Paul Eggleton [Mon, 13 Jun 2016 04:43:41 +0000 (16:43 +1200)]
scripts: ensure not specifying subcommand shows help text
With Python 2, argparse subparsers behaviour in Python 2 was to print
the usage information if the subparsers argument wasn't specified.
However, with Python 3.2.3 and later a subparsers argument is not
required by default, leading to errors when no arguments are specified:
AttributeError: 'Namespace' object has no attribute 'func'
Restore the previous desired behaviour of showing the help text for
devtool, recipetool and the devtool-stress script by setting
subparsers.required to True.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sun, 12 Jun 2016 22:57:53 +0000 (23:57 +0100)]
staging: Add BB_SETSCENE_VERIFY_FUNCTION2 for bitbake runqueue changes
We're about to rewrite the data structures in taskdata/runqueue in bitbake
and we 'leaked' knowledge about those structures to this single function.
Add a 'v2' function definition for use with the newer bitbake, the older
one can remain for compatibility for a while, then be removed. The function
is comparatively simple and rarely changes.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Zhenhua Luo [Sun, 12 Jun 2016 09:02:34 +0000 (17:02 +0800)]
binutils: fix the incorrect assembling for ppc wait instruction
The wait mnemonic for ppc targets is incorrectly assembled into 0x7c00003c due
to duplicated address definition with waitasec instruction. The issue causes
kernel boot calltrace for ppc targets when wait instruction is executed.
Signed-off-by: Zhenhua Luo <zhenhua.luo@nxp.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Aníbal Limón [Fri, 10 Jun 2016 15:12:10 +0000 (10:12 -0500)]
oe/distro_check.py: Fixes for python3
create_socket: Use urllib because urllib2 is now urllib in python3
and proxies as argument are deprecated so export them in the environ
instead.
get_links_from_url: Change usage of sgmllib for parsing HTML because
is deprecated in python 3, use instead bs4 that is already imported
in the bitbake tree.
[YOCTO #9744]
Signed-off-by: Aníbal Limón <anibal.limon@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
André Draszik [Fri, 10 Jun 2016 15:12:39 +0000 (16:12 +0100)]
mips: add tunes for (some) 24K cores
- add 24kc big and little endian, which is based on mips32r2 w/o FPU
- add 24kec which is 24kc + DSP
- both can have the MIPS16e ASE enabled in their tunes
Signed-off-by: André Draszik <adraszik@tycoint.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
André Draszik [Fri, 10 Jun 2016 15:12:38 +0000 (16:12 +0100)]
mips: add a tune for using MIPS16e ASE instructions
The MIPS16e instruction set still has to be enabled by setting
MIPS_INSTRUCTION_SET = 'mips16e'
in e.g. distro.conf and can be disabled on a per-recipe basis as
needed.
This is a similar approach as is available on ARM for Thumb support.
Note that contrary to the ARM Thumb support in OE, we do add a new
OVERRIDE (mips16e), as there are some recipes in OE that need to be
compiled slightly differently if mips16e mode is requested.
Signed-off-by: André Draszik <adraszik@tycoint.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
André Draszik [Fri, 10 Jun 2016 15:12:13 +0000 (16:12 +0100)]
uclibc: never build with SSP
This doesn't work, as the initial gcc that is used for compiling
uclibc doesn't have support for SSP yet (since that will only
be available once uclibc has been compiled). Since during that
same compilation step uclibc is trying to build its own utils
those are failing to compile with SSP enabled as the
initial gcc doesn't have access to the required libraries,
yet.
We never used to set UCLIBC_BUILD_SSP in the past, this was
only changed as part of the upgrade to uclibc-ng in
commit 63bdadc (uclibc: Switch to using uclibc-ng), so here
we now simply restore the previous behaviour.
Note that we still enable SSP support inside uclibc for
everybody else to use, though.
Signed-off-by: André Draszik <adraszik@tycoint.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
André Draszik [Fri, 10 Jun 2016 15:12:12 +0000 (16:12 +0100)]
uclibc: no need for CONFIG_MIPS_ISA_xxx config options
The config option for the mips ISA have been completely removed
from uclibc-ng. uclibc doesn't add gcc options based on those
config options anymore. Hence we don't need to create them here
either.
Signed-off-by: André Draszik <adraszik@tycoint.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
André Draszik [Fri, 10 Jun 2016 15:11:38 +0000 (16:11 +0100)]
image/image_types.bbclass: fix fatal error during cpio debugfs creation
If /init is just a symlink to /sbin/init, debugfs creation
fails with the following error:
ERROR: Error: The image creation script '<...>/debugfs.create_image.cpio' returned 1:
touch: cannot touch '<...>/cpio_append/init': Permission denied
WARNING: exit code 1 from a shell command.
ERROR: Function failed: do_rootfs
The reason is that IMAGE_CMD_cpio() is run twice on the same
WORKDIR. The first run creates a symlink in WORKDIR/cpio_append/init
to point to /sbin/init, while the 2nd run then tries to 'touch'
that link, which will fail, of course since /sbin/init is not
usually writable by non-root users.
Fix this by providing knowledge to the IMAGE_CMD_xxx() scripts
with regards to the fact that they are being executed in the
context of debugfs creation. The IMAGE_CMD_cpio() can now be
intelligent in the sense that it can avoid all additional symlink
handling during the debugfs run. The symlinks do not need to
be part of the debugfs, so we can skip that part altogether
in that case.
Signed-off-by: André Draszik <adraszik@tycoint.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Andre McCurdy [Fri, 15 Apr 2016 23:59:54 +0000 (16:59 -0700)]
gnutls: update 3.4.9 -> 3.4.11
* Version 3.4.10 (released 2016-03-03)
** libgnutls: Eliminated issues preventing buffers more than 2^32 bytes
to be used with hashing functions.
** libgnutls: Corrected leaks and other issues in gnutls_x509_crt_list_import().
** libgnutls: Fixes in DSA key handling for PKCS #11. Report and patches
by Jan Vcelak.
** libgnutls: Several fixes to prevent relying on undefined behavior of C
(found with libubsan).
* Version 3.4.11 (released 2016-04-11)
** libgnutls: Fixes in gnutls_record_get/set_state() with DTLS. Reported
by Fridolin Pokorny.
** libgnutls: Fixes in DSA key generation under PKCS #11. Report and patches
by Jan Vcelak.
** libgnutls: Corrected behavior of ALPN extension parsing during session
resumption. Report and patches by Yuriy M. Kaminskiy.
** libgnutls: Corrected regression (since 3.4.0) in gnutls_server_name_set()
which caused it not to accept non-null-terminated hostnames. Reported
by Tim Ruehsen.
** libgnutls: Corrected printing of the IP Adress name constraints.
** ocsptool: use HTTP/1.0 for requests. This avoids issue with servers
serving chunk encoding which ocsptool doesn't support. Reported by Thomas
Klute.
** certtool: do not require a CA for OCSP signing tag. This follows the
recommendations in RFC6960 in 4.2.2.2 which allow a CA to delegate OCSP
signing to another certificate without requiring it to be a CA. Reported
by Thomas Klute.
Signed-off-by: Andre McCurdy <armccurdy@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Jussi Kukkonen [Fri, 6 May 2016 12:28:10 +0000 (15:28 +0300)]
adwaita-icon-theme: Split symbolic icons into 2 packages
Symbolic icons are tiny but there's a lot of them and they actually
take multiple times their apparent size on a typical filesystem.
Split the symbolic icons: adwaita-icon-theme-symbolic contains the
icon sizes that seem to be typically used by GTK+ and applications
(16x16 & 24x24). Other sizes are in adwaita-icon-theme-symbolic-hires.
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Patrick Ohly [Fri, 10 Jun 2016 08:04:51 +0000 (10:04 +0200)]
npm.bbclass: avoid str/byte conversion problems for PKGV and SUMMARY
In Python3, str.encode() returns byte strings, which later are not
converted back to strings automatically, leading to "TypeError: Can't
convert 'bytes' object to str implicitly" in code which reads PKGV and
SUMMARY and expects to find strings there.
The npm.bbclass must use values for d.setVar() that meet that
expectation, and thus the redundant (and in Python3, harmful)
.encode() gets removed.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Tom Hochstein [Thu, 9 Jun 2016 19:27:06 +0000 (14:27 -0500)]
weston-init: Don't change XDG_RUNTIME_DIR if it already exists
In the case that XDG_RUNTIME_DIR already exists, the mkdir --parents
at least does nothing, but the chmod could overwrite the desired
file mode settings.
Signed-off-by: Tom Hochstein <tom.hochstein@nxp.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Most python recipes download its source from pypi.python.org,
but they changed the way theyre storing the source code, so
we need to use pythonhosted instead, see:
https://bitbucket.org/pypa/pypi/issues/438/backwards-compatible-un-hashed-package
Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
oe-pkgdata-util: new option to provide full info for binary package(s)
New option can be used for displaying full information about binary
package(s), including name, full version, recipe name, recipe full
version and package size.
This information can be useful inside custom buildhistory class
to produce detailed image manifest for further analysis.
List of packages can be specified as command line arguments or can
be read from file (e.g. from existing image manifest).
Warning: In case of image manifest is used as input outside of build
process, be aware that pkgdata might change since the time image
was built.
Upgrades python-numpy and python3-numpy to 1.11.0,
Drops using distutils class, to use setuptools class instead,
this fixes a problem where after compilation the package was created
as an egg, see YOCTO #8673
Fixes source location for both packages and also adds a new patch
to fix the shebang on the f2py script, see YOCTO #9729
Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bill Randle [Sat, 4 Jun 2016 22:45:03 +0000 (15:45 -0700)]
perl: fix several perl test failures
Several ExtUtils-MakeMaker tests fail when cross-compiled and run on
the target machine. Backport an upstream patch to fix the issues. Also
update the customized.dat hash file for the files modified by this patch
and other existing patches so the porting/customized.t test passes.
[YOCTO #8656]
Signed-off-by: Bill Randle <william.c.randle@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bill Randle [Sat, 4 Jun 2016 22:45:01 +0000 (15:45 -0700)]
perl: set proper perl subversion number in config files
During the upgrade from Perl 5.22.0 to 5.22.1 in commit f4c9908eae1ae3dcc38877abe2d5fbeb46851dd4 the config.sh file was hand edited
to change the subversion numbers. However, the edit was not entirely
correct. As a result the Perl version test failed. Set the correct
version strings.
[YOCTO #8656]
Signed-off-by: Bill Randle <william.c.randle@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ed Bartosh [Tue, 7 Jun 2016 11:59:13 +0000 (14:59 +0300)]
oe-pkgdata-util: fix AttributeError
The default arparse behaviour in Python 2 and early versions of Python 3
was to print usage message if no subcommands provided by user.
In Python >= 3.2.3 subparser has to have attribute 'required' set to
True to behave similar way.
By default subparser.required attribute is set to False, usage message is
not printed and code continues to be executed.
oe-pkgdata-util code raises AttributeError exception in
this case:
AttributeError: 'Namespace' object has no attribute 'func'
Fixed this by setting subparser.required to True.
[YOCTO #9711]
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Mariano Lopez [Mon, 6 Jun 2016 09:15:54 +0000 (09:15 +0000)]
testexport.bbclass: Create tarballs for easy release
This create tarballs in the testexport directory in order
to make easier to distribute the test in another systems.
There are three tarballs, one for the metadata that is not
arch dependant, another for packages needed by the DUT
(this depends of target MACHINE), and the last one for the
SDK needed by the systems that perform the tests.
This also create only the tarballs that are needed.
Mariano Lopez [Mon, 6 Jun 2016 07:15:44 +0000 (07:15 +0000)]
testexport.bbclass: Add support for testexport-tarball
Add support to export the SDK tarball needed when a test
system doesn't have the required software to perform runtime
tests.
The support is when exporting the test and when running
the test on a remote system. The user of this feature just
need to set TEST_EXPORT_SDK_ENABLED to "1" and declare
the sdk packages in TEST_EXPORT_SDK_PACKAGES.
Mariano Lopez [Mon, 6 Jun 2016 07:15:43 +0000 (07:15 +0000)]
testexport-tarball.bb: Add recipe
This new recipe is used when exporting runtime test outside
packages that won't be installed in the testing system but
are required for the runtime testing.
This new recipe is almost identical to buildtools-tarball,
but is able to define the SDK packages in local.conf.
Mariano Lopez [Mon, 6 Jun 2016 07:15:41 +0000 (07:15 +0000)]
oetest.py: Add install/uninstall functionality for DUTs
Add the functionality to install/unistall packages in the
DUTs without the use of the package manager. This is possible
with the extraction introduced in package manager class.
testimage and testexport bbclasses has been modified in order
to support this new feature.
[YOCTO #8694]
Signed-off-by: Mariano Lopez <mariano.lopez@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
alsabat has now a "standalone" mode, in which analysis with libfftw is
not needed. Therefore, it's now possible to enable alsabat without
fftw. However, there's no way to explicitly tell the configure script
to enable alsabat, but disable fftw. Without support for explicit
request to disable fftw, fftw would become a floating dependency,
which is why the recipe still only supports the "disable alsabat" and
"enable alsabat with fftw" options.
Dropped 0001-alsabat-rename-to-avoid-naming-conflict.patch (it's
included in the new release).
Signed-off-by: Tanu Kaskinen <tanuk@iki.fi> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bruce Ashfield [Mon, 6 Jun 2016 08:15:29 +0000 (04:15 -0400)]
linux-yocto/4.4: mmc backports
Backporting the following mcc commits to the 4.4 repo. These fix issues
being seen on broxton based boads:
1f3e98df094c mmc: core: Add a facility to "pause" re-tuning b27fcd162327 mmc: block: Pause re-tuning while switched to the RPMB partition 9aa07b4e274e mmc: block: Always switch back to main area after RPMB access
Signed-off-by: California Sullivan <california.l.sullivan@intel.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sat, 4 Jun 2016 08:04:26 +0000 (09:04 +0100)]
libarchive: Add PACKAGECONFIG for lz4 to ensure determinism
This avoids:
WARNING: opkg-1_0.3.1-r0 do_package_qa: QA Issue: libopkg rdepends on lz4, but it isn't a build dependency, missing lz4 in DEPENDS or PACKAGECONFIG? [build-deps]
