| Age | Commit message (Collapse) | Author |
|---|
|
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
singletask.lock is written out while certain tasks execute for recipes
that have externalsrc.bbclass enabled - this includes recipes in
devtool's workspace. It appears that there's a race where
singletask.lock will be there one minute and then when we try to get the
file checksum of it (since we want to know if anything in the source
tree has changed) it will be gone, and git chokes. To fix that, add
singletask.lock to .git/info/exclude in the repository, regardless of
whether we created the repository or not. In any case singletask.lock
should never be tracked by git, so this is a good thing to be doing for
that reason as well.
This fixes oe-selftest failures in test_devtool_modify that we've seen
on the Yocto Project autobuilder:
bb.data_smart.ExpansionError: Failure expanding variable
do_compile[file-checksums], expression was ${@srctree_hash_files(d)}
which triggered exception CalledProcessError: Command
'['git', 'add', '-A', '.']' returned non-zero exit status 128.
Note that this only fixes this issue for devtool; if you are using
externalsrc independently of devtool there's a chance this will still
be an issue unless you add singletask.lock to your .gitignore.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
This allows postinsts scripts to explicitly request a deferral to first boot
(by calling 'postinst_intercept delay_to_first_boot') instead of 'exit 1' which
should be used only to indicate actual script failures.
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
distrocompare.sh is added to compare the added list of recipes
between two releases. The output of the script will share the
information of the licenses required and other distributions
that are using the package.
If a single input is provided, it will compare the current
branch with the provided branch/commit-ish package list.
To run : distrocompare.sh <older hash> <newer hash>
E.g. distrocompare.sh morty 92aa0e7
E.g. distrocompare.sh morty pyro
E.g. distrocompare.sh morty
output : The script will produce a file ending with
new_recipe_list.txt preceeded by the branch name from input
Signed-off-by: Tan Shen Joon <shen.joon.tan@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
When searching for variables, include colon to ensure the script doesn't
find a variable that starts with the same name.
Signed-off-by: Amanda Brindle <amanda.r.brindle@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
In lookup_recipe, package_info, and list_pkg_files, check if the package
name exists in runtime-rprovides. If so, and the provider package has a
different name than the inputted package, print a note that says the
specified package is in another package's RPROVIDES. If the provider
package has the same name as the inputted package, continue as before.
Fixes [YOCTO 11943]
Signed-off-by: Amanda Brindle <amanda.r.brindle@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Refactor functions lookup_recipe and package_info to be consistent with
list_pkg_files. Print the appropriate information as soon as it's found,
rather than storing it in a mappings variable and wait to print after
searching all packages.
Signed-off-by: Amanda Brindle <amanda.r.brindle@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
See previous commit (python2 version) for more info, since mostly
everything applies here as well.
Old manifest file had several issues:
- Its unorganized and hard to read and understand it for an average
human being.
- When a new package needs to be added, the user actually has to modify
the script that creates the manifest, then call the script to create
a new manifest, and then submit a patch for both the script and the
manifest, so its a little convoluted.
- Git complains every single time a patch is submitted to the manifest,
since it violates some of its guidelines.
- It changes or may change with every release of python, its impossible
to know if the required files for a certain package have changed
(it could have more or less dependencies), the only way of doing so
would be to install and test them all one by one on separate individual
images, and even then we wouldnt know if they require less dependencies,
we would just know if an extra dependency is required since it would
complain, lets face it, this isnt feasible.
- The same thing happens for new packages, if someone wants to add a new
package, its dependencies need to be checked manually one by one.
Features/Fixes:
- A new manifest format is used (JSON), easy to read and understand.
This file is parsed by the python recipe and python packages
read from here are passed directly to bitbake during parsing time.
- It provides an automatic manifest creation task (explained on previous
commit), which automagically checks for every package dependencies and
adds them to the new manifest, hence we will have on each package
exactly what that package needs to be run, providing finer granularity.
- Dependencies are also checked automagically for new packages
(explained on previous commit).
This patch has the same features as the python2 version but it differs
in the following ways:
- Python3 handles precompiled bytecode files (*.pyc) differently.
for this reason and since we are cross compiling, wildcards couldnt be
avoided on python3 (See PEP #3147 [1]).
Both the manifest and the manifest creation script handle this
differently, the manifest for python3 has an extra field for cached
files, which is how it lets the user install the cached files or not
via : INCLUDE_PYCS = "1" on their local.conf.
- Shared libraries nomenclature also changed on python3, so again, we
use wildcards to deal with this issue ( See PEP #3149 [2]):
- Fixes python3 manifest, python3-core should be base and everything
should depend on it, hence several packages were deleted:
python3-enum, re, gdbm, subprocess, signal, readline.
- When building python3-native it adds as symlink to it called
nativepython3, which is then isued by the create_manifest task.
- Fixes [YOCTO #11513] while were at it.
References:
[1] https://www.python.org/dev/peps/pep-3147/
[2] https://www.python.org/dev/peps/pep-3149/
Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The reason we have a manifest file for python is that our goal is to
keep python-core as small as posible and add other python packages only
when the user needs them, hence why we split upstream python into several
packages.
Although our manifest file has several issues:
- Its unorganized and hard to read and understand it for an average
human being.
- When a new package needs to be added, the user actually has to modify
the script that creates the manifest, then call the script to create
a new manifest, and then submit a patch for both the script and the
manifest, so its a little convoluted.
- Git complains every single time a patch is submitted to the manifest,
since it violates some of its guidelines.
- It changes or may change with every release of python, its impossible
to know if the required files for a certain package have changed
(it could have more or less dependencies), the only way of doing so
would be to install and test them all one by one on separate individual
images, and even then we wouldnt know if they require less dependencies,
we would just know if an extra dependency is required since it would
complain, lets face it, this isnt feasible.
- The same thing happens for new packages, if someone wants to add a
new package, its dependencies need to be checked manually one by one.
This patch fixes those issues, while adding some additional features.
Features/Fixes:
- A new manifest format is used (JSON), easy to read and understand.
This file is parsed by the python recipe and python packages read
from here are passed directly to bitbake during parsing time.
- It provides an automatic manifest creation task (explained below),
which automagically checks for every package dependencies and adds
them to the new manifest, hence we will have on each package exactly
what that package needs to be run, providing finer granularity.
- Dependencies are also checked automagically for new packages (explained below).
- Fixes the manifest in the following ways:
* python-core should be base and all packages should depend on it,
fixes lang, string, codecs, etc.
* Fixes packages with repeated files (e.g. bssdb and db, or
netclient and mime, and many others).
- Sitecustomize was fixed since encoding was deprecated.
- The JSON manifest file invalidates bitbake's cache, so if it changes
the python package will be rebuilt.
- It creates a solution for users that want precompiled bytecode files
(*.pyc) INCLUDE_PYCS = "1" can be set by the user on their local.conf to
include such files, some argument they get faster boot time, even when the
files would be created on their first run?, but they also sometimes give a
magic number error and take up space, so we leave it to the user to
decide if they want them or not.
- Fixes python-core dependencies, e.g.
When python is run on an image, it TRIES to import everything it needs,
but it doesnt necessarily fails when it doesnt find something, so even if
we didnt know, we had errors like (trimmed on purpose):
# trying /usr/lib/python2.7/_locale.so
# trying /usr/lib/python2.7/lib-dynload/_locale.so
# trying /usr/lib/python2.7/_sysconfigdata.so
while it didnt complain about _locale it should have imported it,
after creating a new manifest with the automated script we get:
# trying /usr/lib/python2.7/lib-dynload/_locale.so
dlopen("/usr/lib/python2.7/lib-dynload/_locale.so", 2);
import _locale # dynamically loaded from /usr/lib/python2.7/lib-dynload/_locale.so
How to use (after a new release of python, or maybe before every OE
release):
- A new task called create_manifest was added to the python package,
which may be invoked via:
$ bitbake python -c create_manifest
This task runs a script on native python on our HOST system, and since
the python and python-native packages come from the same source, we can
use it to know the dependencies of each module as if we were doing it
on an image, this script is called create_manifest.py and in a very
simplistic way it does the following:
1. Reads the JSON manifest file and creates a dictionary data structure
with all of our python packages, their FILES, RDEPENDS and SUMMARY.
2. Loops through all of them and runs every module listed on them
asynchronously, determining every dependency that they have.
3. These module dependencies are then handled, to be able to know which
packages contain those files and which should RDEPEND on one another.
4. The data structure that comes out of this, is then used to create a
new manifest file which is automatically copied onto the user's python
directory replacing the old one.
Create_manifest script features:
- Handles modules which dont exist anymore (new release for example).
- Handles modules that are builtin.
- Deals with modules which were not compiled (e.g. bsddb or ossaudiodev)
- Deals with packages which include folders.
- Deals with packages which include FILES with a wildcard.
- The manifest can be constructed on a multilib environment as well.
- This method works for both python modules and shared libraries used
by python.
How to add a new package:
- If a user wants to add a new package all that has to be done is
modify the python2-manifest.json file, and add the required file(s)
to the FILES list, the script should handle all the rest.
Real example:
We want to add a web browser package, including the file webbrowser.py
which at the moment is on python-misc.
"webbrowser": {
"files": ["${libdir}/python2.7/lib-dynload/webbrowser.py"],
"rdepends": [],
"summary": "Python Web Browser support"}
Run bitbake python -c create_manifest and the resulting manifest
should be completed after a few seconds, showing something like:
"webbrowser": {
"files": ["${libdir}/python2.7/webbrowser.py"],
"rdepends": ["core","fcntl","io","pickle","shell","subprocess"],
"summary": "Python Web Browser support"}
Known errors/issues:
- Some special packages are handled differently: core, misc,
modules,dev, staticdev.
All these should be handled manually, because they either include
binaries, static libraries, include files, etc. (something that we
cant import).
Specifically static libraries are not not supported by this method
and have to be handled by the user.
- The change should be transparent to the user, other than the fact
that now we CANT build python-foo (it was pretty dumb anyway, since
what building python-foo actually did was building the whole python
package anyway), but doing IMAGE_INSTALL_append = " python-foo"
would create an image with the requested package with no issues.
[YOCTO #11510] [YOCTO #11694] [YOCTO #11695]
Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Add it to handle recipe-depends.dot and task-depends.dot. E.g.:
* Print why rpm is built
$ oe-depends-dot -k rpm --why/-w recipe-depends.dot
Because: core-image-sato libdnf libsolv dnf
* Print bzip2-native's depends
$ oe-depends-dot -k bzip2-native --depends/-d recipe-depends.dot
Depends: automake-native gnu-config-native libtool-native quilt-native autoconf-native
* Remove duplicated dependencies to reduce the size of the dot files.
For example, A->B, B->C, A->C, then A->C can be removed. The dot files are too
big, we nearly couldn't use 'dot -T' to generate pictcures for target recipes,
remove the duplicated dependencies makes is it possible.
$ bitbake core-image-sato -g
$ oe-depends-dot -r recipe-depends.dot
Saving reduced dot file to recipe-depends-reduced.dot
$ du -sh recipe-depends*.dot
608K recipe-depends.dot
32K recipe-depends-reduced.dot
It has been recuded from 608K to 32K, now we can generate a picture,
otherwise, it is too big:
$ dot -Tpng recipe-depends-reduced.dot -O
It also can handle task-depends.dot.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Colour value removes in red and additions in green, making it easier to scan the
output for relevant changes.
This adds a --colour option to specify whether colouring should be on, off, or
detected. The default is detected, and depends on whether stdout is a TTY (same
behaviour as git).
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Under some conditions, ioctl FIGETBSZ can't return real value.
We can try to use fallback via os.stat() to get block size.
Source of patch:
https://github.com/intel/bmap-tools/commit/17365f4fe9089df7ee9800a2a0ced177ec4798a4
Signed-off-by: Dogukan Ergun <dogukan.ergun@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
When we run wic within eSDK:
$ wic create mkefidisk -e core-image-minimal
ERROR: BUILDDIR not found, exiting. (Did you forget to source oe-init-build-env?)
In order to figure out variable values, one must have sourced
the OE build environment setup script. However, when we are in
within the eSDK environment which isn't initialised like the
normal OE build environment, we can't use wic utility with eSDK.
Reference:
https://www.yoctoproject.org/docs/latest/mega-manual/mega-manual.html#wic-requirements
While wic ought to be fixed to be able to run without bitbake
& native tools [YOCTO #11281], but this is a workaround to set
BUILDDIR in the environment so that bitbake environment is setup
for wic to build its required native tools.
Signed-off-by: Chang Rebecca Swee Fun <rebecca.swee.fun.chang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
wic modules in scripts/lib/ are needed for wic to work, but path to
the python module is not exported in eSDK environment and we were
using an absolutized path of wic script within the sysroots.
We now changed to use real script path instead, where the wic modules
are located. This will also resolved the tracebacks found when running
wic from within the eSDK environment.
Traceback (most recent call last):
File "/tmp/deploy/sdk/poky_sdk/sysroots/x86_64-pokysdk-linux/usr/bin/wic", line 58, in <module>
from wic import WicError
ImportError: No module named 'wic'
Signed-off-by: Chang Rebecca Swee Fun <rebecca.swee.fun.chang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
wic needs a set of tools to be available from sysroots.
wic will find bitbake executable within the environment,
and wic was unable to locate bitbake executable within eSDK
because it wasn't setup with the OE build environment script.
Hence, we need to add bitbake file path into the environment
PATH for wic to be able to discover it and import bb modules.
Signed-off-by: Chang Rebecca Swee Fun <rebecca.swee.fun.chang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Use the scriptpath module in order to standardize the adding of
bitbake and meta/lib path to sys.path.
Signed-off-by: Chang Rebecca Swee Fun <rebecca.swee.fun.chang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
On pre 4.15 host kernels, an APIC window emulation bug can cause qemu
to hang. On 64 bit we can use the x2apic, for 32 bit, we just have to
disable the other timer sources and rely on kvm-clock.
[YOCTO #12301]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In case of 'new_rootfs' the psuedo directory is not copied. Thus
PSEUDO_LOCALSTATEDIR should still point to the dsa
'native_sysroot'/../pseudo. Otherwise PSEUDO_LOCALSTATEDIR points to a not
existing director ('new_rootfs'/../pseudo) and UID and GUID attributes are not
applied to files of the image.
Signed-off-by: Adrian Fiergolski <adrian.fiergolski@cern.ch>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
First, allow for wic to be given a filesystem UUID to be used when
creating a filesystem. When not provided, wic will generate the UUID to
be used. Next, when --use-uuid is passed, we update the fstab to mount
things via UUID (and if not found, then use PARTUUID) as UUID is more
portable.
Signed-off-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
First, we support squashfs as root, so mention that. Second, the btrfs
rootfs creation function had a copy/paste of the previous function
comment, remove the irrelevant line.
Signed-off-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The SquashFS filesystem does not support UUIDs so make this combination
be an error.
Signed-off-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Specifically, 'devtool upgrade' will now do these things:
1) determine if any of the license checksums need updating; if so,
write the new checksums into the LIC_FILES_CHKSUM value in the recipe
that is written to the workspace;
2) print a notice to the standard output:
NOTE: New recipe is /home/ak/development/poky/build/workspace/recipes/puzzles/puzzles_git.bb
NOTE: License checksums have been updated in the new recipe; please refer to it for the difference between the old and the new license texts.
3) and the cool part: devtool will create a diff of the old and new licenses,
and write the diff into the workspace recipe as a comment, like this:
======
FIXME: the LIC_FILES_CHKSUM values have been updated by 'devtool upgrade'.
The following is the difference between the old and the new license text.
Please update the LICENSE value if needed, and summarize the changes in
the commit message via 'License-checksum-change:' tag.
(example: 'License-checksum-change: copyright years updated.')
The changes:
--- LICENCE
+++ LICENCE
@@ -1,8 +1,9 @@
This software is copyright (c) 2004-2014 Simon Tatham.
Portions copyright Richard Boulton, James Harvey, Mike Pinna, Jonas
-Klker, Dariusz Olszewski, Michael Schierl, Lambros Lambrou, Bernd
-Schmidt, Steffen Bauer, Lennard Sprong and Rogier Goossens.
+Kölker, Dariusz Olszewski, Michael Schierl, Lambros Lambrou, Bernd
+Schmidt, Steffen Bauer, Lennard Sprong, Rogier Goossens and Michael
+Quevillon.
Permission is hereby granted, free of charge, to any person
obtaining a copy of this software and associated documentation files
======
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
|
|
Specifically, 'devtool upgrade' will use the latest upstream release if available
or latest commit if upstream never makes releases.
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
|
|
This command queries the upstream server for what the latest release is and prints
the output; it is a much neater way to find out these things than fumbling with distrodata,
'bitbake -c checkpkg' and awkward to read csv output in a file.
Examples:
python3 (tarballs):
NOTE: Current version: 3.5.3
NOTE: Latest version: 3.6.3
rpm (git):
NOTE: Current version: 4.13.90
NOTE: Latest version: 4.14.0
NOTE: Latest version's commit: da3720f62e57648fb1dc2a632744d38866139971
puzzles (git without version tags):
NOTE: Latest commit: ee8ea9b9785964694cb2b3ad77c3fb2460f49510
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
|
|
If you specify 'tag=' for a git URL and passed to recipetool create, you
will get into Bitbake expansion error shown below:
----- snip -----
$ devtool add --version 2.4.2 mbedtls "git://github.com/ARMmbed/mbedtls;tag=mbedtls-2.4.2"
...
bb.data_smart.ExpansionError: Failure expanding variable SRCPV, expression was ${@bb.fetch2.get_srcrev(d)} which triggered exception FetchError: Fetcher failure: Conflicting revisions (abeccb9dbd7e19ae91ac50e1edd3803111c5f9b6 from SRCREV and mbedtls-2.4.2 from the url) found, please specify one valid value
----- snip -----
Assuming the tag is valid, we should get the tag commit hash and
drop the usage of 'tag=' from SRC_URI. By using a commit hash
corresponding to the tag will prevent bitbake from accessing
remote repository in order to expand SRCPV.
Signed-off-by: Chang Rebecca Swee Fun <rebecca.swee.fun.chang@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
* runpy allows running modules/scripts with 'python -m foo'
* python3-setuptools RDEPENDS on plistlib (present in python2)
* pip3 RDEPENDS on _markupbase (add to python3-core)
Signed-off-by: Tim Orling <timothy.t.orling@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The generators create python-*-manifest.inc files with
lines over 2500 characters long which breaks sending
patches via git send-email (because of smtp limitation).
This patchset formats all the long lines into multiple lines.
Signed-off-by: Tim Orling <timothy.t.orling@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The '--label' argument should work for '--source rawcopy' as it does for
'--source rootfs', so add a method in RawCopyPlugin to update the label
on the temporary filesystem images.
Signed-off-by: Martin Hundebøll <mnhu@prevas.dk>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Update the runqemu script to allow the user to specify a device tree
to boot when calling runqemu.
This involves creating a seperate check_dtb() function incase the user
has specified 'none' for the kernel but still wants a device tree.
Signed-off-by: Alistair Francis <alistair.francis@xilinx.com>
Reviewed-by: Ben Levinsky <ben.levinsky@xilinx.com>
Cc: Ben Levinsky <ben.levinsky@xilinx.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Update the runqemu script to allow the user to specify a Kernel to boot
when calling runqemu.
Signed-off-by: Alistair Francis <alistair.francis@xilinx.com>
Cc: Ben Levinsky <ben.levinsky@xilinx.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Removed the -m option since this script now searches through all
recipes in the configuration. Also removed dead code, which includes
the functions recipe_bbvars() and collect_bbvars().
Signed-off-by: Amanda Brindle <amanda.r.brindle@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
This makes debug easier.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Similarly to handling "../", handle "." to resovle to the qemuconf
file's current directory.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If a variable starts with "../", its likely its a path and we want to
set it to an absolute path relative to the qemuconf file.
This means we don't have to use bitbake as often to figure out variables.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The regexp in the script misses some tap devices, e.g. we see output like:
runqemu - INFO - Acquiring lockfile /tmp/qemu-tap-locks/tap25.lock failed: [Errno 11] Resource temporarily unavailable
runqemu - INFO - Acquiring lockfile /tmp/qemu-tap-locks/tap26.lock failed: [Errno 11] Resource temporarily unavailable
runqemu - INFO - Acquiring lockfile /tmp/qemu-tap-locks/tap27.lock failed: [Errno 11] Resource temporarily unavailable
runqemu - INFO - Acquiring lockfile /tmp/qemu-tap-locks/tap28.lock failed: [Errno 11] Resource temporarily unavailable
runqemu - INFO - Acquiring lockfile /tmp/qemu-tap-locks/tap40.lock failed: [Errno 11] Resource temporarily unavailable
runqemu - INFO - Acquiring lockfile /tmp/qemu-tap-locks/tap41.lock failed: [Errno 11] Resource temporarily unavailable
What happened to tap29 to tap39?
The issue is was we were missing devices with '0' in the number,
like "10:" and so on in the output from "ip link".
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If a setup dict in a python setup.py file pulled in the contents of
another dict (e.g. **otherdict), then we got an error when mapping
the keys because the key is None in that case. Skip those keys to avoid
the error (we pick up the values directly in any case).
A quick reproducer for this issue:
recipetool create https://files.pythonhosted.org/packages/source/p/pyqtgraph/pyqtgraph-0.10.0.tar.gz
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If you have a recipe that uses overrides to conditionally extend
SRC_URI to add additional patches, then you will often need to update
those patches if you're making other changes to the source tree (for
example if you're upgrading the underlying source). Make this possible
with devtool by creating devtool-override-* branches for each override
that conditionally appends/prepends SRC_URI, and have devtool
update-recipe / finish check each branch out in turn and update the
corresponding patches.
A current example of a recipe that does this is the quota recipe - it
applies an additional patch if musl is the selected C library (i.e.
libc-musl is in OVERRIDES).
Note that use of this functionality does require some care - in
particular, updates to patches that appear on the main branch (named
"devtool" by default) should be made there and not only on one of the
specific devtool-override-* branches that are created for each override.
The recommended procedure is to make the changes you want to make to the
main branch first, then check out and rebase each devtool-override-*
branch, testing each one by activating the corresponding configuration,
and then finally run devtool finish.
Fixes [YOCTO #11516].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If you're not sure what changes devtool finish is going to make, or
you're not sure you're finished with your modifications, it is useful to
be able to see what devtool finish is going to do beforehand, so add
a -N/--dry-run option to make that possible.
(It's also very useful for debugging devtool finish itself.)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If a file is going to be effectively removed from the destination by
devtool finish, we should report that rather than just reporting that
we're removing files from the workspace. This is a little tricky because
the way we actually operate when finishing is to:
(1) remove all original files (as recorded by devtool upgrade, if that
was used)
(2) as part of updating the recipe file, remove the files from next to
the new recipe (i.e. in the workspace for an upgrade, real recipe
otherwise) corresponding to commits not in the git tree
(3) copy over remaining files from the workspace to the destination
To report the files removed with respect to what was originally there,
we need to swap steps 1 and 2 so we can see what no longer exists after
the deletion, and suppress the reporting currently done in step 2 -
however, we still want to report removal in step 2 for the non-upgrade
case, so the latter is conditional.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If the files that the devtool-source class is supposed to create in the
source tree aren't found in the temporary directory then we know that
the class hasn't worked properly - say that explicitly.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
* Only log one warning message instead of one per line
* Be a bit more verbose
* "if list" is more pythonic than "if len(list)"
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If the directory where the source code extracts to changes (for
example, when upgrading iucode-tool from 1.5 to 2.1.1, the subdirectory
in the tarball changed from "iucode_tool-${PV}" to "iucode-tool-${PV}")
then handle this automatically. Also handle when it changes to match the
default S value (i.e. "${WORKDIR}/${BP}") in which case we just drop
setting S in the recipe.
Fixes [YOCTO #10939].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Give the user a little more insight into what's being done.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
github archive URLs are not guaranteed to be stable [1] and thus we
should show a warning if a user specifies one to recipetool create (or
devtool add).
[1] http://lists.openembedded.org/pipermail/openembedded-core/2017-September/142519.html
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
I noticed that using bitbake-layers add-layer followed by a devtool
command resulted in bitbake re-parsing all of the recipes, which is
annoying. Upon closer inspection I could see that devtool was moving the
workspace layer path to the end of BBLAYERS if it happened to be
somewhere in the middle - there's no need for it to be doing this. This
occurred because we were passing the current workspace path to remove
and the "new" path to add even if the path is not being changed, and I
think earlier versions of bb.utils.edit_bblayers_conf() didn't move the
existing entry under these circumstances as it clearly does now. Fix it
so we only pass the path to be removed if we're actually changing the
path.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
This looks like some debug printing that was left in by accident.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
devtool upgrade did not properly handle setting SRC_URI checksums for
recipes that use named SRC_URI entries and also use those names in the
SRC_URI checksums. A further complication was where the name contained
an expression that changed with the version e.g. ${PV} (probably quite
rare, but the dnsmasq recipe in meta-networking is currently one such
recipe.) All of these are now handled properly.
Additionally, drop the _get_checksums() function that wasn't being
called from anywhere in the code.
Note that this now turns nowrap_vars in recipeutils.py to be a list of
regexes, hence things such as [ and ] need to be appropriately escaped.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
Richard Purdie
Ross Burton
Paul Eggleton
Alexander Kanavin
Tan Shen Joon
Amanda Brindle
Alejandro Hernandez
Robert Yang
Dogukan Ergun
Chang Rebecca Swee Fun
Adrian Fiergolski
Tom Rini
Tim Orling
Martin Hundebøll
Alistair Francis