them to use the common symlinks code so that they behave
like the rest of the system.
(From OE-Core rev: 6a05cb64dfafd531d50454ef7141ff0290d01ca9)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When RM_OLD_IMAGE is set, old manifest files should be removed
along with old image files and our QA tests expect this.
This patch ensures this happens.
(From OE-Core rev: fc951851a0e172641e2f24a4edce87ca2eb5cecd)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Similarly to DATETIME, don't expand TMPDIR in image commands.
This ensures some of the stamp comparisons we make in the
QA tests work correctly.
(From OE-Core rev: a8c377beadb85b0ff503ec8ddd1a2cd05e363c19)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The do_image_* tasks contained the expanded version of DATETIME. Due
to the expansion, we couldn't exclude the value from the task checksum
which meant the task would rerun.
We fix this by deleting the DATETIME value during expansion so we don't
expand any references to at that time. This means the task's hash can be
stable rather than having hardcoded date/time values. It will get expanded
at execution time.
This also fixes errors shown by -S:
NOTE: Reparsing files to collect dependency data
Writing locked sigs to /home/pokybuild/yocto-autobuilder/yocto-worker/nightly-oe-selftest/build/build/locked-sigs.inc
ERROR: Bitbake's cached basehash does not match the one we just generated (/home/pokybuild/yocto-autobuilder/yocto-worker/nightly-oe-selftest/build/meta/recipes-sato/images/core-image-sato.bb.do_image_tar)!
ERROR: The mismatched hashes were 77872792556367f1dde49a1425caf1a0 and 9bb0aca6286ab7dd22d3c69964beb665
(From OE-Core rev: ecbc1db7ed1f9848dee69507de8eb289b8ddeba0)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The wic environment function needs to run after the rootfs size is
setup. We move this code to a specific task, and depend on that task
from the wic images and other places its needed.
This fixes:
======================================================================
FAIL: test_image_env (oeqa.selftest.wic.Wic)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/pokybuild/yocto-autobuilder/yocto-worker/nightly-oe-selftest/build/meta/lib/oeqa/utils/decorators.py", line 106, in wrapped_f
return func(*args, **kwargs)
File "/home/pokybuild/yocto-autobuilder/yocto-worker/nightly-oe-selftest/build/meta/lib/oeqa/selftest/wic.py", line 218, in test_image_env
self.assertTrue(var in content, "%s is not in .env file" % var)
AssertionError: False is not true : ROOTFS_SIZE is not in .env file
(From OE-Core rev: 606f9e2d7d8d389c8d4f5c3090139d3bb780e09c)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This cleans up the do_configure_qa code so that the exit status from
package_qa_handle_error is handled correctly.
It also converts package_qa_check_license to use our standard QA check
configuration interface through package_qa_handle_error rather than
a mix of return values, bb.fatal and bb.build.FuncFailed.
Due to the merging of multiple lines into one message, we need to
tweak a QA test to account for this.
(From OE-Core rev: fbdf977a8094bb1e0baca85ffb3ed7f326986639)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Whilst in theory return values are nice, handling the complex
aggregation of the return values is ugly. The code already uses QA_SANE
as a marker to signal problems. Whilst that isn't as idealistic, it
makes the code massively cleaner, so rely on this instead.
(From OE-Core rev: ebee3caeaa18daf4974ea90e264d64467681f57a)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We had upgraded e2fsprogs from 1.42.9 to 1.43 (WIP), and used:
PV = "1.42+1.43-git${SRCPV}"
But:
$ dpkg --compare-versions 1.42+1.43 gt 1.42.9 && echo greater || echo less
less
It shows that 1.42+1.43 is less than 1.42.9, so this is a downgraded. Use
PV = "1.42.13+git${SRCPV}"
to fix the problem. 1.42.13 is the latest e2fsprogs 1.42 version.
(From OE-Core rev: aafe22fb5f2f58e1e7206a34f410e9cae0d3cfe3)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixed:
e2fsprogs-1.42+1.43-gitAUTOINC+0f26747167: e2fsprogs rdepends on fuse, but it isn't a build dependency? [build-deps]
The fuse is in meta-filesystems layer.
(From OE-Core rev: 4612448ee7e26601694e4ab3d41a92982712c8a5)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The license manifest is wrong when there is no spaces before and after separates |&()
and we can get warning like:
WARNING: The license listed Artistic-1.0GPLv1+ was not in the licenses collected for recipe
(From OE-Core rev: 20d6a4ce4e395251604d92111ce5ecacde5ce172)
Signed-off-by: Jackie Huang <jackie.huang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
It uses command glib-genmarshal to build glib-2.0 which is from
glib-2.0-native. If the command doesn't exist on host, build fails
with error:
| /bin/sh: --prefix=test_marshal: command not found
| make[4]: *** [testmarshal.c] Error 127
So add the dependency back.
(From OE-Core rev: 86136f6013b216d8ee258783e5612c3cd0043a58)
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Move do_kernel_link_vmlinux() from kernel-yocto.bbclass into
kernel.bbclass so that it's available to any kernel recipe.
Note that the task is not enabled by default in kernel-yocto.bbclass,
so don't enable by default in kernel.bbclass either. To enable, see
the example in linux-yocto.inc, ie:
addtask kernel_link_vmlinux after do_compile before do_install
(From OE-Core rev: a29371848deda618a42f9a30f9856a44c2342fe6)
Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The do_kernel_link_vmlinux() task modifies the build directory (not
the source tree) and should not be skipped when externalsrc is being
used.
(From OE-Core rev: 70b8a60d42831c701d10066eb57aaaad7a6fb7ae)
Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some products might need to use a tar replacement that
1. supports xattrs and
2. has more permissive license than GNU tar.
And the bsdtar binary produced from libarchive meets these
requirements.
(From OE-Core rev: ebf68926139ae5374c4d2669930d7f3817ad8f5a)
Signed-off-by: Dmitry Rozhkov <dmitry.rozhkov@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
SDK artifacts were being dumped alongside the image file dump.
However, the dump was not being triggered correctly, as it
should be attached as a postfunc to do_populate_sdk, rather than
do_rootfs, as do_rootfs doesn't always fire when populating
the SDK directory.
By adding a postfunc to do_populate_sdk, the postfunc is fired
in the correct situations.
Also fixes the directory which is scanned for SDK artifacts, as
this is different from the image file directory.
[YOCTO #7603]
(From OE-Core rev: cd2eb45ce1f9be15e133d6fe940c1cff2199099b)
Signed-off-by: Elliot Smith <elliot.smith@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
- Support autotools:
*) Touch NEWS/AUTHORS/ChangeLog to workaroud failure
caused by 'automake --add-missing'
*) Add patch to fix configure failure caused by missing
macro definition in m4 directory
- Add PACKAGECONFIG flags for libcap, libseccomp and pyelftools,
and disable them by default.
- Add PACKAGECONFIG flag for largefile.
- Since we add PACKAGECONFIG flags for pyelftools, we do not need
python in RDEPENDS.
BTW:
Currently oe-core does not support libseccomp.
(https://github.com/seccomp/libseccomp)
Currently oe-core does not support pyelftools.
(https://github.com/eliben/pyelftools)
(From OE-Core rev: d966a4ed8f0de7dc9b08880505c668421e6d4488)
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
3.9.2 -> 3.10.0
1. Fixed a parallel build problem
2. Added a PACKAGECONFIG for readline
(From OE-Core rev: ed46ac71846845ffc41efcd55a36e27883d9980a)
Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Installing python3-modules should install *all* of the Python standard library,
not just most of it.
(From OE-Core rev: b2c1c48d90aa9a94f90836557856e0319ddcb524)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This removes unnecessary spaces inserted before semicolons in the
modified USERADD_PARAM_${PN} and GROUPADD_PARAM_${PN} variables. This
should not affect the handling of the variables as the only one that
actually sees the semicolons is the code in useradd.bbclass that uses
cut to split the variables at them, and any whitespace preceeding or
following the semicolons will be properly ignored.
(From OE-Core rev: acc17ef91a6f506e3cacdc0d4ebfa268b3f3affd)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Read and merge the passwd/group files before parsing the user and
group definitions. This means they will only be read once per
recipe. This solves a problem where if a user was definied in multiple
files, it could generate group definitions for groups that should not
be created. E.g., if the first passwd file read defines a user as:
foobar:🔢:::
and the second passwd file defines it as:
foobar:::nogroup:The foobar user:/:/bin/sh
then a foobar group would be created even if the user will use the
nogroup as its primary group.
(From OE-Core rev: 3149319ad997379a01d87f2b1b5d14f66541997f)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The original code was near impossible to follow, and missed a couple
of cases. For example, if one added the following line to the passwd
file specified in USERADD_UID_TABLES:
foobar❌12345:nogroup::/:/bin/sh
and then specified the user as:
USERADD_PARAM_${PN} = "--system foobar"
one would then assume that the foobar user would be created with the
primary group set to nogroup. However, it was not (the primary group
would be foobar), and the only way to get it correct was to explicitly
add --gid nogroup to the USERADD_PARAM_${PN}.
(From OE-Core rev: e98e98ce65cf9ffdef29462310a622ffddd1412b)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The [<on_true>, <on_false>][not <condition>] construct may solve the
problem of implementing a conditional operator, but it is not very
readable. At least I find this:
uaargs.groupid = field[3] or uaargs.gid or uaargs.groupname
a lot more readable than this:
uaargs.groupid = [uaargs.gid, uaargs.groupname][not uaargs.gid]
uaargs.groupid = [field[3], uaargs.groupid][not field[3]]
Also, the official conditional operator since Python 2.5 (<on_true> if
<condition> else <on_false>) does not evaluate both <on_false> and
<on_true> as [<on_true>, <on_false>][not <condition>] does.
(From OE-Core rev: 274d38a6e2183ec88335a08b963f26c34b328558)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If --no-user-group is specified in USERADD_PARAM_${PN} for a user and
no --gid is specified, then we should not assume that the group name
for the user is the user name.
(From OE-Core rev: 0e5402c433739a416a76df532378533cb25365c7)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The useradd options --create-home/--no-create-home and
--user-group/--no-user-group are mutually exclusive and should be
treated as such.
(From OE-Core rev: 908eca7fb4af8a60026f53e2bb2cf1d5daf089ab)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changelog: http://alsa-project.org/main/index.php/Changes_v1.0.29_v1.1.0
The most significant change is probably the addition of the "topology
API", which is used to change the DSP topology from userspace. My
understanding is that the API is (or will be) needed to support some
new hardware, but I don't know any specifics about that.
Rebased Check-if-wordexp-function-is-supported.patch.
Dropped 0001-build-Do-not-try-to-detect-cross-compiler.patch, since it
is included in the new release.
(From OE-Core rev: 13a2786493bd91a54e3a5342da11c06a9b6c99b8)
Signed-off-by: Tanu Kaskinen <tanuk@iki.fi>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This commit in gtk+ moved the cache files below <libdir>
"
commit c8849046860a9b17fa943247d85ddadb29262b48
Author: Matthias Clasen <mclasen@redhat.com>
Date: Thu Jul 4 09:27:17 2013 -0400
Move the module cache files below libdir
These files contain architecture-dependent paths, and thus placing
them into sysconfdir causes unnecessary hassle. Now the immodule cache
file is looked for in libdir/gtk-2.0/2.10.0/immodules.cache.
Belated backport of a change that was done in the run-up to 3.0.
"
Since gtk-immodules-cache.bbclass creates the cache file in
"/etc/gtk-2.0/gtk.immodules", the environment variable "GTK_IM_MODULE_FILE"
can be set to point GTK+ at the file.
[YOCTO #6774]
(From OE-Core rev: d83d6ef50e6736d3c5f75050a01d5811a2a668f7)
Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License checksum changed because text layout changed.
(From OE-Core rev: 243ccaceee9bab9c8f1021a1502c07e1055614e6)
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License checksum changed because of text layout change.
(From OE-Core rev: 5afb75e552573dda6770c59b5eed2d1be0bb26f7)
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The 4.4 kernel has been under test via linux-yocto-dev, and is the next
LTS/LTSI kernel.
So we take our -dev support and created a named/versioned recipe for use
in upcoming releases.
(From OE-Core rev: 2dbc788bc3988532f0932e8e51e24ba702c28f6a)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Upgrade to 1.43 (master) to make "mke2fs -d" support xattr, so that the
layer which requires xattr such as meta-selinux can populate images
easily.
* Remove the following patches since they are alredy in the source.
0001-e2fsprogs-fix-cross-compilation-problem.patch
0001-libext2fs-fix-potential-buffer-overflow-in-closefs.patch
0001-mke2fs-add-the-ability-to-copy-files-from-a-given-di.patch
0002-misc-create_inode.c-copy-files-recursively.patch
0003-misc-create_inode.c-create-special-file.patch
0004-misc-create_inode.c-create-symlink.patch
0005-misc-create_inode.c-copy-regular-file.patch
0006-misc-create_inode.c-create-directory.patch
0007-misc-create_inode.c-set-owner-mode-time-for-the-inod.patch
0008-mke2fs.c-add-an-option-d-root-directory.patch
0009-misc-create_inode.c-handle-hardlinks.patch
0010-debugfs-use-the-functions-in-misc-create_inode.c.patch
0011-mke2fs.8.in-update-the-manual-for-the-d-option.patch
0012-Fix-musl-build-failures.patch
CVE-2015-0247.patch
copy-in-create-hardlinks-with-the-correct-directory-.patch
fix-icache.patch
misc-mke2fs.c-return-error-when-failed-to-populate-fs.patch
* Remove cache_inode.patch since it is not needed any more
* Updated mkdir.patch and ptest.patch
* Add --enable-libblkid to EXTRA_OECONF since libblkid is not created by
default.
* Time of core-image-sato-sdk do_rootfs:
- Before upgrade
real 3m18.508s
user 7m42.088s
sys 1m1.984s
- After upgrade
real 3m21.552s
user 7m38.496s
sys 1m0.644s
The are nearly the same
* The "fsck -f" shows the image is OK, and also can boot.
[YOCTO #8622]
(From OE-Core rev: a1f235ad736d322bb50eb4a4293b6b2f4e5200aa)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The code in native.bbclass adds -native suffix to the package
names that don't have it.
Renamed python-native-runtime -> hostpython-runtime-native to avoid
mangling it and to conform with the naming convetion for native
packages.
(From OE-Core rev: 8a474057d86b3ebf6271656d6b9adf384ea9ad6d)
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When set PATCHTOOL = "patch" in conf/local.conf an error will occur
in the step of do_patch.
The error is: "quilt: command not found", so fix the missing
dependency.
(From OE-Core rev: 3cc6b73f9feaa1f09edd649060e63f5143d7cd15)
Signed-off-by: Li Xin <lixin.fnst@cn.fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
To be compatible with ncurses 5, add option '--with-abi-version=5' to EXTRA_OECONF
(From OE-Core rev: 743e97cc80744ca88d72b8db0ee95bb58f350ef9)
Signed-off-by: Dengke Du <dengke.du@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The absolute path (/path/to/configure) caused VPATH in Makefile to be an
absolute path, and then it will be in elf files, use relative path to
run configure can fix the problem.
This will reduce a lot of buildpaths QA issues in a world buld.
[YOCTO #8894]
(From OE-Core rev: 2c6ad43af0a0c4db5d7fd342be8585bec66debfb)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
For a bare-bone kernel recipe which specifies 32 bit x86 target,
a 64 bit .config will be generated from do_configure task when
building 32-bit qemux86, once all of these conditions are true:
* arch of host is x86_64
* kernel source tree used in build has commit ffee0de41 which
actually chooses i386 or x86_64 defconfig by asking host when
ARCH is "x86" (arch/x86/Makefile)
* bare-bone kernel recipe inherits directly from kernel without
other special treatments.
Build will fail because of the mismatched kernel architecture.
The patch sets ARCH i386 or x86_64 explicitly to configure
task to avoid this host contamination. Kernel artifact is also
changed so that it can map i386 and x64 back to arch/x86 when
needed.
(From OE-Core rev: 8d310b24927d0f348fb431895f0583733db2aad0)
Signed-off-by: Jianxun Zhang <jianxun.zhang@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
