It is not uncommon for some of the QEMU-based jobs to fail not because
of a code issue but rather because of a timing issue or similar problem
that is out of our control. Make use of the keywords that Azure and
GitLab provide so that we will automatically re-run these when they fail
2 times. If they fail that often it is likely we have found a real issue
to investigate.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
As this is the current version of the public cross toolchains we use,
upgrade to this now.
Suggested-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Acked-by: Alexey Brodkin <abrodkin@synopsys.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Use the latest OpenSBI v1.2 release binaries for the RISC-V CI.
Signed-off-by: Bin Meng <bmeng@tinylab.org>
Reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com>
We do not want to merge documentation that produces Sphinx warnings.
scripts/kernel-doc uses environment variable KDOC_WERROR to determine
if warnings should be treated as errors.
Reported-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Add M5208EVBE board to CI. This does not use default config due to
limitations of QEMU emulation, instead the timer is switched from
DMA timer to PIT timer and RAMBAR accesses are inhibited.
Local QEMU launch command is as follows:
$ qemu-system-m68k -nographic -machine mcf5208evb -cpu m5208 -bios u-boot.bin
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Acked-by: Angelo Dureghello <angelo@kernel-space.org>
As this is now the stable release, move to using that now for our tests.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
To conserve bandwidth and potentially avoid rate limits, allow a local
mirror of Docker Hub to be specified globally. The default value is
unchanged.
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
The default behavior of Gitlab runners is to only run jobs which match
the configured tag, although there is an option to run untagged jobs
[1].
To support running the CI in more complex environments where different
types of runners may be present that support different tags, allow the
DEFAULT_TAG for all jobs in the pipeline to be set globally using an
environment variable. An empty default value is provided to retain
support for untagged runners.
[1] https://docs.gitlab.com/ee/ci/runners/configure_runners.html#use-tags-to-control-which-jobs-a-runner-can-run
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Copy build artifacts for all test.py tests, so they show up in
artifacts storage for later inspection. The test.py tests output
in CI is basically useless, but it is far more useful in the html
output for analysis and debugging.
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Suggested-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Now that the Dockerfile creates images which have the binaries we
require included, have CI make symlinks for them and update the existing
script to support this.
Signed-off-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Pali Rohár <pali@kernel.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
These should not be used anymore. Add a check to ensure they don't creek
back into U-Boot. Use bootph-... instead.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add a test which runs sandbox, collects a trace and makes sure it can
be processed by trace-cmd. This should ensure that this feature continues
to work as U-Boot and trace-cmd evolve.
Signed-off-by: Simon Glass <sjg@chromium.org>
Now that all remaining in-tree cases where we define or undef a CONFIG
symbol have been migrated to Kconfig or renamed to CFG we can make the
CI check more robust. We will exclude the doc, tools and arch/arm/dts
directories from this check as they are special cases. Further, we can
exclude the scripts/kconfig/lkc.h and include/linux/kconfig.h files as
the CONFIG values they define are special tooling cases and not real
symbols.
In the case of docs, the only places that currently fail this test are
old documentation that should be rewritten so that we can remove this
special case.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Add infodocs target to CI testing.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Update to the latest "jammy" tag. This requires us to list libc6-i386 as
a required package to install (for nokia_rx51 tests) that was previously
implicit.
Signed-off-by: Tom Rini <trini@konsulko.com>
Now that all symbols have been migrated to Kconfig, or are part of the
CFG namespace we do not need a complex check for unmigrated CONFIG
symbols. Any instance of #define (or #undef) or a CONFIG value is wrong,
so cause CI to fail.
This test is not as strict as possible yet as we have more symbols that
were not previously caught to deal with.
Signed-off-by: Tom Rini <trini@konsulko.com>
- Update to gcc-12.2, and cherry-pick a fix in grub for risc-v
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
We have a number of jobs that will have git complain about needing to
set safe.directory and this being untrue as a fatal error, but then
complete. Set this flag correctly now as it should be used, and may
prevent a future failure.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Add a new flag to buildman so that we will in turn pass
BINMAN_ALLOW_MISSING=1 to 'make'. Make use of this flag in CI.
Allow the settings file to control this.
Cc: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Cc: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
For building riscv32 targets we should use the riscv32 toolchain.
Add it to the Docker image.
Drop the riscv toolchain-alias as we do not need it in future.
While in here, update to the latest "jammy" tag.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Rick Chen <rick@andestech.com>
[trini: Update to latest jammy tag]
Signed-off-by: Tom Rini <trini@konsulko.com>
- We now have a new enough sbsigntools in the distro, stop building.
- Use the 20220801 tag for Jammy.
- Move to pygit2 1.9.2 (current version) as the old one doesn't build on
"Jammy".
- Add the working directory to the list of safe directories for git.
- Move to pytest 6.2.5 to address other issues.
- This move exposed a number of minor issues in the existing scripts we
used within CI to perform the jobs themselves. The most notable changes
here involve using 'set +e / set -e' to enforce when we should or should
not make non-zero buildman status be a fatal error.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Use the equivalent buildman functionality to check maintainer info.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
The AST2600 has a Qemu model that allows testing. Create a SPI NOR image
containing the combined SPL and u-boot FIT image.
Reviewed-by: Chia-Wei Wang <chiawei_wang@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Joel Stanley <joel@jms.id.au>
To make addressing the problem of migrated symbols being present in
board config header files, update the CI test to them print what symbols
are causing it to fail. Also report all failures in the tree, rather
than stopping at the first failing file.
Signed-off-by: Tom Rini <trini@konsulko.com>
For consistency in runs, we need to always use the same pylint version.
Pin to 2.12.2 as this is what we have been using so far.
Signed-off-by: Tom Rini <trini@konsulko.com>
When calling comm to compare the CONFIG symbols a defconfig uses with
the symbols that have been migrated, we need to suppress all output as
the summary line will have everything we need. Failure to do this leads
to the test blowing up, but in non-fatal ways.
Signed-off-by: Tom Rini <trini@konsulko.com>
Coreboot supports U-Boot as a payload and this recently got a bit of a
facelist. Add a test for this.
For now this uses a binary build of coreboot (v4.15). Future work could
potentially build it from source, but we need to figure out the
toolchain problems first, since coreboot uses its own toolchain. It
turns out that this is tricky, because coreboot fails to build with a
vanilla gcc.
This needs some changes to the hooks scripts as well. An example build
is at https://source.denx.de/u-boot/custodians/u-boot-dm/-/jobs/359687
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Now that all symbols that exist in Kconfig no longer also have boards
setting them in the board config.h file, add a CI test to catch new
instances of this, and fail.
Signed-off-by: Tom Rini <trini@konsulko.com>
Instead of fetching an arm toolchain to use, run the test with the one
that's already in the container image.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Rebuild our current docker image so that ca-certificates will be updated
and Let's Encrypt issued certificates will work again.
Signed-off-by: Tom Rini <trini@konsulko.com>
vexpress_ca9x4 is seemingly the only board except for qemu_arm which
is able to run U-Boot correctly, using the `-M vexpress-a9` option to
QEMU. Building for qemu_arm and running qemu-system-arm with the `-M
virt` argument has a number of downsides, most importantly that it
only supports virtio storage drivers. This significantly reduces its
usefulness in testing memory card and Flash solutions, especially when
the tested images are from a third party source.
So therefore we reintroduce the vexpress_ca9x4 board in this commit,
with the explicit goal of using it with QEMU.
A number of differences to note from the original:
* Since the board was apparently unmaintained, I have now set myself
as the maintainer.
* The board has been converted to use the driver model, which was the
reason it was removed in the first place.
* The vexpress_ca15_tc2 and vexpress_ca5x2 boards, which were removed
in the same commit, are not necessary for the QEMU use case, and
have been omitted.
* An `mmc0` alias was introduced in the dts file. The mmc is not
detected correctly without this, now that it's based on the device
tree instead of the board's init function.
* A couple of other nodes were removed because they were problematic
when trying to run the UEFI bootmgr. Once again, the primary use
case here is QEMU, and these nodes are not needed for that to work.
* Unnecessary board init code has been removed, thanks to driver model
and device tree.
* `CONFIG_OF_EMBED` has been enabled. I know this goes against
recommended practice, but there doesn't seem to be any other way to
pass the dtb to U-Boot in the QEMU scenario. Using the -dtb argument
does not work, I suppose because U-Boot doesn't use the same
mechanics as the kernel when it's booting.
* Load addresses have been changed to fit QEMU use case.
People wanting to get a more detailed, yet somewhat isolated, diff
between this and the original, can run this command:
git diff c6c26a05b89f25a06e7562f8c2071b60fd0c9eac~1 -- \
$( git diff-tree --diff-filter=A -r --name-only HEAD~1 HEAD)
(Make sure to either check out this commit first, or replace HEAD with
the commit ID of this commit)
Signed-off-by: Kristian Amlie <kristian.amlie@northern.tech>
This adds CI tests for SiFive Unleashed board.
QEMU supports booting exact the same images as used on the real
hardware out of the box, that U-Boot SPL loads U-Boot proper
from either an SD card or the SPI NOR flash, hence we can easily
set up CI to cover these 2 boot flows of SiFive Unleashed board.
With this, now we can have regression testing of mmc-spi-slot and
sifive spi drivers, as well as mmc and spi-nor subsystems.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
