Richard Purdie [Sat, 26 Sep 2015 17:12:17 +0000 (18:12 +0100)]
gdk-pixbuf: Only apply native cleaning in normal task, not setscene
The previous fix works well for normal tasks since the ordering is always
correct and gdk-pixbuf-native comes first. Setscene ordering is reversed
so we can't rely on this there. Remove the fix in that case since
it breaks more than it solves.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
oeqa/testimage: Remove absolute path to oeqa from json
In the json file the whole datastore is serialized which contains
absolute path to the oeqa folder. This breaks the functionality
when trying to run the tests from other machines.
Signed-off-by: Lucian Musat <george.l.musat@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Previously, the iproute2's configure script is actually run in the
do_compile stage. There's a Makefile rule 'all: Config' which makes
the configure process run to generate the Config file.
However, this makes it hard to fix the dependency problems. We need
to generate the Config file in the do_configure stage.
Add do_configure_append to separate the configure process from the
compile process.
Besides, explicitly disable ATM support for iproute2, otherwise we'll
sometimes meet errors like below when building.
ld: cannot find -latm
collect2: error: ld returned 1 exit status
Makefile:154: recipe for target 'q_atm.so' failed
make[1]: *** [q_atm.so] Error 1
Signed-off-by: Chen Qi <Qi.Chen@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
The ccargs obtained from get_tune_parameters may not be fully expanded,
so that the gcc_multilib_setup function can be confused, and generates
invalid MULTILIB_OPTIONS in GCC Makefile fragment, which will break the
multilib feature of target gcc.
To address problems above, this patch modifies gcc_multilib_setup
function to expand ccargs before use.
Upstream-Status: Inappropriate [configuration]
Signed-off-by: Yuanjie Huang <yuanjie.huang@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bruce Ashfield [Thu, 24 Sep 2015 22:17:54 +0000 (18:17 -0400)]
linux-yocto/4.1: rt update to 4.1.x-rt8
The -rt build suffered the same breakage in master as the standard
kernel builds: due to a bad git history merge, existing patches were
dropped from branches.
This meant that we were missing the -rt changes, which re-introduced
some failures with linux-yocto-rt.
With these SRCREV updates, we have the latest 4.1.x-rt and all
architectures build.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
busybox: Use CC instead of bare LD to be the Linker
This patch was on mailing list, another patch to make sure -r is not
passed directly but via -Wl switch is added.
This was exposed when using clang and gold linker, clang does not have
-r switch to do relocatable objects and problem happens specific to OE
becuase we use LD = CC
now what happens is that busybox assumes that linker will be called
directly, and hence sprinkles linkers options in its kbuild system which
aggregate into LDFLAGS, some of these options are happily ignored by gcc
as well but it passes -r options rightly to linker so it all works,
however when using clang, this falls apart since -r is not known option
for clang so it drops this option and all obects which should be
partially linked becomes ET_EXEC and when they are added to final link
then gold starts to get confused
/mnt/home/kraj/work/angstrom/build/tmp-angstrom-glibc/sysroots/x86_64-linux/usr/bin/arm-angstrom-linux-gnueabi/arm-angstrom-linux-gnueabi-ld:
error: applets/built-in.o: unsupported ELF file type 2
clang-3.7: error: linker command failed with exit code 1 (use -v to see invocation)
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ross Burton [Thu, 24 Sep 2015 08:54:33 +0000 (09:54 +0100)]
sstate: run recipe-provided hooks outside of ${B}
To avoid races between the sstate tasks/hooks using ${B} as the cwd, and other
tasks such as cmake_do_configure which deletes and re-creates ${B}, ensure that
all sstate hooks are run in the right directory, and run the prefunc/postfunc in WORKDIR.
To avoid logs being overwriten when running the automated
tests multiple times, log files include timestamps in
their names and a link is created to point to the latest one.
Signed-off-by: Lucian Musat <george.l.musat@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Patrick Ohly [Thu, 24 Sep 2015 14:53:10 +0000 (16:53 +0200)]
image types: add hdddirect
When image-vm.bbclass was introduced, it indirectly also introduced a
".hdddirect" image type based on boot-directdisk.bbclass. However, one
could only get that image when also enabling at least one of the
virtual machine images.
The .hdddirect images are useful by themselves. By registering
image-vm.bbclass as implementation of it, it becomes possible to
select them with:
IMAGE_FSTYPES = "hdddirect"
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Alex Franco [Wed, 23 Sep 2015 16:13:56 +0000 (18:13 +0200)]
sanity.bbclass: show warning when chmod fails
* for some reason this part of:
http://patchwork.openembedded.org/patch/102561/
wasn't ever merged.
[YOCTO #7669]
Signed-off-by: Alex Franco <alejandro.franco@linux.intel.com> Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Patrick Ohly [Mon, 21 Sep 2015 14:30:10 +0000 (16:30 +0200)]
systemd: apply persistent storage udev rules also for /dev/hd*
This fixes booting with initramfs and root=UUID on machines with IDE
disks, like "runqemu hdddirect", and kernels which still use the
deprecated CONFIG_IDE.
v2: Rebased against current master-next.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Markus Lehtonen [Tue, 15 Sep 2015 12:29:55 +0000 (15:29 +0300)]
rpm: search for gpg if gpg2 is not found
Some (host) systems only have a binary named 'gpg' (e.g. Fedora) while
some only have 'gpg2' (Ubuntu) and others have both of them (openSUSE).
Currently the behavior of rpm-native with regards to GnuPG depends on
the host platform: rpm(-native) is configured to use GnuPG binary of the
host system if 'gpg2' is found in $PATH. Otherwise, rpm(-native) will
default to using '%{_bindir}/gpg2' which will be pointing to a sysroot
binary which usually does not exist.
This patch changes rpm to look for both 'gpg' and 'gpg2' when searching
for the GnuPG binary in PATH. This makes possible to create signed RPM
packages on different host platforms, using the GnuPG binary of the
host, without the need to explicitly define the gpg binary in bitbake
configuration (via GPG_BIN variable).
[YOCTO #8134]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Richard Purdie [Sat, 26 Sep 2015 16:49:25 +0000 (17:49 +0100)]
qemurunner: Handle qemu start failure correctly
If qemu didn't start correctly, we may not have registered the child
signal. This results in a nasty traceback which confuses the underlying
issue. Cleanup this code and make the handler cleanup conditional.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sat, 26 Sep 2015 16:47:39 +0000 (17:47 +0100)]
gcc-runtime: Add multilib C++ header mapping
The SDK was unable to find the C++ header pieces correctly since its
using a generic compiler, not one specifically targeting the multilib
vendor prefix. This adds in the right mapping to ensure multilib SDKs
work as expected. This fixes multilib SDK automated tests.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sat, 26 Sep 2015 16:46:53 +0000 (17:46 +0100)]
oeqa/oetest: Fix SDK command execution
The SDK environment wasn't being sourced when running these test
commands, meaning in some cases the cross compiler was being tested,
not the SDK tools. This is clearly not the intent so fix this by
ensuring the SDK environment is present. This fixes test failures
in multilib SDKs.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sat, 26 Sep 2015 16:44:59 +0000 (17:44 +0100)]
mulitlib: Ensure SDKTARGETSYSROOT is set correctly
When building something like lib32-core-image-minimal -c populate_sdk, we
expect one sysroot with both multilibs installed. We therefore
need a single SDKTARGETSYSROOT value which doesn't change when multilibs
are enabled.
This makes the image generation code match what the meta-environment
files set the SDK up to use.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Fri, 25 Sep 2015 22:25:12 +0000 (23:25 +0100)]
gdk-pixbuf: move gdk-pixbuf-query-loaders to $libdir for multilib safety
If for example gdk-pixbuf and lib32-gdk-pixbuf are in an image then only one
${bindir}/gdk-pixbuf-query-loaders will be installed, so only one variant will
actually be usable.
Solve this by moving gdk-pixbuf-query-loaders into ${libdir} as it's intimately
tied to the library and rarely directly invoked by the user, and update the
callers to use the right path.
Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 16 Sep 2015 02:04:12 +0000 (19:04 -0700)]
multilib: Drop populate_sdk variable manipulation
I believe this code dates from previous times when we didn't extend
the TOOLCHAIN_TARGET* variables to cover all multilibs. We now do
this so this code acutally breaks things by removing the non-multilib
variants.
By changing this, a multilib SDK now contains both sets of base libraries
which matches the tools we ship with it. If the user wishes to customise,
this also becomes easier.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Robert Yang [Wed, 16 Sep 2015 02:04:11 +0000 (19:04 -0700)]
package_manager.py: make rpm install mutilib pkgs corectly
When configure multilib, "bitbake <image_bb> -c populate_sdk" should
install all arch toolchains (for example, 32 and 64bit), but rpm not
handle the multilib requires correctly, for example:
lib32-packagegroup-core-standalone-sdk-target requires lib32-libc6, rpm
may pull in libc6 rather than lib32-libc6, there are the similar issue
when:
Richard Purdie [Wed, 23 Sep 2015 22:32:45 +0000 (23:32 +0100)]
gdk-pixbuf: Avoid rebuild failures
If gdkpixbuf-native rebuilds and there are stale (broken) modules lying around,
it can fail to run the postinst. E.g. svg links to harfbuzz and if harfbuzz is
removed from the sysroot but the svg loader isn't, we get a symbol linking issue.
The reproducer is along the lines of build gdk-pixbuf-native along
with harfbuzz-native and librsvg-native, then make a small change to the
gdk-pixbuf recipe that would cause it to rebuild, clean harfbuzz-native
and then build gdk-pixbuf.
To fix this, when we install gdk-pixbuf, we wipe out any previous loaders.
The idea is that gdk would always come first and anything else installing
itself will come later and rerun the postinst if needed. We can therefore
just remove any other loaders.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
systemd: fix tmpfiles location when multilib in use
Systemd's configuration files for creation, deletion and cleaning
of volatile and temporary files are installed in /usr/lib even when
multilib is in use (when /usr/lib64 is available). In this check the
systemd.conf file will not be found if libdir is /usr/lib64 so we fix the
path to match this file's installation path to look for it in
${exec_prefix}/lib
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Kai Kang [Wed, 23 Sep 2015 09:04:37 +0000 (17:04 +0800)]
p11-kit: configure without trust-paths
Configure option --with-trust-paths is only used for test scripts
trust/test-extract which is not packaged by default. If the option is
not provided, it checks 4 files on build machine. If the files don't
exist, configure fails.
Add configure option '--without-trust-paths' to fix this issue.
Signed-off-by: Kai Kang <kai.kang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Robert Yang [Wed, 16 Sep 2015 02:04:10 +0000 (19:04 -0700)]
oe-pkgdata-util: avoid returning skipped packages
The skipped packages may be pulled in by another package, for example,
when libc6-dbg is already installed and should be skipped, but it would
be pulled in by libsegfault, this patch fixes the issue.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Robert Yang [Wed, 16 Sep 2015 02:04:09 +0000 (19:04 -0700)]
toolchain-shar-extract.sh: remove checkbashism
Fixed when sh is bash:
$ sh ./tmp/deploy/sdk/poky-glibc-x86_64-core-image-minimal-core2-64-toolchain-1.8+snapshot.sh
Poky (Yocto Project Reference Distro) SDK installer version 1.8+snapshot
===========================================================
./tmp/deploy/sdk/poky-glibc-x86_64-core-image-minimal-core2-64-toolchain-1.8+snapshot.sh: 77: read: Illegal option -e
You are about to install the SDK to "/opt/poky/1.8+snapshot". Proceed[Y/n]? ../SDK2
Installation aborted!
There is ony one bashism "read -e" in toolchain-shar-extract.sh, but
'-e' is useless here, so remove it and use /bin/sh.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 23 Sep 2015 21:36:07 +0000 (22:36 +0100)]
archiver: stamp-base is dead, remove it
stamp-base was only ever used by the shared workdir code in gcc. This
turned out to be problematic and has been replaced by other approaches
which don't need specialist bitbake knowledge.
stamp-base will likely get removed from bitbake but for now, remove it
from archiver to simplfy the code since gcc no longer uses it.
This stops people getting confused by the obsolete code paths which I'm
getting a lot of questions about.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 23 Sep 2015 21:33:17 +0000 (22:33 +0100)]
gcc-shared-source: Set empty SRC_URI
gcc-source is the only gcc recipe meant to handle the fetch/unpack/patch
tasks, the other gcc recipes then depend on this.
This approach has been creating some confusion for tools like the archiver.
The simplest way to signal to these processes that there is no source
is to empty SRC_URI at the same time we disable the other tasks.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When building libgcc some files and libraries may be built but not packaged.
The original code packaged only some explicitly specified files targeting mostly x86.
This patch does not discriminate between various targets.
It fixes errors such as these:
ERROR: QA Issue: libgcc: Files/directories were installed but not shipped in any package:
/usr/lib/arm-poky-eabi/4.9.3/thumb
/usr/lib/arm-poky-eabi/4.9.3/fpu
/usr/lib/arm-poky-eabi/4.9.3/thumb/libgcov.a
/usr/lib/arm-poky-eabi/4.9.3/thumb/crtn.o
/usr/lib/arm-poky-eabi/4.9.3/thumb/crtend.o
/usr/lib/arm-poky-eabi/4.9.3/thumb/crtbegin.o
/usr/lib/arm-poky-eabi/4.9.3/thumb/libgcc.a
/usr/lib/arm-poky-eabi/4.9.3/thumb/crti.o
/usr/lib/arm-poky-eabi/4.9.3/fpu/libgcov.a
/usr/lib/arm-poky-eabi/4.9.3/fpu/crtn.o
/usr/lib/arm-poky-eabi/4.9.3/fpu/crtend.o
/usr/lib/arm-poky-eabi/4.9.3/fpu/crtbegin.o
/usr/lib/arm-poky-eabi/4.9.3/fpu/libgcc.a
/usr/lib/arm-poky-eabi/4.9.3/fpu/crti.o
Please set FILES such that these items are packaged. Alternatively if they are unneeded,
avoid installing them or delete them within do_install. [installed-vs-shipped]
Signed-off-by: Juro Bystricky <juro.bystricky@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
brian avery [Mon, 21 Sep 2015 23:45:55 +0000 (16:45 -0700)]
meta-selftest: add error recipe and error-image
* The error recipe throws an error during compile
* The error-image includes the error recipe
* Initial target is toaster selenium testing
Signed-off-by: brian avery <avery.brian@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This fixes systemd failing to start on Raspberry Pi 2 if it is compiled
with GCC 5.2.
It would try to start "Journal Service" and "udev Kernel Device Manager"
but fail repeatedly.
[YOCTO #8291]
Signed-off-by: Jonathan Liu <net147@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
oeqa/utils/dump: Add default commands and directory
Currently if qemu fails when running a selftest and
tries to run some commands on the host it will fail
because some variables required by the Dumper class
do not exist because testimage was not included.
This change adds a default parent directory to save
the dumps for the host or target. Also adds default
commands to run if no commands were provided to the
class. With these changes the previous errors using
selftest don't show anymore.
[YOCTO #8306]
Signed-off-by: Mariano Lopez <mariano.lopez@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In order for the test export runner to work oetest needs to be
separated from bitbake environment. There is no need to use bb
import here so we can use a logger instead.
Signed-off-by: Lucian Musat <george.l.musat@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We presently don't package isl. Unfortunately, if the host is already
using gcc-5.2 (as is presently the case on Arch Linux), configure will
autodetect the host's libisl, and do_compile will break because the
system isl headers aren't pulled in. In lieu of packaging isl, disable
it for now.
[YOCTO #8376]
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Wed, 23 Sep 2015 10:05:25 +0000 (11:05 +0100)]
devtool: upgrade: use shutil.move instead of os.rename
Rename fails over filesystem boundaries.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Markus Lehtonen [Wed, 23 Sep 2015 10:05:23 +0000 (11:05 +0100)]
devtool: second fix for running from a different directory
Do not change change current working directory permanently, but, only
for the duration of tinfoil initialization instead. The previous fix
caused very unintuitive behavior where using relative paths were solved
with respect to the builddir instead of the current working directory.
E.g. calling "devtool extract zlib ./zlib" would always create create
srctree in ${TOPDIR}/zlib, independent of the users cwd.
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com> Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Peter Seebach [Tue, 22 Sep 2015 22:59:10 +0000 (17:59 -0500)]
pseudo_1.7.4.bb: fix f*open()
The 0600 modes were coming from fopen/freopen/etc., because those
don't specify a filesystem mode (just an access mode like "r" or
"w"). Use 0666 & ~umask. (And then the PSEUDO_FS_MODE macro masks
in the 0600 bits we want to be sure are present.)
Signed-off-by: Peter Seebach <peter.seebach@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
oeqa/utils/qemurunner.py: Fix HIGH CPU usage on LoggingThread
LoggingThread is used for receive console output from QEMU
over TCP, so add filter to only wake poll on read events,
also change the event mask variable name to be more descriptive.
This fixes HIGH CPU consume caused by wake on ready to write
events.
Signed-off-by: Aníbal Limón <anibal.limon@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:41 +0000 (17:21 +0100)]
devtool: add search command
Adds a subcommand to search to find the target recipe name providing
some file or capability. This is implemented by searching on recipe
name, package name, description, package contents (file names), and
runtime file provides. For example:
This is particularly useful within the extensible SDK but is also made
available in devtool alongside the build system.
Note of course that because this searches pkgdata, useful results depend
upon do_packagedata(_setscene) having executed for the recipe being
searched for.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:40 +0000 (17:21 +0100)]
devtool: add basic means of running runqemu within the extensible SDK
We ship the runqemu script and if we build QEMU itself within the
extensible SDK, then it would be nice to be able to run it. This is a
very thin wrapper around runqemu, supplying the machine and image name
so the user doesn't need to. (This subcommand is only available within
the extensible SDK since it only really makes sense there where it is
otherwise hard to run runqemu directly.)
Implements [YOCTO #6657].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:39 +0000 (17:21 +0100)]
devtool / recipetool: add handling for binary-only packages
Add a means of creating recipes for package files or archives that
contain a directory structure to be installed verbatim, for example an
rpm file. (We mostly just re-use bin_package here and skip some of the
normal build system checks.) This support is available in "recipetool
create" and "devtool add" which wraps the former.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:38 +0000 (17:21 +0100)]
devtool: build-image: delete bbappend at end of build
Upon further reflection, it seems to me that this bbappend ought to just
be deleted at the end of the build. This keeps things simple; you never
have to remember to delete any files to get back to where you were
before with the image. This means we can also drop the slightly awkward
message reminding the user how to do that. I've also updated the test to
look at the image manifest to determine if the command has worked
instead of looking for the (now deleted) bbappend.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* Make image optional for the extensible SDK (auto-determine it based on
the targets the SDK was built for)
* Check that specified recipe is in fact an image
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:36 +0000 (17:21 +0100)]
devtool: build-image: tell user where to find output files
If the user is running "devtool build-image" within the extensible SDK
then they probably won't know where to find the resulting output files,
so we should tell them explicitly.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:34 +0000 (17:21 +0100)]
devtool: add: move important "recipe created" message to the end
If we end up printing a message about the build directory being the same
as the source, we should print that first and then print the message
about the recipe file possibly needing to be edited to the end so that
it has slightly more impact.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:33 +0000 (17:21 +0100)]
devtool: add: set up fetched source as a git repository by default
If the fetched source isn't already a git repository, initialise it as
one and then branch and tag, just as we do with "devtool modify". This
makes it easier to make changes, commit them and then use the
"devtool update-recipe" command to turn those commits into patches
on the recipe.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:32 +0000 (17:21 +0100)]
devtool: better handling for recipes that don't unpack source
Some recipes don't extract any source (for example, opkg-keyrings). We
were producing a traceback in this case because we weren't checking if
the directory existed.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:31 +0000 (17:21 +0100)]
devtool: fix extracting source for work-shared recipes
Recipes that use work-shared (such as libgcc) are capable of unpacking
the source, but it doesn't necessarily unpack to ${WORKDIR}/${BP}. Use
the last part of the actual S value instead which is more likely to
work.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:30 +0000 (17:21 +0100)]
devtool: show proper error when extracting source for recipes with disabled unpack task
If you try to use "devtool modify -x" or "devtool extract" on a recipe
where do_unpack has been set as noexec (e.g. glibc-locale), then we get
an error because the expected source wasn't ever unpacked. Do a check up
front for noexec being set on do_unpack and error out with a reasonable
message if that's the case.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:29 +0000 (17:21 +0100)]
recipetool: create: fix handling of URIs containing #
The # character in a URI denotes a fragment; we don't care about this
since it is never supposed to be sent to the server, so remove it from
the URI before actually trying to fetch it or use it in SRC_URI within
the recipe.
(This has come up because download links on pypi.python.org seem to have
a fragment containing the md5sum of the download; without stripping this
off the fetcher will choke on it.)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The shell considers empty functions to be a syntax error, so for
template shell functions that contain only comments (or no lines at all)
then add a : to act as a no-op which avoids the syntax error.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:27 +0000 (17:21 +0100)]
devtool: add: properly handle separate build directory
When we were adding a recipe for software that would typically be built
in the same directory as the source, we were always using a separate
build directory unless the user explicitly specified not to, leading to
errors for software that doesn't expect to be built that way (such as
Python modules using distutils). Split out the code that makes this
determination automatically from the "devtool modify" and "devtool
upgrade" code and re-use that here so the behaviour is consistent.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:26 +0000 (17:21 +0100)]
devtool / lib/oe/recipeutils: ensure we can parse without bbappends
These functions ostensibly allowed parsing a recipe without bbappends
but this clearly hadn't been tested because a variable was unassigned in
both of them in that case.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:24 +0000 (17:21 +0100)]
devtool: check that source tree still exists
Sometimes, particularly if you extracted the source to /tmp which is on
tmpfs, the external source tree that is being pointed to may no longer
exist when you come to run "devtool build" or "devtool update-recipe"
etc. Make all of the commands that need to check for a recipe being in
the workspace call a single function and have that function additionally
check the source tree still exists where appropriate.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:23 +0000 (17:21 +0100)]
scripts/contrib: add devtool stress tester
Add a script to run "devtool modify" followed by a build on every target
recipe in the environment (with the option to skip/resume from/only
include specific recipes). This takes far too long to run as an
oe-selftest test but is still something that is useful to be able to
run. There's also a slightly quicker mode that just runs "devtool
extract" on each recipe.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:22 +0000 (17:21 +0100)]
lib/oe/patch: fix for git am not cleaning up after itself
Unfortunately it appears that under certain circumstances, a failed
git am followed by git am --abort won't clean up any changes the patch
might have made - this was seen when running "devtool extract" on the
unzip recipe; unzip-6.0_overflow3.diff has a malformed date as far as
git am is concerned but it triggers this condition. Add a
git reset --hard HEAD followed by git clean -f in order to recover from
this scenario.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:21 +0000 (17:21 +0100)]
classes/externalsrc: fix setting of deps varflag as a string
The value of the deps varflag for tasks is (unusually) expected to be a
list object rather than a space-separated string, hence we cannot use
appendVarFlag() here. This fixes a traceback when parsing the gcc recipe
with externalsrc enabled, for example.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:20 +0000 (17:21 +0100)]
classes/externalsrc: scale back warning to a plain note
It turns out that there are folks out there who use externalsrc in
normal builds and don't really need to be warned; additionally within
the extensible SDK or when using devtool, it shouldn't be a warning
situation. Thus, scale it back to a note (we can't use bb.note() here
since that wouldn't actually be piped through to the bitbake UI). Also
touch up the message a little bit.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:19 +0000 (17:21 +0100)]
toolchain-shar-extract.sh: show progress when extracting SDK
Tar has supported a --checkpoint option since version 1.15.91, so it
should be safe to use here to print dots showing that it's still doing
something (technically it's not really progress unless you know how many
dots it's going to print, which even it doesn't know at the start, but
it's better than nothing).
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:18 +0000 (17:21 +0100)]
classes/populate_sdk_ext: drop work-config.inc
This is not actually used for anything - I thought that we would need to
use it within devtool to set global configuration, but we're able to do
everything we need within the bbappends it creates, which also saves on
parse time. If we're not going to use work-config.inc let's just drop it
completely.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:17 +0000 (17:21 +0100)]
classes/populate_sdk_ext: allow custom configuration for extensible SDK
Provide the ability to define a function containing extra configuration
values to be added to the local.conf file that goes into the SDK. For
example, this could be used to set up SSTATE_MIRRORS within the SDK.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 22 Sep 2015 16:21:16 +0000 (17:21 +0100)]
classes/populate_sdk_ext: fix missing environment settings if running installer with sh
If you ran the extensible SDK installer file with sh (instead of bash),
then the additional call to buildtools environment setup, extension of
PATH to support running devtool, and setting of OE_SKIP_SDK_CHECK
weren't being added to the end of the script. This is because apparently
bash is happy to expand wildcards in the target of a redirection, but
bash running in POSIX sh mode won't (although it apparently does work on
the sh command line rather than within a script run as an argument to
sh). In any case using a wildcard here is a bit of a crutch which we
don't need, so replace it with the proper path to the environment setup
script.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
