One use-case for buildman is to continually run it interactively after
each small step in a large refactoring operation. This gives more
immediate feedback than making a number of commits and then going back and
testing them. For this to work well, buildman needs to be extremely fast.
At present, a couple issues prevent it being as fast as it could be:
1) Each time buildman runs "make %_defconfig", it runs "make mrproper"
first. This throws away all previous build results, requiring a
from-scratch build. Optionally avoiding this would speed up the build, at
the cost of potentially causing or missing some build issues.
2) A build tree is created per thread rather than per board. When a thread
switches between building different boards, this often causes many files
to be rebuilt due to changing config options. Using a separate build tree
for each board would avoid this. This does put more strain on the system's
disk cache, but it is worth it on my system at least.
This commit adds two command-line options to implement the changes
described above; -I ("--incremental") turns of "make mrproper" and -P
("--per-board-out-dir") creats a build directory per board rather than per
thread.
Tested:
./tools/buildman/buildman.py tegra
./tools/buildman/buildman.py -I -P tegra
./tools/buildman/buildman.py -b tegra_dev tegra
./tools/buildman/buildman.py -b tegra_dev -I -P tegra
... each once after deleting the buildman result/work directory, and once
"incrementally" after a previous identical invocation.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Acked-by: Simon Glass <sjg@chromium.org> # v1
Tested-by: Simon Glass <sjg@chromium.org> # v1
Acked-by: Simon Glass <sjg@chromium.org>
This option outputs to the log file, not to the terminal. Clarify that in
the help, and add a mention of it in the README.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reported-by: Stephen Warren <swarren@wwwdotorg.org>
Reviewed-by: Stephen Warren <swarren@nvidia.com>
At present buildman allows you to specify the directory containing the
toolchain, but not the actual toolchain prefix. If there are multiple
toolchains in a single directory, this can be inconvenient.
Add a new 'toolchain-prefix' setting to the settings file, which allows
the full prefix (or path to the C compiler) to be specified.
Update the documentation to match.
Suggested-by: Stephen Warren <swarren@wwwdotorg.org>
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>
At present if you try to use buildman with the branch 'test' it will
complain that it is unsure whether you mean the branch or the directory.
This is a feature of the 'git log' command that buildman uses. Fix it
by resolving the ambiguity.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
At present the architecture is deduced from the toolchain filename. Allow it
to be specified by the caller.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com
At present the priority of a toolchain is calculated from its filename based
on hard-coded rules. Allow it to be specified by the caller. We will use
this in a later patch. Also display the priority and provide a message when
it is overriden by another toolchain of higher priority.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
Normally we use a single quote for strings unless there is a reason not to
(such as an embedded single quote). Fix a few counter-examples in this file.
Also add a missing function-argument comment.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
It is convenient to install symlinks to buildman and patman in the search
patch, such as /usr/local/bin. But when this is done, the -H option fails to
work because it looks in the directory containing the symlink instead of its
target. Fix this.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
Since commit 87da2690ab
"openrisc: updating build tools naming convention", openrisc
kernel.org toolchain is out of date and cannot build U-Boot.
Update buildman and moveconfig tools to refer to the new one.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Add links for toolchains not available on kernel.org.
The sh4 toolchains from kernel.org dose not work for some boards,
so use the sh from Sourcery.
Signed-off-by: Thomas Chou <thomas@wytron.com.tw>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
At present buildman can compare configurations between commits but the
feature is less useful than it could be. There is no summary by architecture
and changes are not reported on a per-board basis.
Correct these deficiencies so that it is possible to see exactly what is
changing for any number of boards.
Note that 'buildman -b <branch> -C' is recommended for any build where you
will be comparing configuration. Without -C the correct configuration will
not be reported since changes will often not be picked up.
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Corresponds to ba71a0e (Fix _parse_block() 'parent' documentation re.
ifs.) from upstream, just adding the SPDX tag.
Has performance improvements, code cleanup, Python 3 support, and various
small fixes, including the following:
- Unset user values when loading a zero-byte .config. (5e54e2c)
- Ignore indented .config assignments. (f8a7510)
- Do not require $srctree to be set for non-kernel projects. (d56e9c1)
- Report correct locations in the presence of continuation lines.
(0cebc87)
Signed-off-by: Ulf Magnusson <ulfalizer@gmail.com>
The doc wrongly put sandbox in the '--fetch-arch' command. Remove it.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
On i.MX platforms the SPL binary is called "SPL" so make sure we keep
that.
Cc: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Acked-by: Simon Glass <sjg@chromium.org>
This commit imports some updates of kconfiglib.py from
https://github.com/ulfalizer/Kconfiglib
- Warn about and ignore the "allnoconfig_y" Kconfig option
- Statements in choices inherit menu/if deps
- Add Symbol.is_allnoconfig_y()
- Hint that modules are still supported despite warnings.
- Add warning related to get_defconfig_filename().
- Fix typo in docs.
- Allow digits in $-references to symbols.
Signed-off-by: Ulf Magnusson <ulfalizer@gmail.com>
Signed-off-by: Philip Craig <philipjcraig@gmail.com>
Signed-off-by: Jakub Sitnicki <jsitnicki@gmail.com>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Also read gcc 4.9.0 at kernel.org which also have Microblaze toolchain.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Fixed unit test failure by updating the test:
Signed-off-by: Simon Glass <sjg@chromium.org>
The help text for -V says we will pass V=1 but all it really did was not
pass in -s. Change the logic to pass make V=1 with given to buildman -V or
-s to make otherwise.
Cc: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Acked-by: Simon Glass <sjg@chromium.org>
When told to keep outputs, be much more liberal in what files we keep.
In addition to adding 'MLO', keep anything that matches u-boot-spl.* (so
that we keep the map file as well) and anything we generate about
'u-boot itself. A large number of bootable formats now match this and
thus it's easier to build many targets and then boot them afterwards
using buildman.
Cc: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Acked-by: Simon Glass <sjg@chromium.org>
It is useful to be able to see CONFIG changes made by commits. Add this
feature to buildman using the -K flag so that all CONFIG changes are
reported.
The CONFIG options exist in a number of files. Each is reported
individually as well as a summary that covers all files. The output
shows three parts: green for additions, red for removals and yellow for
changes.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present buildman tries to detect an aborted build and doesn't record a
result in that case. This is to make sure that an abort (e.g. with Ctrl-C)
does not mark the build as done. Without this option, buildman would never
retry the build unless -f/-F are provided. The effect is that aborting the
build creates 'fake errors' on whatever builds buildman happens to be
working on at the time.
Unfortunately the current test is not reliable and this detection can
trigger if a required toolchain tool is missing. In this case the toolchain
problem is never reported.
Adjust the logic to continue processing the build result, mark the build as
done (and failed), but with a return code which indicates that it should be
retried.
The correct fix is to fully and correctly detect an aborted build, quit
buildman immediately and not write any partial build results in this case.
Unfortunately this is currently beyond my powers and is left as an exercise
for the reader (and patches are welcome).
Signed-off-by: Simon Glass <sjg@chromium.org>
In accordance with our other modules supported by U-Boot and as agreed
upon for Apalis/Colibri T30 get rid of the carrier board in the board/
configuration/device-tree naming.
While at it also bring the prompt more in line with our other products.
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Tweak the output slightly so we don't get things like:
- board1 board2+ board3 board4
There should be a space before the '+'.
Signed-off-by: Simon Glass <sjg@chromium.org>
Commit d908898 updated the ScanPath() function but not its documentation
and not all its callers.
This breaks the toolchain check after it is downloaded. Fix it.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Heiko Schocher <hs@denx.de>
When buildman scans a toolchain path, it stops at the
first toolchain found. However, a single path can contains
several toolchains, each with its own prefix.
This patch lets buildman scan all toolchains in the path.
Signed-off-by: Albert ARIBAUD <albert.u.boot@aribaud.net>
Acked-by: Simon Glass <sjg@chromium.org>
Normally buildman runs with 'make -s' meaning that only errors and warnings
appear in the log file. Add a -V option to run make in verbose mode, and
with V=1, causing a full build log to be created.
Signed-off-by: Simon Glass <sjg@chromium.org>
The site at https://www.kernel.org/pub/tools/crosstool/ is a convenient
repository of toolchains which can be used for U-Boot. Add a feature to
download and install a toolchain for a selected architecture automatically.
It isn't clear how long this site will stay in the current place and
format, but we should be able to rely on bug reports if it changes.
Suggested-by: Marek Vašut <marex@denx.de>
Suggested-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
We should create a test setting file when running testes, not use whatever
happens to be on the local machine.
Signed-off-by: Simon Glass <sjg@chromium.org>
Since we need a few modules which might not be available in a bare-bones
distribution, add a note about that to the README.
Signed-off-by: Simon Glass <sjg@chromium.org>
Suggested-by: Wolfgang Denk <wd@denx.de>
In some cases there may be multiple toolchains with the same name in the
path. Provide an option to use the full path in the CROSS_COMPILE
environment variable.
Note: Wolfgang mentioned that this is dangerous since in some cases there
may be other tools on the path that are needed. So this is set up as an
option, not the default. I will need test confirmation (i.e. that this
commit fixes a real problem) before merging it.
Signed-off-by: Simon Glass <sjg@chromium.org>
Suggested-by: Steve Rae <srae@broadcom.com>
If:
1. Toolchains A and B have the same filename
2. Toolchain A is in the PATH
3. Toolchain B is given in ~/.buildman and buildman uses it to build
then buildman will add toolchain B to the end of its path but will not
necessarily use it since U-Boot will find toolchain A first in the PATH.
Try to fix this by putting the toolchain first in the path instead of
last.
Signed-off-by: Simon Glass <sjg@chromium.org>
The assumption that the compiler name will always end in gcc is incorrect
for clang and apparently on BSD.
Signed-off-by: Simon Glass <sjg@chromium.org>
Adjust the -b flag to permit a range expression as well as a branch.
Signed-off-by: Simon Glass <sjg@chromium.org>
Suggested-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Tested-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
When running tests the output directory is often wiped. This is only safe if
a branch is being built. The output directory may contain other things
besides the buildman test output.
Signed-off-by: Simon Glass <sjg@chromium.org>
When building current source for a single board, buildman puts the output
in <output_dir>/current/current/<board>. Add an option to make it use
<output_dir>/<board> instead. This removes the unnecessary directories
in that case, controlled by the --no-subdirs/-N option.
Suggested-by: Tom Rini <trini@ti.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Buildman normally obtains the upstream commit by asking git. Provided that
the branch was created with 'git checkout -b <branch> <some_upstream>' then
this normally works.
When there is no upstream, we can try to guess one, by looking up through
the commits until we find a branch. Add a function to try this and print
a warning if buildman ends up relying on it.
Also update the documentation to match.
Signed-off-by: Simon Glass <sjg@chromium.org>
Suggested-by: Wolfgang Denk <wd@denx.de>
This is not needed since we always do a full (non-incremental) build. Also
it might be dangerous since it will try to delete everything below the
base directory.
Fix this potentially nasty bug.
Signed-off-by: Simon Glass <sjg@chromium.org>
Buildman currently puts current-source builds in a current/current
subdirectory, but there is no need for the extra depth.
Suggested-by: Albert Aribaud <albert.u.boot@aribaud.net>
Signed-off-by: Simon Glass <sjg@chromium.org>
We probably don't need to enable this option by default. It is useful to
display only failure boards (not errors) and it is easy to add -e if it
is required. Also update the docs.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reported-by: Albert Aribaud <albert.u.boot@aribaud.net>
Ensure that we don't print duplicate board names when -l is used.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reported-by: Albert Aribaud <albert.u.boot@aribaud.net>
When saving binary files we likely want to keep any .img files that have
been generated as well.
Signed-off-by: Tom Rini <trini@ti.com>
Acked-by: Simon Glass <sjg@chromium.org>
When using summary mode (-s) we don't always want to display errors.
Allow this option to be omitted.
Series-to: u-boot
Series-cc: albert
Change-Id: I6b37754d55eb920ecae114fceba55834b43ea3b9
Signed-off-by: Simon Glass <sjg@chromium.org>
Reported-by: Albert Aribaud <albert.u.boot@aribaud.net>
Ensure that we don't print duplicate board names when -l is used.
Change-Id: I56adb138fc18f772ba61eba0fa194cdd7bc7efc6
Signed-off-by: Simon Glass <sjg@chromium.org>
Reported-by: Albert Aribaud <albert.u.boot@aribaud.net>
This check should now be done whatever mode buildman is running in, since
we may be displaying information while building.
Signed-off-by: Simon Glass <sjg@chromium.org>
Commit f219e01311 (tools: Import Kconfiglib)
added SPDX GPL-2.0+ to this library by mistake.
It should be ISC.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Ulf Magnusson <ulfalizer@gmail.com>
Kconfiglib is the flexible Python Kconfig parser and library
created by Ulf Magnusson.
(https://github.com/ulfalizer/Kconfiglib)
This commit imports kconfiglib.py from
commit ce84c22e58fa59cb93679d4ead03c3cd1387965e,
with ISC SPDX-License-Identifier.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Signed-off-by: Ulf Magnusson <ulfalizer@gmail.com>
Cc: Ulf Magnusson <ulfalizer@gmail.com>
Cc: Wolfgang Denk <wd@denx.de>
try run => dry run
no nothing => do nothing
"..." => '...'
The last one is for consistency with the other option helps.
Change-Id: I1d69047d1fae6ef095a18f69f44ee13c448db9b7
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
When creating build directories also create parents as necessary. This
fixes a failure when building a hierarchical branch (i.e. foo/bar).
Signed-off-by: Thierry Reding <treding@nvidia.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Tom Rini <trini@ti.com>
Tags like Series-version are normally expected to appear once, and with a
unique value. But buildman doesn't actually look at these tags. So ignore
conflicts.
This allows bulidman to build a branch containing multiple patman series.
Reported-by: Steve Rae <srae@broadcom.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
At present buildman naively uses the branch name as part of its directory
path, which causes problems if the name has an embedded '/'.
Replace these with '_' to fix the problem.
Reported-by: Steve Rae <srae@broadcom.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Now that buildman supports removing the build directory prefix from output,
add a test for it. Also ensure that output directories are removed when the
test completes.
Signed-off-by: Simon Glass <sjg@chromium.org>
This adds coverage of core features of the builder, including the
command-line options which affect building.
Signed-off-by: Simon Glass <sjg@chromium.org>
For testing it is useful to clean the output directory before running a
test. This avoids a test interfering with the results of a subsequent
test by leaving data around.
Add this feature as an optional parameter to the control logic.
Signed-off-by: Simon Glass <sjg@chromium.org>
When a build is to be performed, buildman checks to see if it has already
been done. In most cases it will not bother trying again. However, it was
not reading the return code from the 'done' file, so if the result was a
failure, it would not be counted. This depresses the 'failure' count stats
that buildman prints in this case.
Fix this bug by always reading the return code.
Signed-off-by: Simon Glass <sjg@chromium.org>
Rather than reading boards.cfg, which may take time to generate and is not
necessarily suitable for running tests, create our own list of boards.
Signed-off-by: Simon Glass <sjg@chromium.org>
These files may not exist in the environment, or may not be suitable for
testing. Provide our own config file and our own toolchains when running
tests.
Signed-off-by: Simon Glass <sjg@chromium.org>
Move the bsettings code back to the main buildman.py file, so we can do
something different when testing.
Signed-off-by: Simon Glass <sjg@chromium.org>
Buildman currently lacks testing in many areas, including its use of git,
make and many command-line flags.
Add a functional test which covers some of these areas. So far it does
a fake 'build' of all boards for the current source tree.
This version reads the real ~/.buildman and boards.cfg files. Future work
will improve this.
Signed-off-by: Simon Glass <sjg@chromium.org>
We want to be able to issue parser commands from within buildman for test
purposes. Move the parser code into its own file so we don't end up needing
the buildman and test modules to reference each other.
Signed-off-by: Simon Glass <sjg@chromium.org>
Adjust the basic test so that it checks all console output. This will help
to ensure that the builder is behaving correctly with printing summary
information.
Signed-off-by: Simon Glass <sjg@chromium.org>
To allow us to verify the builder's console output, send it through a
function which can collect it when running in test mode.
Signed-off-by: Simon Glass <sjg@chromium.org>
Some boards unfortunately build with warnings and it is useful to be able
to easily distinguish the warnings from the errors.
Use a simple pattern match to categorise gcc output into warnings and
errors, and display each separately. New warnings are shown in magenta (with
a w+ prefix) and fixed warnings are shown in yellow with a w- prefix.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add a -l option to display a list of offending boards against each
error/warning line. The information will be shown in brackets as below:
02: wip
sandbox: + sandbox
arm: + seaboard
+(sandbox) arch/sandbox/cpu/cpu.c: In function 'timer_get_us':
+(sandbox) arch/sandbox/cpu/cpu.c:40:9: warning: unused variable 'i' [-Wunused-variable]
+(seaboard) board/nvidia/seaboard/seaboard.c: In function 'pin_mux_mmc':
+(seaboard) board/nvidia/seaboard/seaboard.c:36:9: warning: unused variable 'fred' [-Wunused-variable]
+(seaboard) int fred;
+(seaboard) ^
Signed-off-by: Simon Glass <sjg@chromium.org>
The full path is long and also includes buildman private directories.
Clean this up, so that only a relative U-Boot path is shown.
This will change warnings like these:
/home/sjg/c/src/third_party/u-boot/buildman5/.bm-work/00/arch/sandbox/cpu/cpu.c: In function 'timer_get_us':
/home/sjg/c/src/third_party/u-boot/buildman5/.bm-work/00/arch/sandbox/cpu/cpu.c:40:9: warning: unused variable 'i' [-Wunused-variable]
/home/sjg/c/src/third_party/u-boot/files/arch/sandbox/cpu/cpu.c: In function 'timer_get_us':
/home/sjg/c/src/third_party/u-boot/files/arch/sandbox/cpu/cpu.c:40:9: warning: unused variable 'i' [-Wunused-variable]
to:
arch/sandbox/cpu/cpu.c: In function 'timer_get_us':
arch/sandbox/cpu/cpu.c:40:9: warning: unused variable 'i' [-Wunused-variable]
Signed-off-by: Simon Glass <sjg@chromium.org>
Some boards are known to be broken and it is convenient to be able to
exclude them from the build.
Add an --exclude option to specific boards to exclude. This uses the
same matching rules as the normal 'include' arguments, and is a comma-
separated list of regular expressions.
Suggested-by: York Sun <yorksun@freescale.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
These characters are commonly used in variables, so permit them. Also
document the permitted characters.
Reported-by: Tom Rini <trini@ti.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
When buildman finds errors/warnings when building, set the return code to
indicate this.
Suggested-by: York Sun <yorksun@freescale.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
This commit makes sure boards.cfg is up to date before starting
the build tests. tools/genboardscfg.py exits immediately printing
"boards.cfg is up to date. Nothing to do." when boards.cfg is
already new.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
"buildman [options]" is displayed by default.
Append the rest of help messages to parser.usage
instead of replacing it.
Besides, "-b <branch>" is not mandatory since commit fea5858e.
Drop it from the usage.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Signed-off-by: Roger Meier <roger@bufferoverflow.ch>
Reviewed-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Tested-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Tom Rini <trini@ti.com>
Cc: Wolfgang Denk <wd@denx.de>
Cc: Simon Glass <sjg@chromium.org>
In Python, sys.exit() function can also take an object other
than an integer.
If an integer is given to the argument, Python exits with the return
code of it. If a non-integer argument is given, Python outputs it
to stderr and exits with the return code of 1.
That means,
print >> sys.stderr, "Blah Blah"
sys.exit(1)
is equivalent to
sys.exit("Blah Blah")
The latter is a useful shorthand.
Note:
Some error messages in Buildman and Patman were output to stdout.
But they should go to stderr. They are also fixed by this commit.
This is a nice side effect.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
It is useful to be able to build only some of the commits in a branch. Add
support for the -c option to allow this. It was previously parsed by
buildman but not implemented.
Suggested-by: York Sun <yorksun@freescale.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: York Sun <yorksun@freescale.com>
Currently buildman allows a list of boards to build to be specified on the
command line. The list can include specific board names, architecture, SOC
and so on.
At present the list of boards is dealt with in an 'OR' fashion, and there
is no way to specify something like 'arm & freescale', meaning boards with
ARM architecture but only those made by Freescale. This would exclude the
PowerPC boards made by Freescale.
Support an '&' operator on the command line to permit this. Ensure that
arguments can be specified in a single string to permit easy shell quoting.
Suggested-by: York Sun <yorksun@freescale.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: York Sun <yorksun@freescale.com>
The current README is a bit sparse in this area, so add a few more
examples.
Suggested-by: Tom Rini <trini@ti.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
If buildman finds no problems it prints nothing. This can be a bit confusing,
so add a message that all is well.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add a new --config-file option (-G) to specify a different configuration
file from the default ~/.buildman.
Reported-by: Tom Rini <trini@ti.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Normally buildman operates in two passes - one to do the build and another
to summarise the errors. Add a verbose option (-v) to display build problems
as they happen. With -e also given, this will display errors too.
When building the current source tree (rather than a list of commits in a
branch), both -v and -e are enabled automatically.
Signed-off-by: Simon Glass <sjg@chromium.org>
We need the output options to be available in several places. It's a pain
to pass them into each function. Make them properties of the builder and
add a single function to set them up. At the same time, add a function which
produces summary output using these options.
Signed-off-by: Simon Glass <sjg@chromium.org>
Originally buildman had some support for building the current source tree.
However this was dropped before it was submitted, as part of the effort to
make it faster when building entire branches.
Reinstate this support. If no -b option is given, buildman will build the
current source tree.
Reported-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Use "make <board>_defconfig" instead of "make <board>_config".
Invoke tools/genboardscfg.py to generate boards.cfg when it is missing.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
Since the command name 'make' may not be GNU Make on some platforms
such as FreeBSD, buildman should call scripts/show-gnu-make to get
the command name for GNU MAKE (and error out if it is not found).
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Jeroen Hofstee <jeroen@myspectrum.nl>
At present buildman always builds out-of-tree, that is it uses a separate
output directory from the source directory. Normally this is what you want,
but it is important that in-tree builds work also. Some Makefile changes may
break this.
Add a -i option to tell buildman to use in-tree builds, so that it is easy
to test this feature.
Signed-off-by: Simon Glass <sjg@chromium.org>
Normally buildman wil try to configure U-Boot for a particular board on the
first commit that it builds in a series. Subsequent commits are built
without reconfiguring which normally works. Where it doesn't, buildman
automatically reconfigures and retries.
To fully emulate the way MAKEALL works, we should have an option to disable
this optimisation.
Add a -C option to cause buildman to always reconfigure on each commit.
Signed-off-by: Simon Glass <sjg@chromium.org>
After a build fails buildman will reconfigure and try again, if it did not
reconfigure before the build. However it doesn't actually keep track of
whether it did reconfigure on the previous attempt.
Fix that logic to avoid a pointless rebuild. This speeds things up quite a
bit for failing builds. Previously they would always be built twice.
Change-Id: Ib37f21320baa7c60bed98f4042c0b7ed7c0dc85e
Signed-off-by: Simon Glass <sjg@chromium.org>
Generally a build failure with a particular commit cannot be fixed except
by changing that commit. Changing the commit will automatically cause
buildman to retry when you run it again: buildman sees that the commit
hash is different and that it has no previous build result for the new
commit hash.
However sometimes the build failure is due to a toolchain issue or some
other environment problem. In that case, retrying failed builds may yield
a different result.
Add a flag to retry failed builds. This differs from the force rebuild
flag (-f) in that it will not rebuild commits which are already marked as
succeeded.
Series-to: u-boot
Change-Id: Iac4306df499d65ff0888b1c60f06fc162a6faad8
'-elf' appears twice in the toolchain priority_list.
The second one is rudundant.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Simon Glass <sjg@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
Toolchains.__init__ is expected to display a warning message
when the [toolchain] section is missing from ~/.buildman file.
But it never works.
In that case, instead, buildmain fails with an error message
which is difficult to understand:
Traceback (most recent call last):
File "tools/buildman/buildman", line 126, in <module>
control.DoBuildman(options, args)
File "/home/foo/u-boot/tools/buildman/control.py", line 78, in DoBuildman
toolchains = toolchain.Toolchains()
File "/home/foo/u-boot/tools/buildman/toolchain.py", line 106, in __init__
config_fname)
NameError: global name 'config_fname' is not defined
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Simon Glass <sjg@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
Add an option to specify the output directory to override the
default path '../'. This is useful for building in a ramdisk.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
A common use-case is to build all boards for a particular SoC. This can
be achieved by:
./tools/buildman/buildman -b mainline_dev tegra20
However, when the SoC is a member of a family of SoCs, and each SoC has
a different name, it would be even more useful to build all boards for
every SoC in that family. This currently isn't possible since buildman's
board selection command-line arguments are compared to board definitions
using pure string equality.
To enable this, compare using a regex match instead. This matches
MAKEALL's handling of command-line arguments. This enables:
(all Tegra)
./tools/buildman/buildman -b mainline_dev tegra
(all Tegra)
./tools/buildman/buildman -b mainline_dev '^tegra.*$'
(all Tegra20, Tegra30 boards, but not Tegra114)
./tools/buildman/buildman -b mainline_dev 'tegra[23]'
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Acked-by: Simon Glass <sjg@chromium.org>
python used in buildman doesn't need to be placed in
/usr/bin/python, So use env to ensure that the interpreter
will pick the python from environment.
Usefull with several versions of python's installed on system.
Signed-off-by: Jagannadha Sutradharudu Teki <jaganna@xilinx.com>
Acked-by: Simon Glass <sjg@chromium.org>
When a toolchain invocation fails, an exception is thrown but not caught
which then aborts the entire toolchain detection process. To solve this,
request that exceptions not be thrown, since the toolchain init code
already error-checks the command result. This solves e.g.:
- found '/usr/bin/winegcc'
Traceback (most recent call last):
...
Exception: Error running '/usr/bin/winegcc --version'
Change-Id: I579c72ab3b021e38b14132893c3375ea257c74f0
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Simon Glass <sjg@chromium.org>
(formatted to 80cols)
There are a few make options such as BUILD_TAG which can be provided when
building U-Boot. Provide a way for buildman to pass these flags to make
also.
The flags should be in a [make-flags] section and arranged by target name
(the 'target' column in boards.cfg. See the README for more details.
Signed-off-by: Simon Glass <sjg@chromium.org>
Commit 27af930e9a changed the boards.cfg format
but missed to change the parsing in buildman. A follow-on commit
03c1bb2425 fixed this but missed fixing the
tests.
This patch updates the tests to fit the new Board constructor.
./tools/buildman/buildman -t
<unittest.result.TestResult run=1 errors=0 failures=0>
Signed-off-by: Simon Glass <sjg@chromium.org>
Commit 27af930e9a changed the boards.cfg format
but missed to change the parsing in buildman.
This patch changes c'tor of Board class to the new sequence, but omits
maintainer field.
Signed-off-by: Andreas Bießmann <andreas.devel@googlemail.com>
Put all informations about targets, including state (active or
orphan) and maintainers, in boards.cfg; remove MAINTAINERS;
adjust the build system accordingly.
Signed-off-by: Albert ARIBAUD <albert.u.boot@aribaud.net>
Conflicting tags can prevent buildman from building two series which exist
one after the other in a branch. There is no reason not to allow this sort
of workflow with buildman, so ignore conflicting tags in buildman.
Change-Id: I2231d04d8684fe0f8fe77f8ea107e5899a3da5e8
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@ti.com>
This tool handles building U-Boot to check that you have not broken it
with your patch series. It can build each individual commit and report
which boards fail on which commits, and which errors come up. It also
shows differences in image sizes due to particular commits.
Buildman aims to make full use of multi-processor machines.
Documentation and caveats are in tools/buildman/README.
Signed-off-by: Simon Glass <sjg@chromium.org>
