Signed-off-by: Randy Li <ayaka@soulik.info> Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Randy Li <ayaka@soulik.info> Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
ayaka [Tue, 5 Feb 2019 02:57:23 +0000 (18:57 -0800)]
ThunderX: update the armv8a rules
I am not familiar with the big endian, so I don't update it.
I don't have much information about the Cavium ThunderX,
it looks like it supports all the ARM instructions.
Signed-off-by: Randy Li <ayaka@soulik.info> Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
ayaka [Tue, 5 Feb 2019 02:57:22 +0000 (18:57 -0800)]
arch-armv8a.inc: add tune include for armv8
There are some addtional instructions apart from bare armv8,
also there is armv8.1, armv8.2.
Most the processor would support crc, except X-gene 1.
Signed-off-by: Randy Li <ayaka@soulik.info> Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
kernel-devsrc: restore compatibility with kernel < v4.10
The kernel commit that added the syscall table generation tools
for ARM platform (ARM: convert to generated system call tables)
wasn't merged until kernel release v4.10.
Current assumption in kernel-devsrc recipe is that these syscall
scripts exist, which breaks the build for external board layers that are
still using 4.9.x kernels:
cp: cannot stat ‘arch/arm/tools/syscall*’: No such file or directory
so only try to install the scripts when they actually exist.
Signed-off-by: S. Lockwood-Childs <sjl@vctlabs.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Khem Raj [Tue, 5 Feb 2019 00:36:26 +0000 (16:36 -0800)]
multilib_header_wrapper.h: Use #pragma once
Avoid infinite include loops, especially with bits/wordsize.h which is
now possible with the synthesized headers since we now also synthesize
bits/wordsize.h itelf for some arches e.g. arm/aarch64
In cases where extra preprocessing tools are used such as clang-tidy
e.g. and these tools are not passed the knowledge about architecture
then case comes where we enter into header include loop for
bits/wordsize.h, since this template does explicitly include
bits/wordsize.h
To fix this emits the pragma once at beginning of file, this is better
solution than include guards, and pragma once is practically supported
on all compilers except few e.g. cray c/c++ compiler
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 5 Feb 2019 13:33:37 +0000 (13:33 +0000)]
package_rpm/archiver: Apply bandaid to src.rpm creation
| error: create archive failed on file /media/build1/poky/build/tmp/work/all-poky-linux/xcursor-transparent-theme/0.1.1+gitAUTOINC+23c8af5ba4-r0/deploy-sources/allarch-poky-linux/xcursor-transparent-theme-0.1.1+gitAUTOINC+23c8af5ba4-r0/xcursor-transparent-theme-0.1.1+git0+23c8af5ba4-r0.src.rpm: cpio: read failed - No such file or directory
| Building target platforms: noarch-poky-linux
| Building for target noarch-poky-linux
So do_deploy_archives can run from sstate, created a .src.rpm in WORKDIR/deploy-sources,
then it was removed when rpm was running. This leads to a broken Source line in the
spec file as the original file was found by the os.listdir().
This fix is just a bandaid over much more fundamental problems sadly.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
0007-Add-the-armv5e-architecture-to-binutils.patch is dropped since we
armv5 is not a used option anymore in OE armv5te is default, and gcc9
will drop it completely anyway
CVE patches were backports from master which is already past of 2.32
release
Other dropped patches were partial or full backports as well
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Randy MacLeod [Fri, 1 Feb 2019 21:16:28 +0000 (16:16 -0500)]
libsolv: Drop now uneeded musl patch
Drop the musl specific fopencookie patches since musl
added support for fopencookie in: 06184334 implement the fopencookie extension to stdio
in December 2017 so it has been in musl since v1.1.19.
There was no change in libsolv's configure log when building
with musl and these patches dropped.
Signed-off-by: Randy MacLeod <Randy.MacLeod@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
- selected bug fixes:
* fixed a couple of null pointer derefs and potential memory
leaks
* made disfavoring recommended packages work if strong recommends
is enabled
* no longer disable infarch rules when they don't conflict with
the job
- new features:
* do favor evaluation before pruning allowing to (dis)favor
specific package versions
Signed-off-by: Alejandro del Castillo <alejandro.delcastillo@ni.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Fri, 1 Feb 2019 13:09:39 +0000 (13:09 +0000)]
e2fsprogs: fix file system generation with large files
When copying files into the file system the file offset was being truncated to a
signed 32-bit value, so any files that are larger than 2^31 bytes were the right
size, but no content after that point.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Manjukumar Matha [Thu, 31 Jan 2019 22:21:26 +0000 (14:21 -0800)]
ruby.inc: Add dependency on readline-native
Add dependency on readline-native to fix the following issue
uninitialized constant Logfile
| Check ext/fiddle/mkmf.log for more details.
| readline:
| Could not be configured. It will not be installed.
|
build/tmp/work/x86_64-linux/ruby-native/2.5.1-r0/ruby-2.5.1/ext/readline/extconf.rb:62:
Neither readline nor libedit was found
| Check ext/readline/mkmf.log for more details.
| *** Fix the problems, then remove these directories and try again if
you want.
Signed-off-by: Manjukumar Matha <manjukumar.harthikote-matha@xilinx.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
André Draszik [Wed, 16 Jan 2019 12:51:16 +0000 (12:51 +0000)]
update-alternatives: try to update FILES_${PN} when renaming a file
When using update-alternatives, FILES_${PN} must be
referencing the new name after update-alternatives has
renamed files.
This is more or less OK when having static lists of files to
be packaged into a package, but makes it quite hard to
dynamically generate FILES_${PN}, e.g. using do_split_packages(),
as in that case we can not easily modify what goes into
FILES_${PN}, because that list is based on filenames as seen
at the time do_split_packages() is executing.
Of couse one could explicitly specify the (renamed) file(s)
in the recipe, but that contradicts the intended usage of
do_split_packages().
Instead, if FILES_${PN} contains the file name as it was pre
renaming, we here modify this to reflect the new name.
This will allow usage of do_split_packages() to populate
FILES_${PN}.
[YOCTO #13058]
Signed-off-by: André Draszik <andre.draszik@jci.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
André Draszik [Wed, 16 Jan 2019 12:51:15 +0000 (12:51 +0000)]
update-alternatives: convert file renames to PACKAGE_PREPROCESS_FUNCS
At the moment, the update-alternatives file renaming is
happening right after copying into PKGD during packaging
time using an _append OVERRIDE to the copy function
perform_packagecopy().
This is not really readable and hard to maintain.
Additionally, this makes it impossible to e.g. populate
PACKAGES dynamically using do_split_packages() and still
do update-alternatives - update-alternatives file renaming
requires the PACKAGES variable to have been fully populated
to work correctly. On the other hand, do_split_packages()
can only execute after perform_packagecopy(), as it needs
PKGD populated; so it's impossible to insert
do_split_packages() early enough in a deterministic way in
this use-case.
As there doesn't seem to be a reason not to, convert
this to a proper function and use PACKAGE_PREPROCESS_FUNCS
instead - after all, that's what this is meant for.
No other classes or recipes in oe-core or meta-openembededd
seem to have a hard requirement on update-alterantives
executing before any other PACKAGE_PREPROCESS_FUNCS, so
this should be perfectly fine.
The only implication is that if compress_doc is inherited,
compressed man-page file names will end up being, e.g.
eject.1.util-linux.gz
or
eject.1.gz.util-linux
based on the include of compress_doc.bbclass vs.
update-alternatives.bbclass order, but the symlink created
(alternative name) will always be correct.
This solves both problems:
* the code is easier to read / follow
* the above described use-case can be accomodated easily
[YOCTO #13058]
Signed-off-by: André Draszik <andre.draszik@jci.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
André Draszik [Wed, 16 Jan 2019 12:51:12 +0000 (12:51 +0000)]
util-linux: remove minix remnants
While commit fc66762d7c11 ("util-linux: Disable minix support.")
(or e88cee8cc31d in poky) removed most of the bits, there are
still references to minix remaining.
Remove them.
Signed-off-by: André Draszik <andre.draszik@jci.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
André Draszik [Wed, 16 Jan 2019 12:51:11 +0000 (12:51 +0000)]
util-linux: remove erroneous patch (qsort_r)
The patch in question is patching the code to use
qsort() instead of qsort_r(), without adopting the
compare function. This is a major issue, because
the compare function as written is evaluating /
accessing a third argument, which is not passed
with this OE patch, causing access to random
memory.
Given this patch was added so as to support (old)
linux (host) distros which might not provide
qsort_r(), according to the git history, and given
these days util-linux detects availability of
qsort_r() during configure phase, and given musl
builds (which doesn't provide qsort_r() either)
work without problem, the right solution is to
simply drop this invalid patch.
Do so.
Signed-off-by: André Draszik <andre.draszik@jci.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 30 Jan 2019 22:01:23 +0000 (22:01 +0000)]
oeqa/runtime/ptest: Ensure OOM errors are logged
Currently processed being killed by the OOM killer may not be spotted by
ptest-runner. After we complete the tests, check the logs and report if there
were any. This ensures the user is aware of OOM conditions affecting the
ptest results.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Alistair Francis [Wed, 30 Jan 2019 17:00:13 +0000 (17:00 +0000)]
libgpg-error: Bump to version 1.35
The file src/gpg-error.h.in SHA changed as some new funtion definitions
were added to the header file and the copyright year was updated to
include 2019.
Signed-off-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Host-assisted GL in Qemu in theory works with both SDL
and GTK; in practice SDL shows an empty screen. This
prepares the switchover of graphical qemu to use
native gtk, which also provides a neat set of menus to
control the emulator.
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Note that only the most minimal necessary subset of mesa is built;
particularly we don't build any drivers as that is offloaded
to the GL implementation provided by the host.
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Xulin Sun [Tue, 29 Jan 2019 07:31:22 +0000 (15:31 +0800)]
openssl: fix multilib file install conflicts
To avoid issue like below if run "bitbake lib32-core-image-minimal"
with series userspace packages(LAMP,krb5...) added.
Add multilib_script support for openssl's c_rehash which is a perl script.
Error: Transaction check error:
file /usr/bin/c_rehash conflicts between attempted installs of
lib32-openssl-bin-1.1.1-r0.armv7at2hf_neon and openssl-bin-1.1.1-r0.aarch64
Signed-off-by: Xulin Sun <xulin.sun@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jonas Bonn [Tue, 29 Jan 2019 07:24:41 +0000 (08:24 +0100)]
base-files: filesystems serves no real purpose
mount can generally figure out the filesystem type from the superblock;
otherwise, /proc/filesystems is a more correct fallback than
/etc/filesystems and presumably always available.
Signed-off-by: Jonas Bonn <jonas@norrbonn.se> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jan Kiszka [Tue, 29 Jan 2019 16:48:35 +0000 (17:48 +0100)]
oe-git-proxy: Avoid resolving NO_PROXY against local files
NO_PROXY may contain * elements, and if we are unlucky (or want to match
all hosts with *), we will pick up local files rather than doing the
match in match_host. Quoting helps here.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Chen Qi [Wed, 30 Jan 2019 02:56:24 +0000 (10:56 +0800)]
pkgdata.py: avoid target-sdk-dummy-provides to mess things up
Sometimes we meet the following failure for the test_lookup_recipe
test case.
AssertionError: 'zlib\nbusybox is in the RPROVIDES of target-sdk-provid[32 chars]ummy' != 'zlib\nbusybox'
zlib
+ busybox- busybox is in the RPROVIDES of target-sdk-provides-dummy:
- target-sdk-provides-dummy
This is because target-sdk-provides-dummy rprovides busybox.
So clean things up to avoid failure.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Aníbal Limón [Wed, 30 Jan 2019 17:14:32 +0000 (11:14 -0600)]
ptest-runner: Upgrade to 2.3.1
Changeset,
05b112b utils.c: Print DURATION after ERROR acb5efb utils.c: run_child redirect stderr to stdout e1062f7 ptest-runner: make DEFAULT_DIRECTORY be able to be defined when compiling
Signed-off-by: Aníbal Limón <anibal.limon@linaro.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 29 Jan 2019 16:52:18 +0000 (16:52 +0000)]
oeqa/logparser: Various misc cleanups
Get rid of further unneeded code complications:
* value mappings we could just direct use
* ftools when we can write files easily ourself
* test result status filtering we don't use
* variable overwriting module imports
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 29 Jan 2019 13:24:45 +0000 (13:24 +0000)]
oeqa/logparser: Reform the ptest results parser
Now we have a dedicated ptest parser, merge in the remaining ptest
specific pieces to further clarify and simplify the code, moving to
a point where we can consider extending/enhancing it.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
logparser is only used by ptest. Its slightly overcomplicated as it was
intended to be reusable but wasn't. Simplify it as a dedicated parser is
likely to me more readable and maintainable.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
libepoxy is a requirement of virglrenderer. Note that we strip
RPATH from the library, as this allows fall-through to the host
GL implementation, instead of attempting (and failing) to use
mesa-native.
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
sangeeta jain [Mon, 28 Jan 2019 08:11:15 +0000 (16:11 +0800)]
oeqa/manual/sdk.json: Updated toolchain tarball to core-image-sato
Manual test step for SDK is updated. Previously toolchain was "poky-glibc-x86_64-core-image-sato-sdk<type-arch>-toolchain-<release-version>.sh"
But toochain for core-image-sato-sdk is not available in releases after 2.1, hence changed it to "poky-glibc-x86_64-core-image-sato-<type-arch>-toolchain-<release-version>.sh
Other less significant update is to add exepcted result in intermediate test steps.
Richard Purdie [Mon, 28 Jan 2019 13:57:44 +0000 (13:57 +0000)]
systemd: Update recent CVE patches
* Added CVE tag, Upstream-Status tag and Sign-off-by tags.
* Removed the verification of the entry length in the header
* Squashed CVE-2018-16865 patches into one
* CVE-2018-16866 patch now taken from systemd-stable and includes
an additional heap buffer overflow fix.
Signed-off-by: Marcus Cooper <marcusc@axis.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Khem Raj [Sun, 27 Jan 2019 20:56:15 +0000 (12:56 -0800)]
arm-tunes: Remove -march option if mcpu is already added
tune files which inherit the arch definitions already define appropriate
-mcpu option, which is equivalent of right -march and -mtune combination
and is preferred since gcc is getting stricter and stricter with option
check semantics and can now find incompatible -march and -mcpu options
better with every release. It does internal feature consistency check
and if it finds out discrepency between what -mcpu would expand to as
compared to -march it will flag the options to be incompatible, for
naked eye it sounds wrong but gcc would translate -mcpu to a given
-march internally and it might not match to what we set in these arch
files.
The effects are quite subtle, where this can result in configure test
failing to compile due to these incompatible options and a feature
option getting disabled for a recipe for no reason.
e.g. with gcc9 which can now detect that -mcpu=cortex-a5 and
-march=armv7-a are incompatible, many features in libstdc++ ends up
disabled due to configure check failures e.g. size_t size, ptrdiff_t
sizes, which inturn results in compiling libstdc++ with unwanted
disabled features.
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Fri, 25 Jan 2019 19:47:45 +0000 (13:47 -0600)]
Change default debug split to make separate source packages
Changes the default PACKAGE_DEBUG_SPLIT_STYLE to generate separate
source and debug packages. SDKIMAGE_FEATURES is updated to include the
source packages so that there is not change for the SDK contents.
[YOCTO #12931]
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ruslan Bilovol [Sat, 26 Jan 2019 12:58:00 +0000 (14:58 +0200)]
perl: drop lost patch
Commit 68552c353255 "perl: remove the previous version of the recipe"
dropped 0001-Makefile.SH-Pod-Simple-requires-Getopt-Long.patch
from recipe, but left the patch itself in source tree.
Remove this patch since nobody uses it.
Cc: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Ruslan Bilovol <ruslan.bilovol@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ruslan Bilovol [Sat, 26 Jan 2019 12:57:59 +0000 (14:57 +0200)]
subversion: drop lost patch
Commit 5bb47984af79 "subversion: 1.9.7 -> 1.10.0" dropped
serf.m4-Regex-modified-to-allow-D-in-paths.patch
from recipe, but left the patch itself in source tree.
Remove this patch since nobody uses it.
Cc: Richard Purdie <richard.purdie@linuxfoundation.org> Signed-off-by: Ruslan Bilovol <ruslan.bilovol@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ruslan Bilovol [Sat, 26 Jan 2019 12:57:58 +0000 (14:57 +0200)]
connman: drop lost patch
Commit 85b76e52d206 "connman: update to 1.36" dropped
0001-inet-Add-prefixlen-to-iproute_default_function.patch
from recipe, but left the patch itself in source tree.
Remove this patch since nobody uses it.
Cc: Oleksandr Kravchuk <dev@sashko.rv.ua> Signed-off-by: Ruslan Bilovol <ruslan.bilovol@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ruslan Bilovol [Sat, 26 Jan 2019 12:57:57 +0000 (14:57 +0200)]
dhcp: drop lost patch
Commit 7cb42ae87ef9 "dhcp: update 4.4.1" dropped
0008-tweak-to-support-external-bind.patch
from recipe, but left the patch itself in source tree.
Remove this patch since nobody uses it.
Cc: Armin Kuster <akuster808@gmail.com> Signed-off-by: Ruslan Bilovol <ruslan.bilovol@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ruslan Bilovol [Sat, 26 Jan 2019 12:57:56 +0000 (14:57 +0200)]
bind: drop lost patch
Commit "c37207d0aca5 bind: update to ESV version 9.11.3" dropped
0001-build-use-pkg-config-to-find-libxml2.patch
from recipe, but left the patch itself in source tree.
Remove this patch since nobody uses it.
Cc: Armin Kuster <akuster808@gmail.com> Signed-off-by: Ruslan Bilovol <ruslan.bilovol@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ruslan Bilovol [Sat, 26 Jan 2019 12:57:55 +0000 (14:57 +0200)]
lttng: drop lost patch
Commit "f63965c0f9fc lttng: uprev to 2.10.7" dropped
0001-Fix-btrfs-Remove-unnecessary-fs_info-parameter.patch
from recipe, but left the patch itself in source tree.
Remove this patch since nobody uses it.
Cc: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ruslan Bilovol <ruslan.bilovol@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
lib/oe/patch.py: Fix applying a directory as a patch
If a SRC_URI content ends with '.patch' bitbake is
trying to apply it as it's a patch file.
It causes that if we use git repository for 'patch' package
the bare clone is extracted to a directory
(i.e. build/downloads/git2/git.mirror.org.patch/) which is considered
to be a patch file, so patch.py tries to apply that directory as a patch
which ends up with a failure.
Signed-off-by: Tomasz Dziendzielski <tomasz.dziendzielski@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Robert Yang [Thu, 24 Jan 2019 06:57:31 +0000 (14:57 +0800)]
ccache: 3.5 -> 3.6
* Rebased 0002-dev.mk.in-fix-file-name-too-long.patch and sent it to upstream,
and got merged.
* The LIC_FILES_CHKSUM is changed because of year updated.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
lib/oe/reciputils.py: parallelize upstream version checks
Previously this was done via bitbake tasks, and when this
was rewritten to a for loop, performance sufered significantly:
from 90 seconds to about 12 minutes for oe-core. This change
restores the previous run time, and makes it possible
to perform such checks with command line utilities in an
interactive way.
Implementation note: we have to create a copy of the recipe
data, as Tinfoil API can't be used from multiple threads
and only allows one process to access the data at a time.
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
cmake: do not look into native sysroot in the nativesdk environment
I am not sure why we do this in the first place, but it is causing
cmake to erroneously pick up items from the native sysroot
when building for the target and the target item is missing, for example:
lib/oe/package_manager: turn postinst_intercept warnings into failures for nativesdk
The few cases where they failed should be now all fixed. The only allowed
exception is when building mingw32 SDKs, as there is currently no support for running
postinst_intercepts through wine.
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
