As reported by Paul Barker, my attempt to allow the patching
of kernel meta data broke other repository configurations, since
the meta data patch routing was matching too broadly and the
same repo would end up being patched twice.
Using his suggested fix, we are up and running with both types
of repos again.
(From OE-Core rev: 0fd09633e3c2fb4668c7700516d288f8c5356ff6)
Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
For testing purposes, it is often easier to patch a fragment
in a kernel-meta repository versus needing to make a copy or
modify the source repository.
We can allow this sort of patching when a patchdir of kernel-meta
is passed (to indicate the nested kernel-meta repository).
Also note that we must patch the meta data before they are
processed/gathered, since migrated copies to the kernel source
directory will be used later.
(From OE-Core rev: 324f9c8181150c23e3ce1ace4c45794cc8894167)
Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When using Kmeta directories, devtool finish will add every single file
in the directory to the bbappend. This is because in the current
implementation, the get_recipe_local_files function treats the kmeta
directory like a file. Modifying the function to loop through the
provided directories and return all included files instead of just the
top level directory. This will enable correct file to file comparison
when determing which files are new/changed and need to be added to the
bbappend.
Adding an extra check in devtool-source.bbclass to not copy the cfg file
if its already included somewhere in the kmeta directory
Also during 'modify', when moving necessary files in the kmeta directory
from the workdir to oe-local-files, the dangling parent directories are
left behind. This in itself is not an issue as the temporary devtool
workspace is automatically deleted, but this causes an incorrect include
directory to be added in kernel-yocto.bbclass. Changing the order of
the if statements to catch the correct conditional. This is safe to do
as when not in the devtool context, there will be no oe-local-files
directory.
(From OE-Core rev: 4671011b5b02d0989aa0fdcb50e18cb385a0e95e)
Signed-off-by: Jaewon Lee <jaewon.lee@xilinx.com>
Signed-off-by: Alejandro Enedino Hernandez Samaniego <alejandr@xilinx.com>
Acked-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
During the re-working of kernel fragments and status output cleanup the
visibility for BSP level errors was dropped/removed.
The audit phase can detect errors, redefinition and invalid
configuration fragments. We control the visibility of these reports via
the existing KCONF_BSP_AUDIT_LEVEL variable.
By default, errors and invalid configuration values will be displayed as
a warning. If redefinition values are to be shown the audit level must
be raised above the deafult value of '2'. We inhibit these by default,
since there are many valid reasons why a BSP will change a default /
base config .. and showing them offers no value to the BSP user.
(From OE-Core rev: 4933b686b0f75d6841630917a2d90832df41cac7)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Replace usage of os.utils.getstatusoutput() with direct subprocess
calls. Pass a modified environment and working directory where necessary
to bypass the need to execute in a shell.
(From OE-Core rev: 21de5cc43cfedc703e5bc0515507a6dae36afb74)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The devtool-source class moves all local files specified in SRC_URI to
an oe-local-files directory. When using devtool and a recipe space kernel-meta,
devtool modify throws an error because the paths the kernel-yocto class
is looking for feature directories in, don't include the oe-local-files
directory which devtool is using.
This patch checks for feature directories in oe-local-files,
and if present, adds that path to include directives.
[YOCTO #12855]
(From OE-Core rev: 4eceae7e3e6dab935e2cf49b75148139192fc6e7)
Signed-off-by: Jaewon Lee <jaewon.lee@xilinx.com>
Signed-off-by: Alejandro Enedino Hernandez Samaniego <alejandr@xilinx.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
With recent kernels (i.e. 4.17+) the configuration phase of the kernel
will check for capabilities/options of the compiler for CVE and other
mitigation support.
For a general kernel, we want to ensure that CC is fully defined when
the config targets are invoked (so the proper compiler will be checked).
For linux-yocto, we also need to specify the compiler/tools dependencies
for the configme task since it executes before configure and hence the
main kernel build DEPENDS will not always be in the sysroot before it
executes. Without those dependencies the kernel will be incorrectly
configured (i.e. bison is missing) or the configuration will fail the
mitigation tests.
[YOCTO #12757]
(From OE-Core rev: ff1bdd75d50f0ebac3d599e461685ace29559a82)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Initialize the git config prior creating a git repository in a plain
directory.
(From OE-Core rev: 0e3a9bed988b5fd4de770fe8cd1160b9f6c8c365)
Signed-off-by: Justin Bronder <jsbronder@cold-front.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
commit 7e98c295c1bb511e [kernel-yocto: ensure that only a single
defconfig is processed] has an overly broad replacement strategy
for 'defconfig' and hence will chop up any fragment name that
happens to contain that string.
If we change the processing to split on whitespace and drop
any full work 'defconfig' fragments, we'll get the behaviour
we want (no duplicate defconfigs, but fragments with defconfig
in their name are fine).
[YOCTO #12487]
(From OE-Core rev: fd2678190db6575dc909104b60c284d0c41f605f)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As a follow up to the changes to ensure that KBUILD_DEFCONFIG will
always get its defconfig into the processing, we also should ensure that
we don't get both a KBUILD_DEFCONFIG and a SRC_URI defconfig in the
configuration queue. If both are in the configuration queue, we end up
with competing values and a potentially long running check of 5000+
options x 2 against the final .config.
By removing the defconfig found from the SRC_URI when one is found
via KBUILD_DEFCONFIG, we ensure that only one will be processed.
(From OE-Core rev: 7e98c295c1bb511ece51b5f8c97f26c173ddaf76)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The bberror does not stop the build correctly, this should be a
bbfatal_log to ensure the failure correctly stops the build and logs
the failure.
Part of
[YOCTO #12162]
(From OE-Core rev: 44ff06635e7f575e67b7ebba5d6900b8ddbc4a06)
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When using KBUILD_DEFCONFIG, $sccs should be set to the $WORKDIR/defconfig
regardless if it compares or is copied. Otherwise $sccs is not set and the
defconfig is not found correctly.
Part of
[YOCTO #12162]
(From OE-Core rev: b63cc051f1eb58c768f49db2c04843336e62d3df)
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
commit 44aea7b873 [kernel-yocto: ensure that only valid BSPs are built]
introduced a new check to ensure that a valid BSP definition was found,
rather than building something that 'closely' matched the current
MACHINE.
This check breaks valid configurations which do not have a bsp
definition but are otherwise completely configured machines.
To allow both elements to co-exist (and not add warnings or errors
to otherwise valid builds), we first check to see if an empty bsp
definition was found, but then check to see if a defconfig was
provided.
If a defconfig has been provided, that is a sign that the board
configuration is complete and we should continue the build without
otherwise bothering the user.
Tested on meta-raspberrypi and linux-yocto*
(From OE-Core rev: 156cf9e25f875a50f907e570d9b62cbc2c051133)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There was a bug in the search routines responsible for locating
BSP definitions which returned a valid match if only the ktype
matched.
This meant that someone looking for "qemux86foo" (which is an
invalid definition) would potentially end up building "qemuarm"
and be none the wiser (until it didn't boot).
With this fix to the tools search routine, and improved return
code testing, we will now stop the build and report and error to
the user.
[YOCTO: #11878]
(From OE-Core rev: 44aea7b87307795fe4e089c51d45afccaa2f6525)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As pointed out by klapperichpaul@johndeere.com, missing configuration
fragments were being picked up twice, once by the tools and once by the
bbclass. Unfortunately, the tools error message was being detected as
configs, and hence no error was reported at all.
Rather than catching the output of the tools, we can instead check the
return code and propagate the error message from the tools directly to
the user.
[YOCTO #11649]
(From OE-Core rev: 960652416e2390337df6d9734375d6829ceb6420)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
It was reported that do_validate_branches was failing with the following
error:
Log data follows:
| DEBUG: Executing shell function do_validate_branches
| HEAD is now at fe0fb8d Merge tag 'v4.10.9' into standard/base
| mkdir: cannot create directory .: File exists
|
| [ERROR] Can't find patch dir at ./patches/standard/base
| usage: kgit s2q
| WARNING: exit code 1 from a shell command.
| ERROR: Function failed: do_validate_branches
This was triggered by the execution of 'kgit-s2q --clean' after forcing
the SRCREV to something other than the tip of the branch. --clean is
being run to remove any sentinel files from previous kernel builds to
ensure that the tree is in a consistent state.
There were two bugs, --clean was being executed and not exiting the
script as it was supposed to. Hence validation for applying patches
was done, and threw the error that eventually makes it to the console.
And the second bug is that since do_validate_branches actually calls
kgit-s2q --clean, the dependency on kern-tools-native needs to be on
that function (versus do_kernel_metadata which runs later).
With the tweaked kern-tool + the dependency fix, we no longer see this
error.
(From OE-Core rev: 4d5890b54cbdac01ee748759578b7b22ed8e61a2)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
A typo introduced an extra } in the logging of the bsp_definition.
While this won't cause problems .. it needs to be fixed!
(From OE-Core rev: 9b95f3fca2a94ca922e05321096e07a11e98f9bb)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In some cases it is seen that kernel_checkout and validate_branches are
run again in simultaneous builds. During do_patch the kgit-s2q mechanism
looks for a sentinel file inside the .git directory, finds a fence post
and starts picking up patches after that.
This can create trouble as validate_branches checks out the HEAD of the
branch and so the patches should be reapplied rather than skipped due to
finding of the fence post.
We can call kgit-s2q --clean to remove the sentinel file when the
branches are checked out.
(From OE-Core rev: d57aeafdd2b49010a9fa6d1cd9d10f3cfd5754a5)
Signed-off-by: Awais Belal <awais_belal@mentor.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When debugging a kernel configuration issue, one of the first questions
is "what BSP was used". To answer this qusetion, we log the BSP .scc
file that was used to generate the kernel configuration in the kernel
source meta directory.
(From OE-Core rev: 1003615b866bfe196994bdc25d305ff5eace2715)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The ability to merge two branches directly from a .scc file was
dropped during the streamlining of the tools.
As was pointed out by David Vincent <freesilicon@gmail.com>, there is
once again a valid use case for this functionality, so we restore the
capability.
(From OE-Core rev: a0059ebbb52c659282e355664bba1a2fa282170e)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There were a few straggling expansion parameter removals left for
getVar/getVarFlag where the odd whitespace meant they were missed
on previous passes. There were also some plain broken ussages such
as:
d.getVar('ALTERNATIVE_TARGET', old_name, True)
path = d.getVar('PATH', d, True)
d.getVar('IMAGE_ROOTFS', 'True')
which I've corrected (they happend to work by luck).
(From OE-Core rev: 688f7a64917a5ce5cbe12f8e5da4d47e265d240f)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
getVar() now defaults to expanding by default, thus remove the True
option from getVar() calls with a regex search and replace.
Search made with the following regex: getVar ?\(( ?[^,()]*), True\)
(From OE-Core rev: 7c552996597faaee2fbee185b250c0ee30ea3b5f)
Signed-off-by: Joshua Lock <joshua.g.lock@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
To trap errors and halt processing, do_kernel_metadata was recently
switched to exit on any non zero return code. While the concept is
sound, there are subcommands that have legitimate non-zero return
codes.
Instead of removing set +e, we'll explicitly check the return code
of the commands that can error, and throw a bbfatal to alert the
user.
(From OE-Core rev: a4705e62d0973c290011fc0d250501d358b659e8)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Historically the processing of kernel meta data contained some
commands that exited with a non-zero return code. Special processing
was required to properly deal with their exit.
That is no longer true, and instead of handling all return codes
and doing an explicit 'exit' call, we can remove set -e from the
routine and have all errors be trapped and stop processing.
(From OE-Core rev: 476ffd57cf5b6fba40d4e3f5dd913824ab8a8d3d)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If attempting to patch a git repo without a proper git config setup,
an error will occur saying user.name/user.email are needed by git
am/apply. After some code was removed from kernel-yocto, it was
simple enough to reproduce this error by creating a kernel patch and
using a container to build.
This patch abstracts out functionality that existed in buildhistory
for use in other classes. It also adds a call to this functionality
to the kernel-yocto class.
Fixes [YOCTO #10346]
introduced in OE-core revision
0f698dfd1c8bbc0d53ae7977e26685a7a3df52a3
(From OE-Core rev: 25b43cb05c645e43f96bc18906441b8fdc272228)
Signed-off-by: Stephano Cetola <stephano.cetola@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Various recipes depend on the kernel's do_shared_workdir
task, a quick grep suggests all external kernel modules
(via module.bbclass), but also perf, and potentially any
additional headers as outlined in linux-libc-headers.inc
are affected.
Having do_shared_workdir in SRCTREECOVEREDTASKS means this
task is removed when externalsrc is enabled, making all
those recipes fail as the task they depend on,
virtual/kernel:do_shared_workdir, doesn't exist.
Remove do_shared_workdir from SRCTREECOVEREDTASKS so that
all those recipes work even if externalsrc is activated.
According to the comment in here, the reason for
do_shared_workdir to be removed as a task is because it
modifies the source tree, but that doesn't seem to be
case.
(From OE-Core rev: 29e99d7a57803e450920600b5d35c5b4e58a0ede)
Signed-off-by: Andre Draszik <git@andred.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Previously merge_config.sh was wrapped by the configme script, configme
took the different KCONFIG_MODES as options, and used --allnoconfig
or --alldefconfig.
With the switch to merge_config.sh no longer being wrapped, the new
processing wasn't matching the existing values and only supported
allnoconfig or alldefconfig.
To avoid breaking existing layers, and also keep any working that
have already switched, we can make the case statement match both.
(From OE-Core rev: 614227f28a023fe148307e0d85a5e9b8d9b74372)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Before the kernel tools were simplified and streamlined, there was code
which not only migrated a patch/cfg/scc to the kernel build tree, it
also migrated any subdirectories of those patches.
The effect of this data migration was that any other meta data in
a patch's directory structure would be available for processing.
While we don't want to do this migration anymore, it is possible to
check the path of any SRC_URI patches, and if they include a "kernel-meta"
subdirectory add it to the search path.
This restores the functionality without the old complexity.
(From OE-Core rev: 7ef7af5c03bad28faf380986f792f7f3d4d5944d)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
do_kernel_configme calls merge_config.sh (installed in the sysroot by
the kern-tools-native recipe) which may invoke the compiler to complete
the configuration process.
Depending on the build (and dependencies), this may error due to sysroot poisoning [1].
The errors are similar to:
make[1]: Entering directory '4.1+gitAUTOINC+a7e53ecc27-r0/linux-x64-standard-build' HOSTCC scripts/basic/fixdep
work-shared/x64/kernel-source/scripts/basic/fixdep.c:106:23: fatal error: sys/types.h: No such file or directory
compilation terminated.
make[2]: *** [work-shared/x64/kernel-source/scripts/basic/Makefile:22: scripts/basic/x86_64-nilrt-linux-fixdep] Error 1
Adding $TOOLCHAIN_OPTIONS to $CFLAGS before calling merge_configs.sh
fixes the error because $TOOLCHAIN_OPTIONS defines the sysroot and make
uses it to correctly compile & fill all missing kernel config options.
[1] http://lists.openembedded.org/pipermail/openembedded-core/2014-October/098253.html
(From OE-Core rev: 4b770d62472d1b1a26366de0a1742db240aa5239)
Signed-off-by: Ioan-Adrian Ratiu <adrian.ratiu@ni.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
With the updated kernel tools, we generate a list of sccs, patches,
configs and BSP definitions as part of the meta data generation.
It is valid if there aren't any of these artifacts found (i.e. you
are just building a branch and a default config), but invoking the
tools with no inputs isn't a good idea.
To avoid this issue, we generate a string based on the artifacts
and skip calling the tools if there's nothing to do.
(From OE-Core rev: 58715183493de1deb90f2ab075048462b4bf6c73)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We've been running with a set of kern-tools that were designed to work
with build systems that knew nothing about git, trees, commits, etc.
As such, there's been a set of shims/wrappers in place to work with
within bitbake/oe-core. These were the *me scripts: createme, updateme,
patchme and configme.
With this commit, we strip that legacy code and use the tools directly.
This means less complexity, fewer corner cases .. and no surprises
when the tools are arunning. As another benefit, the tools consume
much less time during a typical build and have no noticeable impact
on the overall build time.
Existing .scc files, features, and processing are not impacted as
these tools are compatible with existing feature descriptions and
kerne configuration fragments.
The audit of kernel configuration fragments is now detached
from the linux-yocto build structure and process. This means that
they can eventually be tweaked to offer kernel audit to any type of
kernel build and configuration process.
Additionally, the kernel symbol audit phase can now resolve symbol
dependencies and offer guidance when a symbol is missing:
WARNING: linux-yocto-4.4.15+gitAUTOINC+b030d96c7b_f5e2c49d58-r0 do_kernel_configcheck: [kernel config]: specified values did not make it into the kernel's final configuration:
---------- CONFIG_BT_6LOWPAN -----------------
Config: CONFIG_BT_6LOWPAN
From: /home/bruce/poky/build/tmp/work-shared/qemux86-64/kernel-source/.kernel-meta/configs/standard/features/bluetooth/bluetooth.cfg
Requested value: CONFIG_BT_6LOWPAN=y
Actual value:
Config 'BT_6LOWPAN' has the following conditionals:
BT_LE && 6LOWPAN (value: "n")
Dependency values are:
BT_LE [y] 6LOWPAN [n]
(From OE-Core rev: 0f698dfd1c8bbc0d53ae7977e26685a7a3df52a3)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When working on the yocto-bsp and kernel-lab update for yocto 1.2
we found it was impossible for a end-user BSP to isolate patches
on a branch, since with the following commit:
[kernel-yocto: enforce SRC_URI specified branch]
Any new branch would be switched to whatever was specified on the
SRC_URI and undoing the work that the yocto-bsp tool did to support
board specific patches.
To fix this, we'll keep the enforcing of branch consistency enabled
by default, but introduce a variable "KMETA_AUDIT" that when not
set will skip the check.
There's no impact for existing users, and it is only something that
other plumbing commands and tools will need to use (or care about).
[YOCTO: #9120]
(From OE-Core rev: 1d4c120edeb6e45665eafd6962a10ebb89d758eb)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
During the simplication and cleanup of branches and kernel meta data
handling, the ability to force build a branch that didn't match the
meta data was dropped.
There are valid uses cases when a different branch should be built
(testing, development, etc), so we restore the capability with this
change.
If after the kernel meta data is processed the current branch does
not match the SRC_URI specified branch, a warning is generated
about the impending branch switch and that the user should double
check that they are building what they expect.
WARNING: After meta data application, the kernel tree branch is standard/base. The
WARNING: SRC_URI specified branch standard/gt. The branch will be forced to standard/gt,
WARNING: but this means the board meta data (.scc files) do not match the SRC_URI specification.
WARNING: The meta data and branch standard/gt should be inspected to ensure the proper
WARNING: kernel is being built.
Reported-by: Steve Sakoman" <steve.sakoman@intel.com>
(From OE-Core rev: d91a668bc0f6c2cfc52174b4039c7ea0d84e8d4d)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Move do_kernel_link_vmlinux() from kernel-yocto.bbclass into
kernel.bbclass so that it's available to any kernel recipe.
Note that the task is not enabled by default in kernel-yocto.bbclass,
so don't enable by default in kernel.bbclass either. To enable, see
the example in linux-yocto.inc, ie:
addtask kernel_link_vmlinux after do_compile before do_install
(From OE-Core rev: a29371848deda618a42f9a30f9856a44c2342fe6)
Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The do_kernel_link_vmlinux() task modifies the build directory (not
the source tree) and should not be skipped when externalsrc is being
used.
(From OE-Core rev: 70b8a60d42831c701d10066eb57aaaad7a6fb7ae)
Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The existing code doesn't tell regular (with .git) and bare cases and
just move the unpacked repo to the place of kernel source. But later
steps will fail on a bare-cloned repo because we can not checkout
directly in a bare cloned repo.
This change performs another clone to fix the issue.
Note: This change doesn't cover the case that S and WORKDIR are same
and the repo is bare cloned.
(From OE-Core rev: ccfa2ee5c4f509de4c18a7054b2a66fc874d5d69)
Signed-off-by: Jianxun Zhang <jianxun.zhang@linux.intel.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Before the fetcher validated the specified SRCREV was reachable on a
specified branch, linux-yocto style kernel's were comparing the value
of KBRANCH and branch on the SRC_URI and then allowing a SRC_URI
specified branch to override KBRANCH.
With the introduction of kernel meta data on the SRC_URI, this routine
is incorrectly picking up a kernel-cache repository and then attempting
to apply that branch information to the kernel repository.
The rationalization of the branch specification is largely no longer
required, and will may be removed in the future. But for now, to keep
changes minimal, we can simply not return branch information that comes
from kernel meta data by checking the 'type' parameter and skipping
if it is of type 'kmeta'.
(From OE-Core rev: 87363ec019e04b486dee5f07b7226465be7acd6c)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Make sure that 'do_unpack' is executed before 'do_kernel_metadata'.
Enabling externalsrc for kernel disables 'do_validate_branches' task
which caused 'do_kernel_metadata' to fail as the dependency chain to
'do_unpack' got broken.
[YOCTO #6658]
(From OE-Core rev: 8939ea428f642fd6fa48807ded1f9040f09ed375)
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The linux-yocto tree has always been a combined set of kernel changes
and configuration (meta) data carried in a single tree. While this
format is effective at keeping kernel configuration and source
modifications synchronized, it isn't always obvious to developers on
how to manipulate the meta data versus the source.
With this change, we remove the meta data processing from the
kernel-yocto class and use the external meta-data repository that
has always been used to seed the linux-yocto meta branch.
After this change, linux-yocto can no longer process combined trees,
and is simplified as a result.
(From OE-Core rev: 523e4f6a6913b64453579d27a02467e14f7df42e)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Change calls to bbfatal() to either die() or bbfatal_log() where we know
we want the full log to be printed by the UI (calling bberror or bbfatal
would otherwise suppress it since the change to connect these functions
through to the UI.) bbfatal() is still fine to use where there is enough
context information in the message such that the log isn't needed.
(From OE-Core rev: 04ed9a19e1b08003329138b8ab83691d13c11fd9)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
do_patch is currently doing checks with machine_srcrev without initiate
it which leads to below (additional debug added):
DEBUG: Executing shell function do_patch
.
.
.
+ [ 0 -ne 0 ]
+ [ != AUTOINC ]
+ git rev-parse --verify ~0
fatal: Needed a single revision
+ git merge-base HEAD
usage: git merge-base [-a|--all] <commit> <commit>...
or: git merge-base [-a|--all] --octopus <commit>...
or: git merge-base --independent <commit>...
or: git merge-base --is-ancestor <commit> <commit>
or: git merge-base --fork-point <ref> [<commit>]
-a, --all output all common ancestors
--octopus find ancestors for a single n-way merge
--independent list revs not reachable from others
--is-ancestor is the first one ancestor of the other?
--fork-point find where <commit> forked from reflog of <ref>
+ [ = ]
+ set +x
DEBUG: Shell function do_patch finished
Only reason it works today is because 'rev-parse/merge-base' with empty machine_srcrev
will result in "false positive". Solve this by adding a similar non-empty check and
use SRCREV as fallback as in 'do_kernel_metadata'
(From OE-Core rev: 5f7403151fc8bfd5b1d77f6626717c8b84d33d93)
Signed-off-by: Petter Mabäcker <petter@technux.se>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When a patch fails to apply, the kernel-yocto bbclass attempted to
be helpful and suggest that devshell be used to fix the issue.
The only problem is that you can't get to devshell if a patch is
failing.
We drop this bad advise and instead point to the linux source directory.
[YOCTO: #6202]
(From OE-Core rev: d6a0c0f60ea85235479d968f75d067a10ce21574)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As reported by Steffen Pankratz <Steffen.Pankratz@elektrobit.com>, the
previous logic of KBUILD_DEFCONFIG processing would not propagate an in
tree defcofig to WORKDIR if one was not already present.
We fix the propagation by copying the in tee config if a defconfig is
not already in WORKDIR.
Additionally we only warn (versus copying) if an in tree configuration
is specified, is different than the WORKDIR version and isn't copied.
(From OE-Core rev: 58cc903a5156063b24e48011438a709cb6e09f54)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There are two tasks that must run before a linux-yocto kernel is built.
- Kernel checkout and relocation to work-shared (kernel_checkout)
- Meta data gathering and configuration prep (kernel_metadata)
The current task definitions for both are simply "before do_patch",
which is correct, but kernel_checkout must run before and not race with
kernel_metadata.
So we set the definition of kernel_checkout to be more specific and
enforce the proper ordering.
[YOCTO: #7731]
(From OE-Core rev: 9d806f3d05de918952f839f5e595cb71cd1261db)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The kernel_configme task was added twice (once in the .bbclass, one in a .inc)
with different ordering constraints.
Change this to be just one definition in the bbclass with the stronger ordering
constraints.
(From OE-Core rev: b9646b9d31c3e0c70337a8c10ebfc087a0e2b829)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The autobuilder failed like this:
temp/run.do_kernel_metadata.25242: line 165: createme: command not found
createme is provided by kern-tools-native. do_patch has a dependency on
kern-tools-native, but do_kernel_metadata runs before do_patch. So move the
dependency from do_patch to do_kernel_metadata, moving the statement from the
.inc to the class so it's alongside the task definition.
[ YOCTO #7531 ]
(From OE-Core rev: 4a0371847ff0c30d9b60db63559d89dddfcb009f)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In a similar manner to the kernel itself, which does the following to
bring a defconfig into the configuration:
defconfig: $(obj)/conf
ifeq ($(KBUILD_DEFCONFIG),)
$< --defconfig $(Kconfig)
else
@echo "*** Default configuration is based on '$(KBUILD_DEFCONFIG)'"
$(Q)$< --defconfig=arch/$(SRCARCH)/configs/$(KBUILD_DEFCONFIG) $(Kconfig)
endif
We do the same with the linux-yocto configuration processing. If a
defconfig is specified via the KBUILD_DEFCONFIG variable, we copy it
from the source tree, into a common location and normalized "defconfig"
name, where the rest of the process will include and incorporate it
into the configuration process.
If the fetcher has already placed a defconfig in WORKDIR (from the
SRC_URI), we don't overwrite it, but instead warn the user that SRC_URI
defconfigs take precedence.
[YOCTO: #7474]
(From OE-Core rev: 5f5595eebeb81be7e824d998228e4ef9f0bfac7d)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The linux-yocto kernel has a meta-data component which accompanies the
actual tree. That meta-data is processed to generate a series file that
controls the patching and configuration of the kernel.
patching and configuration are two distinct phases, so when working on
kernel configuration, it doesn't make sense to always have to re-run
the patching step just to update configuration data in the meta-series.
To allow a more granular set of tasks, we break the meta-data generation
into a separate task, which runs before do_patch. This allows the task
to be explicitly called when working on configuration, but otherwise
has no impact on the build.
(From OE-Core rev: 1be4c68fb876b63f19d107275b701fe1c1c121db)
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
