The GCS support patches are based on GDB 15.
Copy the GDB 14.2 recipe from meta, upgrade to 15.1, and add the patches.
Signed-off-by: James McGregor <james.mcgregor2@arm.com>
Update the GCS patches to v12.
They've been rebased onto 6.11-rc3, so update to use that too.
GCS support now depends on !UPROBES, so add a config fragment to disable it.
Signed-off-by: James McGregor <james.mcgregor2@arm.com>
The values prctl() expects for shadow stack operations have changed.
Update the glibc patch to match the values in the kernel.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Package up the kselftest from the kernel, and execute the arm64 GCS tests
inside CI.
At the moment some of the tests fail: gcs_write_fault and gcs_stress.
Mark the test as expected to fail until this is resolved.
Signed-off-by: Ross Burton <ross.burton@arm.com>
The bit index for HWCAP2_GCS changed, so update the glibc patch to
reflect the number used in the kernel.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Change gcs-test.yml so that it just enables automatic testimage and does
not duplicate options.
Also change debug-tweaks to the specific features that we want.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Add a basic job for testing the GCS patches that runs testimage with the
gcs test suite.
For speed, disable all the other jobs.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Currently the run_cmd, which is a wrapper for self.target.run()
that uses SSH to spawn commands on the target, can fail spuriously
with error 255 and cause the test to fail on slow systems.
In order to address that, introduce a retry mechanism for the call,
that is able to wait some time for the system to settle and retry
the command when the error code from SSH is 255.
Signed-off-by: Luca Fancellu <luca.fancellu@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Define “DISTRO_UNATTENDED_INST_TESTS” variable in meta-arm-systemready
independently from meta-arm-auto-solutions. This will allow running
the unattended installation without meta-arm-auto-solutions.
Signed-off-by: Amr Mohamed <amr.mohamed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The oeqa test responds to the boot loader prompt error message and
waits till the distro installation is finished.
Signed-off-by: Amr Mohamed <amr.mohamed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add the Fedora kickstart configuration file and define a function to
modify the unpacked ISO image to add the kickstart file inside and
modify the grub.cfg file.
Signed-off-by: Amr Mohamed <amr.mohamed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add a new inc file to unpack and repack the distro ISO image after
adding the kickstart configuration file inside.
Signed-off-by: Amr Mohamed <amr.mohamed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Change defaults repo refspec in fvp-base.yml file from master to
scarthgap.
Signed-off-by: Amr Mohamed <amr.mohamed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This patch enables logging with timestamps for individual pexpect
assertions to ease the debugging of failed tests and the tuning of
timeouts. It measures the execution time of all pexpect calls and logs
the actual duration for each.
Only "callable" pexpect calls are timed (e.g. expect, sendline, but not
before or after).
Signed-off-by: Divin Raj <divin.raj@arm.com>
Signed-off-by: Ziad Elhanafy <ziad.elhanafy@arm.com>
Signed-off-by: Peter Hoyes <peter.hoyes@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This procps fix has been merged upstream in oe-core aaced482, so we can
remove this patch now.
This reverts commit fef5eafc08.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Add Fedora distribution version 39.1.5 installation to fulfill
the SystemReady IR.
Signed-off-by: Amr Mohamed <amr.mohamed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Improve the documentation in the user guide of the following tests:
- SystemReady-IR tests
- Manual capsule update and ESRT checks
- Linux distros tests
- UEFI Secureboot (SB) test
- PSA API tests
In addition, we moved the tests in one section for better readability.
Signed-off-by: Delane Brandy <delane.brandy@arm.com>
Signed-off-by: Harsimran Singh Tungal <harsimransingh.tungal@arm.com>
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This flag should not be set here and the ARM_FVP_EULA_ACCEPT
should be set to True manually before building for the FVP, as it is
mentioned in the Corstone-1000 User guide:
export ARM_FVP_EULA_ACCEPT="True"
Fixes: 6e2a54748 ("kas: Corstone-1000 kas files updated")
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The BB_HASHSERVE_UPSTREAM has issues which cause significantly less of a
match than expected. Update with the correct values to get the expected
behavior.
Fixes: 6e9525115b ("CI: add Yocto Project SSTATE Mirror")
Signed-off-by: Jon Mason <jon.mason@arm.com>
The reset has to be removed from the TF-M side after capsule update
because it caused data abort exceptions on the host side.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
The default branch will be scarthgap so the poky and
openembedded will follow this branch.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The buffer size has to be increased to fit the EFI variables which got
increased metadata sizes.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The increased EFI variable metadata need bigger buffer so it can
be transfered to the Secure Enclave without memory overflow
issues. The heap and buffer sizes had to be aligned with the.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The private authenticated variable changes increased the variables
metadata. The PS max asset size and related buffer sizes have to be
increased because of this.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
OPTEE and ftpm tests are failing in CI on slower systems due to timing
out, but actually finish when given enough time to complete. Increase
the timeout value to be roughly 100 seconds longer than the time it is
currently taking to finish on the slower systems.
Fixes: d450786667 ("oeqa runtime: add optee.py test")
Fixes: ba315f7242 ("oeqa runtime: add ftpm.py test")
Signed-off-by: Jon Mason <jon.mason@arm.com>
Upgrade the openSUSE distribution from version 15.4 to version 15.5
openSUSE Licenses updated to reflect update from 15.4 to 15.5
License now includes: Apache-1.1, BSL-1.0, IPL-1.0, Sleepycat, Zlib
Signed-off-by: Ben Cownley <ben.cownley@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This changeset updates the user guide to test the secureboot for both the
FVP and FPGA.
Signed-off-by: Harsimran Singh Tungal <harsimransingh.tungal@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The psa-iat-api-test was failing because the PLATFORM_HAS_ATTEST_PK
flag was added to the build for Corstone1000.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Crypto-AEAD-APIs tests fails on mps3. Configures CC312 mps3 model
same as predefined cc312 FVP configuration while keeping debug
ports closed.
Signed-off-by: Ali Can Ozaslan <ali.oezaslan@arm.com>
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Corstone1000 does not properly clean the cache and disable gic interrupts
before the reset. This causes a race condition especially in FVP after reset.
This adds proper sequence before resetting the platform.
Signed-off-by: Emekcan Aras <Emekcan.Aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Some networks limit outgoing git: traffic, so use https:.
Fixes: 0cec3e5 ("arm/gem5/boot-wrapper-aarch64: Move main recipe to meta-arm")
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Includes TS and PSA dependency for firmware image build.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Recent changes to the FVP Base homepage have caused it to no longer be
searchable with the current Yocto tooling. Disable it to prevent issues
with `devtool check-upgrade-status`.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add corstone1000-recovery-image image based on core-image-minimal
while disabling the testimage task which is irrelevant in case of
an initramfs bundle.
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Ross Burton <ross.burton@arm.com>
Corstone-1000 no longer uses OpenAMP, and it was the only platform
which needed this library.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Corstone-1000 no longer uses OpenAMP, and it was the only platform
which needed this library.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The OpenAMP is replaced by the RSE Communication Protocol and
the documentation had to by updated to reflect this change.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Added the Corstone-1000 FVP platform to the ACS test build as well as
adding the arm-systemready-firmware variant to the Corstone-1000 FVP
build.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update the Corstone-1000 user guide with the new instructions on how to
build/use an ESP image and how to use the meta-arm-systemready layer to
run the ACS tests.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The SystemReady IR ACS test suite require that there is a valid ESP
partition available to the system. This change creates a new image that
only contains a ESP partition and ensures it's mounted on the second MMC
card so it's available when the SystemReady tests run.
The diagnostic level of the 2 MMC cards have also been lowered to
improve the ACS test duration.
Corrected a spelling mistake in the corstone1000-flash-firmware-image.bb
file.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Added the missing meta-arm-systemready required variable to enable its
use with the corstone1000-fvp machine. Also explicitly set all the
consoles.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The nodistro settings in poky set the TMPDIR variable to include the
TCLIBC value so we need to spot that and swap the TCLIBC for the musl
one used in the firmware multiconfig.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The regex used to validate compatible machines is incorrect as it's only
checking the machine name starts with "fvp" not "fvp-" as intended.
It's also been modified to allow FVPs called xxx-fvp to be compatible
with Corstone-1000.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
After enabling additional features in Trusted Services, the size of BL32
image (OP-TEE + Trusted Services SPs) is larger now. To create more space
in secure RAM for BL32 image, this patch removes NS_SHARED_RAM region which
is not currently used by corstone1000 platform.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This change enables the UEFI secure boot and its related configurations
for corstone1000
Signed-off-by: Harsimran Singh Tungal <harsimransingh.tungal@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The .nopt and capsule are generated during the yocto build. Sync the
documentation with the changes.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Currently, only the Corstone-1000 platform uses the capsule generation
class. Corstone-1000 uses U-Boot instead of EDK2. With this change,
the dependency on EDK2 was removed.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The .nopt image is used during the UEFI Update Capsule generation.
This .nopt image was generated manually when it was needed.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
optee-os test xtest needs additional test trusted applications (TA) from
optee-os-ta package to pass. Execution time for ftpm test is around 21
seconds and 596 seconds for optee-test/xtest on an x86_64 build machine.
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
optee-os test xtest needs additional test trusted applications (TA) from
optee-os-ta package to pass. Execution time for ftpm test is around 18
seconds and 430 seconds for optee-test/xtest on an x86_64 build machine.
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Test checks that ftpm kernel driver interfaces are available.
If fTPM optee TA is missing or crashes, the kernel driver does not
show the interfaces. A more functional tests would be to use tpm2-tools
from meta-security/meta-tpm but those require additional layer
dependencies which are maybe too much for now. tpm2-tools also depend
on starting tpm2-abrmd before the tools work. The ftpm kernel driver
depends on fully running tee-supplicant in userspace and the optee
side ftpm TA which takes some time. When manually running the tests
some of them failed since ftpm was not yet initialized. The boot
was not complete in those cases so added a workaround for that.
Better would be for all of the tests to start only once boot is
complete, not when ssh is available. Also, the qemuarm64-secureboot
machine includes optee and ftpm TA but does u-boot is not configured
to use the TPM device so boot is not measured.
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
If firmware TPM TA is compiled into optee, it needs a bit more
heap to pass optee-test/xtest suite.
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
All other firmware boot components also continue booting
if TPM is not found. It is up to subsequent SW components
to e.g. fail if rootfs can't be decrypted. Enables policies
like fall back to unencrypted rootfs if TPM device is
not found with qemu and swtpm.
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
There are two recipes in meta-arm-systemready that download ISOs for
testing purposes. Build them in CI to verify that the fetch is
successful.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
These recipes look like images but are not constructed the same way,
specifically there is no WORKDIR/rootfs/ directory. If buildhistory is
enabled this will cause it to abort, so disable image data collection in
buildhistory.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
We recently switched the CI to not disable ptest, but this breaks builds
that use the GCC binaries built by Arm (external-arm-toolchain). This is
because the external-arm-toolchain recipe can't build packages for the
target, and the standard oe-core gcc recipes assume that they're being
built with themselves and make assumptions, specifically that libunwind
was enabled and headers can be copied directly from the sysroot.
This is a bigger problem that should be solved somehow, but for now we
can just remove ptest in the external-gccarm CI jobs which removes gcc
from the builds (it comes in via elfutils-ptest RDEPENDS).
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Functions that take 32-bit time_t types are unavoidable in the libc, so
ignore the warnings.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Reviewed-by: Denys Dmytriyenko <denis@denix.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This patch will be merged upstream soon, apply it locally to unblock CI.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add YAML language server comments so that IDEs know what schema to use
for the Kas files.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
trusted_services.py:test_15_crypto_service runs ts-service test with
an incorrect argument list. The -g argument does not accept two group
names. This resulted in a silent failure.
Fix this by relying the pattern matching capability of the argument.
Additionaly remove references to OP-TEE from test messages as TS tests
are SPMC agonistic.
Signed-off-by: Gyorgy Szing <gyorgy.szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update TS and dependencies to latest version of the integration branch.
Remove patches merged upstream.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add a recipe to enable building and deploying the FWU service
implemented in the Trusted Services Project. The FWU service can
help vendors to meet PSA certification requirements.
Signed-off-by: Gyorgy Szing <gyorgy.szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Due to how the timer in u-boot is implemented, it's quite possible for
a two second timeout in the u-boot login to actually take over 15s to
expire.
Take a patch from the mailing list to implement this differently so the
timer runs in an accurate amount of time.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
add support for the remoteproc control feature for the external system
With this feature we can switch on/off the external system on demand:
echo stop > /sys/class/remoteproc/remoteproc0/state
echo start > /sys/class/remoteproc/remoteproc0/state
During Linux boot the remoteproc subsystem automatically start
the external system. The user can use the commands above to
stop then start the remote core.
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
describe the external system as a remoteproc node in the device tree
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
install the external system binaries under /lib/firmware
The kernel's remoteproc subsystem expects the firmware file to be under /lib/firmware
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
When trying to replicate a build locally, having the exact list of Kas
files that was used is very useful.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add the Yocto Project public SSTATE mirror to its own unique yml file.
This allows for developers to use this to speed up builds, while not
adding in the default case. This "off by default" is because it can add
10s of minutes to each build, which might not be beneficial to those who
are using SSTATE dir locally.
Also, removing the removal ptest distro feature, as this change prevents
an optimal usage of the YP SSTATE mirror (~30% match to ~90% match for
qemuarm64).
Signed-off-by: Jon Mason <jon.mason@arm.com>
These BSPs are now obsolete.
Users of generic-arm64 should use genericarm64 from meta-yocto-bsp.
Users of qemu-generic-arm64 should use sbsa-ref from meta-arm-bsp.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This is mostly based on the existing qemu-generic-arm64 machine, but by
not being based on the genericarm64 and instead being specifically a
machine to run on the qemu sbsa-ref machine we get to tune differently.
Specifically, this configures sbsa-ref to be a Neoverse N2 (v9), and the
tune is set to match. Another notable difference to qemu-generic-arm64
is that the kernel configuration is at present defconfig. We may wish
to change this in the future to be the same fragmented configuration as
genericarm64.
We have to ignore two testimage parselogs failures: one from NUMA which
will be fixed in a future EDK2 release, and one from efifb where we
should be using the bochsdrm driver instead (further investigation is
needed)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add the new genericarm64 in meta-yocto-bsp to the CI.
This new BSP is heavily based on the meta-arm generic-arm64 machine, but
with an all-new fragmented kernel configuration.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Some BSPs use a proper initramfs and putting a SSH server into them
via this :append isn't ideal. Adding using += should be sufficient.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Every platform should have the chance to try the -rt patches.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This class is no longer used by any machines, so remove it.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
- The Secure Enclave Proxy Secure Partition fails at psa_call()
because wrong parameter was passed.
- The SMM Gateway initialization failed because a malloc()
returned a NULL pointer. The SMM_GATEWAY_MAX_UEFI_VARIABLES
had to be decreased to avoid this.
- Increase shared memory buffer size and add buildtime check
- Use __packed for the variable_metadata struct
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The Trusted Services v1.0 uses new RPC protocol and the message
fields in u-boot had to be synchronized.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add configuration settings to TF-A, OP-TEE and TS SPs needed to get TS
built and run on the fvp-base machine.
Signed-off-by: Gyorgy Szing <gyorgy.szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Version v11.25 was released and it fixes measured boot. Update the
recipe and integrate the new version.
The pattern of the download URL has changed. Add functionality to
calculate a new URL fragment from the package version.
Signed-off-by: Gyorgy Szing <gyorgy.szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The current version of the TS recipes fails to build if the TS
environment is not set to opteesp. Change the recipes to allow building
the sp environment.
This environment targets "generic" secure partitions and produces SPMC
agnostic SP binaries which should be able to boot under any FF-A v1.0
compliant SPMC implementation.
Signed-off-by: Gyorgy Szing <gyorgy.szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
MbedTLS fails to build when FORTIFY_SOURCE is enabled and the NWd
configuration is used. Disable the compilation option temporary till
the root cause can be fund and a proper fix be made.
The build only fails when building from yocto. The OP-TEE integration
works fine with gcc v13.2_rel1.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
- Update Trusted Services to v1.0.0.
- Update TS "external components" references to fetch the version
dictated by the TS repo.
- Remove patches merged up-stream.
- Update the TS nanopb integration fix (see 210a6ace83)
- Update TS test integration.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
- Update driver version to v2.0.0
- Follow up the name change. The driver has been renamed from
arm_ffa_tee to arm_tstee.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The .nopt and capsule are generated during the yocto build. Sync the
documentation with the changes.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Currently, only the Corstone-1000 platform uses the capsule generation
class. Corstone-1000 uses U-Boot instead of EDK2. With this change,
the dependency on EDK2 was removed.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The .nopt image is used during the UEFI Update Capsule generation.
This .nopt image was generated manually when it was needed.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
A SRCREV for arm-platforms-kmeta was added years ago to get
yocto-check-layer working at the time, but was never removed (and never
updated). Removing now, since it is not necessary.
Signed-off-by: Jon Mason <jon.mason@arm.com>
A patch was dropped when trusted-firmware-m was updated to 2.0 but it
had not yet been merged upstream (2.0 or master).
Restore the patch to fix regression on Corstone-1000
Signed-off-by: Jon Mason <jon.mason@arm.com>
Upgrade the Debian distribution from version 11.7 to version 12.4 in the distribution installation.
Signed-off-by: Amr Mohamed <amr.mohamed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Upstream master and scarthgap have now diverged, so use scarthgap whilst
we prepare for release. At the time of writing there is no scarthgap
branch for meta-clang, so leave that on master.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Increase the number of TLB entries from 0x80 to 0x400 and disable the
checking of memory attributes. In our CI, this makes testimage run in
576s instead of 803s.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Multiconfig is only needed when recovery and the mass storage images
are built together. It is not needed when firmware-only build is used.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Enables host firewall and mpu setup for FVP. It also fixes secure-ram
configuration and disable access rights to secure ram from both normal world
for both mps3 and fvp.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
MM Communicate buffer is accessed by normal world but at the moment so it
should be located in DDR instead of secure-ram. This moves mm communicate
buffer to the DDR for trusted-service components.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
MM Communicate buffer is accessed by normal world but at the moment
it's allocated in the secure ram. This moves mm communicate buffer
to the DDR and also fixes the capsule buffer size since it cannot be
more than the bank size.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Remove the preferred version so that we track the latest release.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Some of the corstone u-boot patches are not properly formatted, causing
scripting issues. Regenerate them via:
devtool modify u-boot
git format-patch -N --no-signature devtool-base..HEAD
mv *.patch ~/meta-arm/meta-arm-bsp/recipes-bsp/u-boot/u-boot/corstone1000/
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update to the latest stable version (1.5.2), comprised of the following
commits:
5d86f27a8c0a opencsd: Update version info and README for 1.5.2
71c50dda716f build: win: Fix name for ocsd-perr utility in windows build
599551d3ea09 opencsd: docs: Update docs for test programs
769faaa6368a opencsd: docs: Update trc_pkt_lister man file
b957577e71bf tests: Fix typo in trc_pkt_lister help output
dca84a74a0d5 build: Fix clean of mem_acc_test
85fd025eed35 opencsd: stm: Fix build warning in 64 bit build of STM
2145b81b4b61 opencsd: etmv4: Fix build warning on decoder
169cc07d3625 docs: Fix formatting in README.md
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update to the latest stable version (2.10.3), comprised of the following
commits:
fc93d0edfc52 docs(changelog): changelog for lts-v2.10.3 release
4a10950a8538 docs(changelog): display all sections
bafc27c8d7cf chore: rename Poseidon to Neoverse V3
a6256d7a2638 feat(cpu): add support for Poseidon V CPU
ef393a3f9fa2 fix(cpu): correct variant name for default Poseidon CPU
81931a13a835 fix(cpus): workaround for Cortex-A715 erratum 2413290
baf14745f117 fix(cpus): workaround for Cortex-A720 erratum 2926083
635c83eb456a chore: update status of Cortex-X3 erratum 2615812
03636f2c3d60 fix(cpus): workaround for Cortex-A720 erratum 2940794
e86990d0911d fix(cpus): fix a defect in Cortex-A715 erratum 2561034
b59307ef8efd fix(cpus): workaround for Cortex-A715 erratum 2413290
44f36c48f280 docs(sdei): provide security guidelines when using SDEI
11cb0962f7ac docs(threat_model): mark power analysis threats out-of-scope
3e3ff298a614 fix(cpus): workaround for Cortex-A715 erratum 2344187
d466c5d4d27b fix(cpus): workaround for Cortex-X4 erratum 2701112
940ebbe2d1d0 fix(cpus): workaround for Cortex-A715 erratum 2331818
04c60d5ef31c fix(cpus): workaround for Cortex-A715 erratum 2420947
b7ed781eea74 fix(gic600): workaround for Part 1 of GIC600 erratum 2384374
58646309aedf chore: rearrange the fvp_cpu_errata.mk file
a234f540b727 fix(cpus): add erratum 2701951 to Cortex-X3's list
a24c8006ea39 refactor(errata-abi): workaround platforms non-arm interconnect
9fe65073d442 refactor(errata-abi): optimize errata ABI using errata framework
301698e15bc8 fix(cpus): workaround for Cortex-A715 erratum 2429384
5f8f745c7e99 fix(cpus): workaround for Cortex-X3 erratum 2372204
Signed-off-by: Jon Mason <jon.mason@arm.com>
The get_testimage_json_result_dir helper in OE core was deleted in
commit 01b1a6a5a4e7cede4d23a981b5144ae9c8306274 in preference for a
common utility. Change the reference within the Arm SystemReady ACS
log handler utlity.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This change upgrades the tftf version to v2.10 for the Corstone-1000.
Signed-off-by: Harsimran Singh Tungal <harsimransingh.tungal@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Since sbsa-acs requires edk2-firmware, it bases its SRC_URI on that.
The first entry of SRC_URI is what is used to determine the latest
version. So, specify an alternative URI to determine the correct
version.
Signed-off-by: Jon Mason <jon.mason@arm.com>
This change upgrades the trusted-firmware-m version to 2.10
for n1sdp.
Signed-off-by: Ali Can Ozaslan <ali.oezaslan@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
As of oe-core ba391d3, hashserv lookups will use authentication from the
.netrc file. However, Kas will write invalid netrc files with comments,
which causes bitbake to emit warnings.
This has been fixed in Kas in e700729 but until Kas 4.3.2 is released we
can ignore this warning specifically when checking the logs.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
When the URL and names of the toolchain tarballs changed, the
UPSTREAM_CHECKs were not modified with the proper values. This causes
the tooling to not show when new versions are available. Modify to get
it working again.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Initial checking providing support for RMM on QEMU's "virt" machine.
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
To reduce build and test times in CI, move the dev kernel outside the
standard matrix. This results in it still being built and tested for
the platform, but only with gcc/glibc (and not against clang and musl).
This greatly reduces the number of permutations that need to be
verified.
Signed-off-by: Jon Mason <jon.mason@arm.com>
gcc-arm-none-eabi v11.2 is no longer needed by tf-m. Remove this
version, as there is a newer one available.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update to the latest version of SCP. In this release, some of the
platforms were grouped into common family directories, which
necessitated adding a variable to specify which one.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update to the latest version. Also, remove the gcc12 workaround, as
that was added in edk2 commit 206168e83f090, which has been included
since the 202305 release.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update the TF-A recipe to the latest stable version (2.10.2).
NOTE: tf-a-tests did not have a corresponding stable release. So,
keeping back at 2.10.0.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Adds a recipe (ts-sp-block-storage) to build the Block Storage
secure partition to enable feature development for downstream users.
Signed-off-by: Emekcan Aras <Emekcan.Aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Remove scp-firmware sha for intermediate version
to update to v2.13 according to the Arm Reference solutions
Feb-2024 manifest.
Signed-off-by: Anusmita Dutta Mazumder <anusmita.duttamazumder@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
With clang 18, optee-os no longer compiles cleanly. It is now seeing:
| In file included from core/arch/arm/kernel/vfp.c:6:
| In file included from core/arch/arm/include/arm.h:131:
| core/arch/arm/include/arm64.h:436:1: error: expected readable system
register
| 436 | DEFINE_U32_REG_READWRITE_FUNCS(fpcr)
| | ^
| core/arch/arm/include/arm64.h:417:3: note: expanded from macro
'DEFINE_U32_REG_READWRITE_FUNCS'
| 417 | DEFINE_U32_REG_READ_FUNC(reg) \
| | ^
| core/arch/arm/include/arm64.h:411:3: note: expanded from macro
'DEFINE_U32_REG_READ_FUNC'
| 411 | DEFINE_REG_READ_FUNC_(reg, uint32_t, reg)
| | ^
| core/arch/arm/include/arm64.h:398:15: note: expanded from macro
'DEFINE_REG_READ_FUNC_'
| 398 | asm volatile("mrs %0, " #asmreg : "=r" (val64));
\
| | ^
| <inline asm>:1:10: note: instantiated into assembly here
| 1 | mrs x8, fpcr
| | ^
Issues are also seen on optee-examples and optee-test.
Forcing GCC for all optee recipes until this issue can be resolved.
Signed-off-by: Jon Mason <jon.mason@arm.com>
The workaround is no longer needed because with the
0043-firmware-psci-Fix-bind_smccc_features-psci-check.patch file the
u-boot PSCI driver is compliant with the PSCI specifications.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The u-boot PSCI driver was not compliant with the PSCI specifications so
this patch had to be added.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
GNU Arm compiler version greater and equal than *11.3.Rel1*
has a linker issue in syscall for TF-M 1.8.0. Let's bump to
TF-M 2.0 which contains the fix for the issue.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Ali Can Ozaslan <ali.oezaslan@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Corstone-1000 uses trusted-firmware-m as secure enclave software
component. Due to the changes in TF-M 2.0, psa services requires
a seperate client_id now. This commit adds smm-gateway-sp client id to
the FMP services since FMP structure accessed by u-boot via
smm-gateway-sp.
Signed-off-by: Ali Can Ozaslan <ali.oezaslan@arm.com>
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Enable the patch-status warning for meta-arm and meta-arm-bsp.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The user guide document for Corstone-1000 has been updated to reflect
the changes required following the multiconfig changes as well as now
running the fvp within the kas shell to ensure all environment variables
are picked up correctly.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The 2 Corstone-1000 kas files files are updated following the
multiconfig changes. The pinned commits have been commented out and
the default branch changed to master to allow the file to build valid
images.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Fix the build with DISTRO_FEATURES containing "usrmerge":
make: *** No rule to make target '/.../optee-os/4.1.0/recipe-sysroot/lib/optee_armtz/bc50d971-d4c9-42c4-82cb-343fb7f37896.stripped.elf', needed by '/.../optee-os/4.1.0/build/core/early_ta_bc50d971-d4c9-42c4-82cb-343fb7f37896.c'. Stop.
Fixes: 6a105f47b9 ("optee-ftpm: Install artifacts into nonarch_base_libdir")
Signed-off-by: Alexander Sverdlin <alexander.sverdlin@siemens.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
To ensure the psa and optee tests are included in the initramsfs based
rootfs included within the flash image so the tests can be run.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
To allow us to continue to ship Corstone-1000 releases that only include
the firmware with the built in Linux image we need a way to build it
outside of the multiconfig builds.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
By building the Corstone-1000 firmware under the firmware multiconfig we
can also build a minimal standard core image to be mounted in the fvp as
a mass storage device.
To do this we had to enable the MMC card interface in the Corstone-1000
kernel configuration.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
To enable building rescue or bootstrap images that can be included into
firmware a "firmware" multiconfig option is required to allow the
building with different options to any mass storage image they may also
be built.
As this multiconfig build will occur under a different TMPDIR, we also
provide a deployment image to allow easy copying of the firmware into
another deploy dir.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
The TF-M configuration step can fail if the doxygen executable is found.
This commit disables the doc generation until this is fixed in the
upstream repos.
Signed-off-by: Bence Balogh <bence.balogh@arm.com>
Bumped kernel version to v6.6 and rebased N1SDP kernel PCIe quirk patches on top of this new version.
Signed-off-by: Anusmita Dutta Mazumder <anusmita.duttamazumder@arm.com>
Since there are no platforms using this version, the related files can
be removed.
Signed-off-by: Anusmita Dutta Mazumder <anusmita.duttamazumder@arm.com>
As we intend to build 2 Corstone-1000 disk images, one for the firmware
in flash and an external mass storage image the existing
corstone1000-image.bb file has been renamed to
corstone1000-flash-firmware-image.bb to make it clear what it's for.
The wks file for specifing the image layout has also been renamed to
make its purpose clearer.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
We can now use the standard poky distro configured to be small by
switching distrobution and using the standard minimal image
from poky.
To do this we also remove and image configuration options from the
machine config and apply them in the kas files.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Moves a number of setting from the machine definition to the actual
recipes they apply too.
Added image configuration and dependancies to the flash image definition
file.
Reordered the settings in the machine definition to group them by
component that are related to.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Moved the U-Boot configuration items from the machine definition to the
Corstone-1000 specific U-Boot append file as it makes it easier to the
U-Boot configuration for a machine in one place and to make it more
consistant with other platforms.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
The capsule_cert and capsule_key file generated by u-boot for
corstone1000 do not get deployed correctly since writingh the output directly
to ${DEPLOY_DIR_IMAGE} causes the sstate mechanism to malfunction
especially in the CI builds. This patch fixes the issue and deploy the
generated files correctly.
Signed-off-by: Ali Can Ozaslan <ali.oezaslan@arm.com>
Signed-off-by: Emekcan Aras <Emekcan.Aras@arm.com>
Setting an API key means we get higher rate limits. Because keys are
private, the key must be set in the environment of the runner.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This changeset fix the tftf tests issue on n1sdp. Before this change, the tftf tests were getting stuck on n1sdp.
The following changes have been done:
1. There were some tftf tests based on multicore which involve powering up the other cores. These tests were creating
issues and the same thing has already been mentioned in the tests-to-skip.txt file for n1sdp platform in tftf source.
Those tests are skipped while executing tftf and patch has been created.
2. The TFTF_MODE variable added for tftf v2.10 recipe file, as did earlier for tftf v2.9. With the help of this, we can
enable debug or relase mode. The configuration based on this has been added for n1sdp in the corresponding bbappend file.
3. Add PREFERRED_VERSION_tf-a-tests for v2.10.
Signed-off-by: Harsimran Singh Tungal <harsimransingh.tungal@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add a Kas fragment to enable the CVE checker. Disable warnings by
default but show them for the layers in meta-arm, because we only care
about meta-arm issues in this CI.
Explicitly hide kernel warnings as the kernel typically has tens of open
CVEs, and if we're carrying a kernel explicitly then it's typically an
interim kernel between releases.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Extend jobs-to-kas so the first argument is still the GitLab job name,
but allow further arguments to specify extra Kas files to use in
addition.
Then add a variable EXTRA_KAS_FILES to the CI configuration that
defaults to the empty string and pass this to jobs-to-kas.
This lets specific pipeline runs add extra Kas files, for example to use
experimental branches or enable extra features without touching the CI
directly.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Linux 6.1 will be removed from oe-core master shortly, so whilst we
still have BSPs that use it (specifically, n1sdp) carry a 6.1 recipe in
meta-arm-bsp.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Since there are no platforms using this version, this reciepe can
be removed.
Signed-off-by: Vikas Katariya <vikas.katariya@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Since linux yocto kernel 6.5 is EOL and other layers like
meta-virtualization dropping support for it, it would be
sensible to downgrade the kernel to 6.1 which is a LTS.
This is a temporary change and later we would move to 6.6 when its
officially supported on N1SDP.
This revert the following commits:
* 1fe76c893c
* 21df60b921
Signed-off-by: Jon Mason <jon.mason@arm.com>
Updated to the latest version. Corstone1000 doesn't seem to boot. So,
pull back the old version to meta-arm-bsp for it to use temporarily.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Changes were needed due to the 2.0.0 version not being available for
download at pypi (though listed as the latest version there).
Signed-off-by: Jon Mason <jon.mason@arm.com>
Scripts processing data for the patch ages need correct information in
the relevant fields to determine the age. Create/correct this
information where missing/incorrect.
Signed-off-by: Jon Mason <jon.mason@arm.com>
The SHAs for both libmetal and open-amp are intermediate SHAs, which are
only a few patches behind the v2021.04.0 tags. Update to those tags and make
the necessary changes to get them working.
Signed-off-by: Jon Mason <jon.mason@arm.com>
oe-core master has upgraded from 6.5 to 6.6, but as we have BSPs that
still use 6.5 (corstone1000, n1sdp) we need to carry the recipe. This
should be a short-term measure as 6.5 was EOL in December 2023.
Also, drop the unused 6.4 linux-yocto which was no longer needed since
N1SDP moved to 6.5 with 21df60b.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The tftf tests were getting crashed on the MPS3 and two issues were found
during investigation. This change set provide fix for those issues:
1. The tftf tests were getting crashed on the MPS3 when compiled with
LOG_LEVEL=50. So, reducing the log level and aligning it with the TF-A
allows us to get rid of the crash.
2. Once the above crash got resolved, it has been found that tftf tests
were getting asserted while reading the value of the register
GICD_ITARGETSR. The reason for the crash is that the value of this register
is zero. As per the GIC documentation, this register is RAZ/WI for uniprocessor
implementation and corstone1000 is uniprocessor implementation for FPGA.
So, this change compiles the tftf tests in release mode which allows us
to compile out the assert definitions.
Signed-off-by: Harsimran Singh Tungal <harsimransingh.tungal@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add CI builds for Arm SystemReady Firmware within the fvp-base CI job and a new
Arm SystemReady IR ACS build job. Add the CI kas config for each of these
builds.
The ACS build can be controlled by the ACS_TEST GitLab variable to specify
whether or not to run the testimage. If this variable is not set, the
testimage step will not run. The job tag can be controlled by the ACS_TAG GitLab
variable.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add the parted-native dependency explicitly which is needed to
use wic commands.
Also explicitly inherit testimage. This means that the kas config
is no longer required to include it in IMAGE_CLASSES.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update tf-a and mbedtls to the latest versions. Also, migrate the
previous version to meta-arm-bsp for corstone1000.
NOTE: in v2.10, the fiptool makefile was changed to reference LDOPTS
instead of LDLIBS.
NOTE: commit 408cde8a59080ac2caa11c4d99474b2ef09f90df in tf-a modifies
the qemu_sbsa starting offset, and per the commit comment, it requires
the edk2 same change. This is why the edk-platforms SHA has been
changed. There are only 19 patches between the previous SHA and this
one (most of which are adding a single platform). So, it shouldn't be
too impactful to bump the SHA (instead of making it a patch to apply
on top of the existing SHA).
NOTE: tf-a-tests added LDFLAGS to the makefile, causing the need for it
to be removed in the recipe.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Now that we've released 4.3 and branched, we can switch master CI back
to master.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
We enable PAC/BTI out of the box, but all of the pieces (such as gcc and
glibc) need to support it for the final binary to be protected.
Add a minimal test recipe to verify that the "Hello, World" binary is
using PAC/BTI, and add it to oe-selftest.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Readthedocs server requires an absolute path from project's root
directory to find correct requirements.txt and conf.py.
Signed-off-by: Emekcan Aras <Emekcan.Aras@arm.com>
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Ross Burton <ross.burton@arm.com>
Upgrade the Corstone-1000 YMLs as follows:
- Set the layers SHAs
- Align with Kas v4
- Update the layers list
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Ross Burton <ross.burton@arm.com>
Add a note for Capsule update negative test scenario and fixes instructions
regarding distro-boot in Corstone-1000 user guide.
Signed-off-by: Emekcan Aras <Emekcan.Aras@arm.com>
Signed-off-by: Delane Brandy <delane.brandy@arm.com>
Signed-off-by: Ross Burton <ross.burton@arm.com>
Adds creating an EFI System Partition for Corstone-1000.
Signed-off-by: Emekcan Aras <Emekcan.Aras@arm.com>
Signed-off-by: Ross Burton <ross.burton@arm.com>
align the release note with the upcoming CORSTONE1000-2023.11 release
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Ross Burton <ross.burton@arm.com>
Use Ethernet over VirtIO on FVP due to lan91c111 Ethernet driver support dropped from U-Boot.
This patch enables virtio-net device in u-boot to pass ACS tests related to
NIC and PXE. The current ethernet device still works in linux kernel and
corstone1000-mps3 Ethernet device is supported by u-boot, so no change is
required regarding existing Ethernet device.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Ross Burton <ross.burton@arm.com>
Adds virtio-net configuration to use virtio-net in corstone1000-fvp.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Ross Burton <ross.burton@arm.com>
Update the user guide to include Linux distro installation on fvp
Signed-off-by: Delane Brandy <delane.brandy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This fixes a race that is observed rarely in the FVP. It occurs in FVP
when tfm sends the notication ack in openamp, and then reset the access
request which resets the mhu registers before received by the host
processor. It implements the fix both in SE and the host processor openamp
wrapper. This solution enables polling on the status register of mhu until
the notificaiton is read by the host processor. (Inspired by
signal_and_wait_for_signal function in mhu_wrapper_v2_x.c in trusted-firmware-m
https://git.trustedfirmware.org/TF-M/trusted-firmware-m.git/tree/platform/ext/target/arm/rss/common/native_drivers/mhu_wrapper_v2_x.c#n61)
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Upgrade the Arm binary toolchains to the latest version. Of note, the
untarred directory has camelcased the "R" in Rel (which was "rel" in the
previous versions).
Signed-off-by: Jon Mason <jon.mason@arm.com>
The fvp-base machine uses a v8.4 tune and the FVP itself is configured
to be v8.4, so also tell TF-A to use v8.4.
This is normally done in the TF-A board configuration, but the fvp board
is configurable.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
We already tell the FVP to be v8.4 cores, so tell the compiler to
tune for that instruction set too.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This CVE is specific to NXP i.MX boards which are documented as being
shipped unsecure, as they're meant for development.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Changes ns-interrupt-action for corstone1000. This will enable
preemption in the SPs which is the default way to handle interrupts in
trusted-firmware and optee documentation.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
When booting, the tee driver from kernel side
invokes a yielding call to OP-TEE, which gets
stuck because OP-TEE never sends Done response:
OPTEE_FFA_YIELDING_CALL_RETURN_DONE
This issue was previously resolved by introducing
an inappropriate patch to the kernel with 1 ms delay
in ffa_msg_send_direct_req.
Further investigation proved that OP-TEE doesn't
get enough processing time and is constantly interrupted
by the kernel requests. To remove this patch, TF-A logging
level is lowered to default (40 in debug builds and
20 in release builds), which eliminates the time consumed
previously by TF-A VERBOSE logs (giving OP-TEE more
processing time).
Signed-off-by: Mariam Elshakfy <mariam.elshakfy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Corstone1000 capsule update implementation does not support use of
scatter_gather_list. This workaround passes 1 as scatter_gather_list value
to pass the NULL checks for scatter_gather_list while
CAPSULE_FLAGS_PERSIST_ACROSS_RESET flag is set (which is introduced lately to
align with UEFI specs). Since these flag checks are not implemented in u-boot
properly and corstone1000 does not support scatter_gather_list during capsule
update, this patch will skip the check only for on-disk capsule update.
This will be fixed with new capsule update design.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add an optional env argument to the run_fvp() function, and check that
DISPLAY is preserved.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The __main__ body used the return value of runfvp() as the exit code,
but this was never set.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Don't pass "" as the cwd as that fails, use None so the cwd doesn't get
changed.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
add debug.yml to the build command so the debug-tweaks image feature
is enabled
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This is no longer needed as Nanbield uses ttyrun to avoid re-spawning
gettys.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
corstone1000's uboot uses efitools-native from meta-efi-secure-boot, so
add the layer dependency to make this clear.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Use IMAGE_CLASSES rather than a direct INHERIT for fvpboot.
This is Yocto best practice as it is used to enable
functionality across all image recipes.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add systemready firmware, systemready ACS, systemready distros, and
fvp-base kas configurations. Update the README file with
instructions on how to build and run using them.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
1. Configure FVP base to allow Arm SystemReady IR ACS console access.
2. Configure FVP base Arm SystemReady IR ACS firmware build.
3. Add the machine-specific report.txt for FVP base.
4. Patch the check-sr-results.yaml and format-sr-results.yaml files
to handle the known differences between FVP base and the
expected ACS functionality.
5. Add a README with instructions of how to use the meta-arm-systemready
layer with fvp-base.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add the meta-arm-systemready layer. This provides the infrastructure to
load and run the Arm SystemReady IR ACS v2.0.0 prebuilt image and
analyze the results.
The recipes included are as follows:
1. arm-systemready-firmware: Enables the ARM_SYSTEMREADY_FIRMWARE machine
conf variable to be used to specify which firmware packages to deploy.
2. arm-systemready-ir-acs: Runs the Arm SystemReady IR ACS tests from
the pre-built images and checks the results adhere to the specification.
3. arm-systemready-linux-distros-[debian|opensuse]: Install the distro of
choice from CD/DVD image to target disk image.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Diego Sueiro <diego.sueiro@arm.com>
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Signed-off-by: Qi Feng <qi.feng@arm.com>
Signed-off-by: Robbie Cao <robbie.cao@arm.com>
Signed-off-by: Tomás González <tomasagustin.gonzalezorlando@arm.com>
Signed-off-by: Vineeth Raveendran <vineeth.raveendran@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
For fvp-base, update the DEFAULT_TEST_SUITES to include
fvp_boot and fvp_devices. This is only the default behaviour;
individual recipes can override this using the TEST_SUITES
variable.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Comments were made after previous optee changes were committed.
Addressing those comments here.
Suggested-by: Denys Dmytriyenko <denis@denix.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update capsule document procedure and ACS image in user guide.
Signed-off-by: Ali Can Ozaslan <ali.oezaslan@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update the recipes and patches for OP-TEE 4.0.0. Migrate the 3.22.0
recipe to meta-arm-bsp for corstone1000 and n1sdp.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Updating to the latest version of hafnium. Also, dropping tc patches,
as they are either experimental or a similar feature has been added.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Updates the status of the patches on the trusted-firmware-m for
corstone1000.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
U-boot mkefitool creates capsule image without packed and byte-aligned structs.
This patch aligns the capsule-update structures and avoids crashes in case of
unaligned pointer access.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Platform-specific capsule-update feature in u-boot does not check the
capsule-update flags properly (as stated in UEFI specs). This patch fixes the
capsule flags checks in u-boot for corstone1000.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Enables on-disk capsule update feature for corstone1000.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Bump kernel version to v6.5 and rebased N1SDP kernel PCIe quirk patches top of this new version.
Signed-off-by: Xueliang Zhong <xueliang.zhong@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Some FVPs are available as x86-64 and aarch64 binaries, so build target
(qemuarm64) and nativesdk (x86-64) packages for these to verify the
checksums are correct.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Now that there's a FVP for aarch64, we don't need to pin the CI pipeline
to x86-64.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Now that the FVP supports both aarch64 and x86-64, this inherit doesn't
need to be conditional.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Upgrade to the 11.22.35 release of the FVP.
Also add the aarch64 binaries as these are now available.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Configure grub as the EFI provider and remove the U-boot boot
args.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Configure FVP base to use vexpress_fvp_defconfig as the U-boot machine.
Configure U-boot:
1. Drop the patch to pick the DRAM size from the devicetree since
the FVP now specifies a devicetree.
2. Enable sysreset to reset by PSCI and patch the vexpress U-boot
machine to leave the reset to PSCI in this case.
3. Enable Virtio RNG and patch the U-boot Virtio RNG driver to
workaround an issue with the FVP that results in RNG calls
hanging.
4. Enable the Arm64 CRC-32 instruction by default and remove the now
redundant config setting.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Set the Trusted Firmware devicetree to fvp-base-gicv3-psci-1t. Patch the
devicetree to include: the stdout path for console access, a virtio net
node and a virtio rng node. This is necessary in the case that the
Trusted Firmware devicetree is passed to Linux from U-boot (rather than
sideloading).
Also rename the include file to change the suffix from "fvp" to "fvp-base".
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Merge the common FVP configuration in fvp-common.inc into
fvp-base.conf since that is the only place it is inherited.
Drop setting MACHINE_FEATURES to "optee" because there is no
optee machine feature.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Have separate machine include files rather than multiple
machine-specific settings in the same file.
Signed-off-by: Debbie Martin <Debbie.Martin@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
meta-clang and meta-virtualisation don't yet have nanbield branches, so
we need to use master for those at the moment.
Signed-off-by: Ross Burton <ross.burton@arm.com>
To allow running the TF-A TFTF tests we need to ensure the images for
N1SDP and Corstone-1000 MPS3 boards build
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
To allow running of the TF-A tests we need to be able to build the TF-A
test recipe for the N1SDP machine.
Signed-off-by: Drew Reed <Drew.Reed@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This change enables N1SDP cache to improve performance
by removing this patch:
HACK-disable-instruction-cache-and-data-cache.patch
Signed-off-by: Mariam Elshakfy <mariam.elshakfy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Since the original location of OP-TEE in DDR3 observes
a HW issue when cache is enabled, this change moves OP-TEE
to run from DDR4. Patches are added to TF-A to reflect that
change and the used region is also reserved in UEFI (EDK2)
to protect against allocations by UEFI applications.
OP-TEE size is modified for consistency across all patches
to be 32 MB (0x02000000) instead of (0x02008000).
Signed-off-by: Mariam Elshakfy <mariam.elshakfy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
the configuration options corresponding to external system are removed
from the kernel and the defconfig is generated with with savedefconfig
bitbake task
Signed-off-by: Harsimran Singh Tungal <harsimransingh.tungal@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Remove the External system patches in uboot as they are not upstreamable.
Signed-off-by: Harsimran Singh Tungal <harsimransingh.tungal@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This support is for Cassini distro using Corstone-1000 platform.
When running parsec test, it reports an error
`PSA_ERROR_DATA_INVALID (-153)`.
This is related to `ITS_MAX_ASSET_SIZE` configuration which is been
set to 512 on the secure enclave (TF-M), which defines the max asset
size and it overflows when running the parsec tests.
The key is generated, but when it is asked to store via `psa_its_set`
it returns `PSA_ERROR_INVALID_ARGUMENT (-135)`, which then propagates
to `PSA_ERROR_DATA_INVALID (-153)`
Increasing the `ITS_MAX_ASSET_SIZE` to 2048 solves this issue.
Signed-off-by: Vikas Katariya <vikas.katariya@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The terribly named qemutiny test case tries to login to the target over
the serial console. It's designed for poky-tiny, so add it to the tests
we run in poky-tiny builds.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Just execute all tests with the meta-arm tag, instead of hardcoding the
list of tests.
Also run two tests in parallel as there's no reason to limit it to one.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Tag all of the tests in meta-arm so that they can be selectively ran
without needing to explicitly list them.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Patch was accepted upstream and has been pulled back the 6.5 and 6.1
kernels. So, it is no longer needed here.
Signed-off-by: Jon Mason <jon.mason@arm.com>
The v6.4 kernel is needed for some platforms in meta-arm-bsp.
Temporarily add it here to give those machines enough time to
update to the latest version. Also, add the patch to the
defconfig.
Signed-off-by: Jon Mason <jon.mason@arm.com>
When ccache is enabled trusted-firmware-a recipe fails with this
error message:
make: *** No rule to make target 'aarch64-poky-linux-gcc'. Stop.
ccache prefix CC variable with 'ccache' word before compiler. Because
there are no quotes assigned to CC, only 'ccache' is assigned. The
compiler becomes a make target, producing the build error.
Add single quotes to LD is a good measure to prevent this kind of error.
Signed-off-by: Javier Tia <javier.tia@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Multiple machines in meta-arm-bsp have need of the 2023.07.02 version of
u-boot. Temporarily add it here to give those machines enough time to
update to the latest version.
NOTE: MTD changes in u-boot require changes to the qemuarm config.
Specifically, not disabling it.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Enables authenticated capsule update and makes necessary changes to
align with new capsule generation tool (mkeficapsule in u-boot).
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Adds signature to device-tree overlay and enables authenticated capsule
update in u-boot for corstone1000.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Adds meta-secure-core to enable capsule update feature.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Adds meta-secure-core to corstone1000.yml to enable signed capsule
update feature.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Meta-secure-core is used to create signed capsule (firmware update
images). This adds meta-secure-core.yml file and since it depends on meta-perl
from oe-core, it adds that layer to meta-openembedded.yml
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The Kas binary is identical, but the container has been rebuilt using
Debian 12 (Bookworm).
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The Kas container in version 4 onwards is based on Debian 12, which
forbids pip from installing files into /usr or ~/.local/.
We want to install the arbitrary dependencies for the documentation
build, so these should be installed in a venv.
The kas container doesn't currently install python3-venv, so we have to
install that manually (patch sent upstream).
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Upgrade nanopb, clean up how it is build, and hopefully fix the build
races. This patch isn't quite ready to be upstreamed but discussion
with the TS maintainer is ongoing.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Now that the apply_local_src_patches class supports being used with
multiple directories, use that instead of reimplementing the logic.
Also remove redundant patchdir assignments as these patches are against
the trusted-services repository, which is ${S}. I suspect these are
exposing a subtle bug in the core patching logic which meant the local
patches were not applying correctly.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This test needs the TF-A sources available to build. When the test is
needed, this commit can be reverted to bring it back.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This recipe won't pass configure without the jsonschema and jinja2
Python modules.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
As with ts-service-test, manually move the binary to $bindir.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The configure log warns that it couldn't find pkgconfig, so add this so
it can.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Pass through the choice of CMake Generator when starting sub-cmakes for
the external components, so that they use Ninja instead of Make.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Pull out the patch application logic so the postfunc by default scans
for patches in LOCAL_SRC_PATCHES_INPUT_DIR and applies them to
LOCAL_SRC_PATCHES_DEST_DIR as before.
This allows recipes to inherit the class and directly call
apply_local_src_patches as needed to process patches in multiple
directories.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The fvp-baser-aemv8r64 machine will not be actively maintained.
Signed-off-by: Divin Raj <divin.raj@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Remove U-Boot specific DT nodes before passing the DT to Linux
This is needed to pass SystemReady IR 2.0 dt-schema tests
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The optee recipe installs the tee image using `${nonarch_base_libdir}`
If usrmerge is enabled this is `/usr/lib`, otherwise it is `/lib`
Several platforms (corstone1000, n1sdp, tc) look for tee-pager_v2.bin in
the hard-coded `/lib/firmware`, hence if usrmerge is enabled it won't be
found.
Fix these platforms by using `${nonarch_base_libdir}` instead of `/lib`
as per the qemu platform code in the generic recipe.
Signed-off-by: Adam Johnston <adam.johnston@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Bump kernel version to v6.4 and rebase the patches on
top of this new version.
Signed-off-by: Xueliang Zhong <xueliang.zhong@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
TFM updated to 1.8.1. Note, TF-Mv1.8.1 tags point to the same SHA as
the TF-Mv1.8.0 tag for tf-m-tests and tf-m-extras.
Signed-off-by: Jon Mason <jon.mason@arm.com>
tee and teeclnt are there to avoid running client applications (CAs) and
tee-supplicant as root.
- The teeclnt group stands for "TEE client" and is for CAs (CAs need
access to /dev/tee[0-9]* but not /dev/teepriv[0-9]*).
- tee is just for tee-supplicant to open its device /dev/teepriv[0-9]*.
No other process is supposed to open that one.
Signed-off-by: Javier Tia <javier.tia@linaro.org>
Reviewed-by: Jerome Forissier <jerome.forissier@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Rationalise the port forwarding to be the same as the runqemu defaults,
so change the SSH port forward to be 2222=22 instead of 8022=22.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
QEMU_USE_SLIRP is no longer needed[1] as adding slirp to
TEST_RUNQEMUPARAMS is sufficient, so remove that.
Setting TEST_SERVER_IP also isn't needed as there's a default value now,
and we disable the package management tests that would use the server
IP. When they work the correct IP can be set.
[1] As of oe-core f4e8650
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update to the latest versions of edk2 and edk2-platforms. This
necessitates updating the patches in sbsa-acs to apply cleanly to the
latest version of edk2.
Signed-off-by: Jon Mason <jon.mason@arm.com>
It's expected to exist multiple /dev/teepriv[0-9]* devices, and the
tee-supplicant service depends on them, which should be activated only
when the device is detected by the kernel using a udev rule.
Improve commit f02d065dce, where it's only considering a path creation
and not a device detection by the kernel.
Signed-off-by: Javier Tia <javier.tia@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The update-repos script currently exits immediately if one of the
underlying Git commands fails (e.g. because of a network issue). If the
repo already exists, then catch this error inside the loop and
carrying on attempting to update other repos, as the network error may
be upstream.
KAS_REPO_REF_DIR is ultimately an optimization and subsequent build
stages should be able to continue if one of the updates fail. Therefore,
ensure the script returns a special error code if at least of the Git
commands fail, and use this to set the allow_failure property of the
job.
If a repo does not exist, fail immediately as before.
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
To make the pipeline slightly more resilient to external networking
issues, allow a local container registry mirror to be specified in the
GitLab settings. If not specified, the upstream container registry is
used automatically.
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Since optee-os for N1SDP has been updated to 3.22,
this patch updates optee-os-tadevkit and optee-test
to match the same version.
Signed-off-by: Mariam Elshakfy <mariam.elshakfy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Bump kernel version to v6.4 and rebased N1SDP kernel PCIe quirk patches
top of this new version.
Signed-off-by: Xueliang Zhong <xueliang.zhong@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The default XSERVER only pulls in the framebuffer driver, which is
pretty broken with modern kernels and the modesetting driver is a lot
more functional.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The sbsa-ref machine can't use KVM because it's an entire emulated
machine, not a virtual machine.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The consoles to use isn't specific to the qemu machine, and without a
value results in no serial consoles when running on real hardware under
sysvinit.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Patches (and recipe support) were added for qemuarm64-secureboot
support, but that is not present in meta-arm-bsp. Remove it.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Without /dev/teepriv[0-9]*, tee-supplicant.service will fail. Prevent
a failure with a condition to check if /dev/teepriv[0-9]* path exists.
Signed-off-by: Javier Tia <javier.tia@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
ARM Reference Solutions' N1SDP 2023.06.22 manifest
uses edk2-firmware version 202305. This patch
aligns with the manifest.
The RemoteDdrSize cast patch is now upstreamed,
hence removed from the patches list
Signed-off-by: Mariam Elshakfy <mariam.elshakfy@arm.com>
Enables the support of a second mmc card, which enables distro installation.
Signed-off-by: Delane Brandy <delane.brandy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Build a custom N1SDP image with only optee-xtests as part of the image. A fresh custom image build is necessary to include the relevant test suite for running the xtests.
Signed-off-by: Anusmita Dutta Mazumder <anusmita.duttamazumder@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add the config needed to run the embedded tests with pseudo trusted application. Without this config, the optee-xtests with pseudo TA get skipped with “skip test, pseudo TA not found” message.
Signed-off-by: Anusmita Dutta Mazumder <anusmita.duttamazumder@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Clean up various patches by
* Create email headers for those not present using the s-o-b as author and date
applied to the tree as the patch date
* regenerating the patch name via git rebase and format-patch
* replacing patch with backported version
* moving patch location to be more accurate
Signed-off-by: Jon Mason <jon.mason@arm.com>
Rebase the corstone500 u-boot patches to 2023.07.02. Some defined
variables changed names, and had to be updated.
Signed-off-by: Jon Mason <jon.mason@arm.com>
As of oe-core a8e7b0f, PV doesn't need to contain SRCPV explicitly as
base.bbclass will append it to PKGV if there is a + in the version. So,
remove the redundant assignments.
Also change the boot-wrapper-aarch64 PV to 0+git as it's possible that
some point they'll tag a release and we want to be able to upgrade to it.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
u-boot patches rebase cleanly to u-boot v2023.01. Update to that and
get rid of the legacy version of u-boot that existed only for this
machine.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Updates optee-os and aligns with changes in v3.22 for Corstone-1000.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
With 6.1.46, the gimple patch has been backported as part of the
release, and is no longer needed as a patch applied here.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update hafnium to v2.8, which allows updating tc1 to that version and
remove the intermediate SHA. It also allows for the removal of some
backported patches.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Updates trusted-firmware-a and aligns with changes in v2.9 for Corstone-1000.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update TF-A version for N1SDP to align with
N1SDP 2023.06.22 manifest
Signed-off-by: Mariam Elshakfy <mariam.elshakfy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Use the patch that Mikko Rapeli is trying to upstream to work around the
Nuvoton defconfig issue instead of reverting the patch that added the
platform to the kernel.
Signed-off-by: Jon Mason <jon.mason@arm.com>
The scp-firmware version in the Arm Reference Solutions N1SDP-2023.06.22 Release has now updated to a beta version beyond v2.12.
Add the SHA override for N1SDP to align to scp-firmware version used in the release.
Signed-off-by: Anusmita Dutta Mazumder <anusmita.duttamazumder@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add new Corstone-1000 firmware GUID and remove previous u-boot GUID to be updated into ESRT table.
SR-IR 2.0 requires the capsule GUID to be unique.
Signed-off-by: Anusmita Dutta Mazumder <anusmita.duttamazumder@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add new Corstone-1000 firmware GUID and remove previous u-boot GUID to be updated into ESRT table.
SR-IR 2.0 requires the capsule GUID to be unique.
Signed-off-by: Anusmita Dutta Mazumder <anusmita.duttamazumder@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update the trusted-firmware-a recipes to 2.9.0
Moving legacy recipes (2.8) for tc1 and corestone1000 to meta-arm-bsp
Signed-off-by: Jon Mason <jon.mason@arm.com>
use filesize environment variable to read the size of the unzipped initramfs bundle
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
If a user does eg "kas shell kas/corstone1000-base.yml" and then calls
runfvp, the spawned xterms don't have a valid DISPLAY set.
Add DISPLAY to the preserved environment variables and DISPLAY will be
passed into the shell, and the xterms will start correctly.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add kernel configuration necessary to build an image with preempt-rt
support for generic-arm64.
And tweak kernel configuration for preempt-rt kernel.
Signed-off-by: Robbie Cao <robbie.cao@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
202211 is only used in meta-arm-bsp, and all other users should be using
the latest version. Move it there until n1sdp can be updated to the
latest.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add back the 11.2 release of GCC, because TF-M has code generation
problems with 11.3 onwards.
This recipe has the major version embedded in the recipe name so that it
has to be specifically asked for in a per-recipe basis.
Signed-off-by: Ross Burton <ross.burton@arm.com>
So that it's obvious which version of GCC is being used, install the
files into a versioned directory under $libexecdir.
Signed-off-by: Ross Burton <ross.burton@arm.com>
This was added to build an Android Common Kernel with the Android
compiler, but we don't need to build that anymore.
Signed-off-by: Ross Burton <ross.burton@arm.com>
The unzipped kernel size increased due to recent changes in oe master.
Since corstone1000 sets a fixed kernel size, this should be increased to
boot the platform correctly.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
The physical memory which is used to run OP-TEE on the N1SDP is known
to the secure world via TOS_FW_CONFIG, but it may not be known to the
normal world.
As a precaution, explicitly reserve this memory via NT_FW_CONFIG to
prevent the normal world from using it. This is not required on most
platforms as the Trusted OS is run from secure RAM.
Signed-off-by: Adam Johnston <adam.johnston@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The upstream u-boot recipe did a minor version update, which is causing
a "preferred version not available" warning. Add a '%' to resolve this
issue.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Temporary workaround for a number binaries in the toolchains that are
using 32bit timer API. This must be done in the CI yml file instead of
the recipe because of all the libraries in the toolchain have the issue.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Pre-populate the standard values of get-binary-toolchains so that it
will work without variables in a standard kas container. This will aid
in ease-of-use for development and testing of binary toolchains.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Port Corstone-1000 patches to U-Boot v2023.07
Include the latest FF-A patchset sent to the mailing list (v15).
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
add SMCCC_ARCH_FEATURES discovery
Signed-off-by: Abdellatif El Khlifi <abdellatif.elkhlifi@arm.com>
Upstream-Status: Inappropriate [A U-Boot patch will be released to fix an issue in the PSCI driver]
Signed-off-by: Jon Mason <jon.mason@arm.com>
Move the 3 patches only needed by the 6.1 kernel into a unique bbappend
The defconfig changes cannot be moved into a config fragment because
they only exist in the defconfig file (because the patches that
integrated their functionality into the kernel were not merged).
Signed-off-by: Jon Mason <jon.mason@arm.com>
The optee-os.inc was including patches which really should've been in
the versioned .bb file. Move those references there and update the
FILESEXTRAPATHS to be more intelligent. While there, rebase the files
via devtool and update the file names as necessary.
Also, remove unreferenced patches.
Signed-off-by: Jon Mason <jon.mason@arm.com>
1- Replace FVP_BASE_R_ARM_EULA_ACCEPT with ARM_FVP_EULA_ACCEPT
in fvp-baser-aemv8r64.md
2- Add instructions to corstone1000/user-guide.rst to set
ARM_FVP_EULA_ACCEPT to "1".
Signed-off-by: Ziad Elhanafy <ziad.elhanafy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
1- Add fvp-eula.yml file which introduces ARM_FVP_EULA_ACCEPT
environment variable
2- Remove any license related settings from fvp-baser-aemv8r64-bsp.yml,
corstone500.yml and corstone1000-fvp.yml
Signed-off-by: Ziad Elhanafy <ziad.elhanafy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Set the FVP EULA environment variable details message
to request the user to accept the EULA in case it has
not been accepted.
Signed-off-by: Ziad Elhanafy <Ziad.Elhanafy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The following error is seen:
ERROR: Nothing PROVIDES 'llvm-native' (but virtual:native:/builds/jonmason00/meta-arm/work/build/../poky/meta/recipes-graphics/mesa/mesa_23.1.3.bb DEPENDS on or otherwise requires it). Close matches:
ovmf-native
rust-llvm-native
vim-native
Work around this by using the llvm-native from meta-clang
Signed-off-by: Jon Mason <jon.mason@arm.com>
Upstream change 26d97acc71379ab6702fa54a23b6542a3f51779c changed the
IMAGE_LINK_NAME to have IMAGE_NAME_SUFFIX and breaks the automatic
finding of the image by name. Work around here until upstream fix can
be added.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Remove IMAGE_NAME_SUFFIX from image path in
wic_nopt.bblass
Signed-off-by: Ziad Elhanafy <ziad.elhanafy@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The fvp_devices test suite can be used to verify the following
functionality at runtime, common to most FVPs:
* CPU hotplug
* virtio-net device presence and functionality
* virtio-rng device presence and functionality
* PL031 RTC device presence and functionality
* SP805 watchdog device presence
The list of devices to be tested can be configured by a BSP using the
variable TEST_FVP_DEVICES.
Add this test suite for fvp-base and fvp-baser-aemv8r64.
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The linuxboot test case is already FVP-specific due to the use of the
OEFVPTarget pexpect interface. Clarify this by renaming to fvp_boot.
So that fvp_boot can be used alongside other OEQA test cases (e.g. those
in OE-core):
* Call self.target.transition("off") at the start of the test to
ensure the model starts from reset
* Call self.target.transition("linux") to reuse the "wait for boot"
logic in OEFVPTarget.
Additionally, minimally validate the firmware boot by checking for
common error patterns in all console logs. Expose the runfvp config in
OEFVPTarget to support this.
Align the list of test cases executed on both fvp-base and
fvp-baser-aemv8r64 by using TEST_CASES:append = " fvp_boot" for both.
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
To better support firmware testing alongside Linux runtime testing,
introduce model state support to OEFVPTarget. The following states are
supported using self.target.transition(state):
* off
* on
* linux
Instead of assuming a specific state in OEFVPTarget.start,
responsibility is delegated to test cases to lazily put the model in
the required state. But to support OE-core test cases, OEFVPTarget.run
automatically puts the model in the "linux" state for running the
command. Firmware and Linux tests can subsequently run alongside each
other without introducing complex test dependencies.
The concept is inspired by Labgrid strategies [1], albeit simplified.
Tweak log file handling so that output is collected across (possibly)
multiple model processes.
[1] https://labgrid.readthedocs.io/en/latest/overview.html#strategies
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The differences between OEFVPTarget, OEFVPSSHTarget and
OEFVPSerialTarget are not obvious and there is a lot of duplication.
Merge all the logic into one OEFVPTarget (again). This has the following
features:
* Run SSH commands
* Run serial console assertions
* Lazily await a Linux login prompt while running test cases (only when
self.target.run is called).
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Using runfvp without explicitly specifying the fvpconf path currently
fails due to a missing fvp.conffile include.
Signed-off-by: Peter Hoyes <Peter.Hoyes@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update to the latest version of OP-TEE, and move the legacy versions
that are still in use by meta-arm-bsp to that directory
Signed-off-by: Jon Mason <jon.mason@arm.com>
By default, optee-test is using an ancient version of openssl (1.2.0o)
in binary form, located in the optee-test build tree. musl is already
working around this. So, use those defaults for everyone.
Signed-off-by: Jon Mason <jon.mason@arm.com>
scp-firmware passes -I/core/include to the compiler which doesn't exist,
and sometimes gcc emits a fatal error. It's unclear why this doesn't
happen for everyone, but this workaround appears to be the correct
solution.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
SCP_COMPILER isn't used anymore, so remove it.
Explicitly set SCP_TOOLCHAIN=GNU for clarity.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Now that this defaults to MACHINE, explicit SCP_PLATFORM assignments can
be removed.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Many platform names are the same as the machine name, so this saves
some typing.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
It is more difficult to update optee recipes to bleeding edge
3.21 when patches are added to SRC_URI via :append and thus they
need to be explicitly removed with :remove and name of the patch file.
For our boards we know 3.21 will work without patches but we still want
to keep using meta-arm side base recipe and just update the SRC_URI
to remove patches and update SRCREV and PV.
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This tool makes it easy to lock a build to a known good configuration,
for example by locking the SHAs to the last good build of master.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
In the update-repos job, generate a Kas lock file to pin the SHAs of the
repositories being build during the build. This ensures that commits to
the repositories _during_ the build doesn't cause differing builds. All
of the kas calls use this lock file to ensure that their builds are
identical.
This lockfile is also added to the artifacts, so that it can be reused
afterwards to replicate the build: either as a known good base or to
replicate failures.
This lock file is only generated if it doesn't exist, so that
development branches can temporarily add a lockfile.yml if for example
master is too unstable to develop on.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Kas 3.3 uses branch and commit keys instead of a general-purpose
refspec. Change the base to use branch:mickledore and remove the
explicit use of master for meta-virtualization now that a mickledore
branch exists there too now.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Kas 3.3 has the following new features:
- kas: Introduce commit and branch as alternative to refspec key
- kas: Warn if a repo uses legacy refspec
- kas: add support for lock files via dump plugin
- kas: track root repo dir config files of menu plugin
- kas: add support for --log-level argument
- kas: add GIT_CREDENTIAL_USEHTTPPATH environment variable
- kas: improve error reporting
- kas: drop support for Python 3.5
- kas-container: fix invocations with --isar for some layers
- kas-container: Purge tmp* on clean
- kas-container: enable colored logging
4.0 is basically the same but the Kas container uses Debian 12, which is
not yet supported as a tested platform. Until Debian 12 is tested we
should continue to use 3.3.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This will still emit the diagnostic but it wont break the build,
additionally pass it to CXXFLAGS since thats the right subset to
disable it for.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Same as for OP-TEE client, this eases debugging and is required to use
the OP-TEE symbolize.py script.
Signed-off-by: Rouven Czerwinski <r.czerwinski@pengutronix.de>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The TA builds always create ELF files, add them to the deploy dir to
ease debugging via the OP-TEE symbolize.py script.
Signed-off-by: Rouven Czerwinski <r.czerwinski@pengutronix.de>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The optee-os compilation provides not only the core files, TAs are also
usually build. Create a separate package which contains them, so they
can be installed.
Signed-off-by: Rouven Czerwinski <r.czerwinski@pengutronix.de>
Signed-off-by: Jon Mason <jon.mason@arm.com>
TF-A version 2.8.6 hangs on boot for corstone1000 non-deterministically.
This sets TF-A version to v2.8.0 which is a tested working version for
corstone1000.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Adds missing update service definitions for using stateless platform
services and initializes the capsule udpate provider in se-proxy-sp
for corstone1000.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Adds missing compilation option to fix psa_raw_key_aggrement test for
corstone1000.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
From: Gowtham Suresh Kumar <gowtham.sureshkumar@arm.com>
These 2 patches causes the secure world to enter into an infinite loop
when the PSA arch tests are triggered. This is a temporary fix and the
issue needs to be investigated before the patches can be enabled.
Signed-off-by: Gowtham Suresh Kumar <gowtham.sureshkumar@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This patch is required to handle one of the corner cases of the
GetNextVariableName EFI service as specified in the UEFI spec.
Signed-off-by: Gowtham Suresh Kumar <gowtham.sureshkumar@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
From: Gowtham Suresh Kumar <gowtham.sureshkumar@arm.com>
This patch adds the required configs to the corstone1000 u-boot
defconfig to enable the EFI services. This is done to fix the SCT
failure reported by the SetTime_Conf and SetTime_Func.
Signed-off-by: Gowtham Suresh Kumar <gowtham.sureshkumar@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
From: Gowtham Suresh Kumar <gowtham.sureshkumar@arm.com>
The patch fixes the ACS InstallMultipleProtocolInterfaces_Conf failures
in corstone1000 platform by dropping a workaround u-boot patch. The NVMXIP
initialization had some issues during u-boot boot stage which led to the
workaround patch.
Upstream-Status: Pending [Not submitted to upstream yet]
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Gowtham Suresh Kumar <gowtham.sureshkumar@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
It's useful to have known good logs when debugging problems, so always
preserve the logs.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The previous commit fixed the build of GN with GCC 13, but broke it for
any other version.
Remove the patch and simply disable the fatal warning that causes the
breakage. Interestingly, this warning is already disabled for Windows
builds.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
EXTERNAL_TOOLCHAIN variable should provide absolute path to external Arm
toolchain install directory. So make that absolute path check explicit.
Signed-off-by: Sumit Garg <sumit.garg@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
CFG_MAP_EXT_DT_SECURE=y should be set per platform, as it requires CFG_DT=y
to also be set, which is not the case for all the platforms out there using
optee-os. Moreover CFG_MAP_EXT_DT_SECURE is already being set conditionally
in optee-os-ts.
Signed-off-by: Denys Dmytriyenko <denys@konsulko.com>
Signed-off-by: Denys Dmytriyenko <denis@denix.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This recipe is a rebuild from source of the Arm GCC release, with
patches from oe-core added to make it work well in our environment.
Most people are happy with the GCC in oe-core, and this release is often
behind: at the time of writing oe-core has mainline GCC 13.1, but Arm
GCC is 12.2. Users who actually want the improvements in Arm GCC will
likely want to use the binary toolchain so that they can have support
from Arm, and they're welcome to do so via the "external-arm" binary
toolchain.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Acked-by: Sumit Garg <sumit.garg@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
There are not any alternatives for the virtual/arm-non-eabi-gcc provider,
so just use the real recipe name.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The binary Arm compiler is based on GCC 12. Remove this GCC 13-specific
option until the next release.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This recipe uses the binary Arm compiler, which is based on GCC 12.
Remove this GCC 13-specific option until the next release.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This recipe uses the binary Arm compiler, which is based on GCC 12.
Remove this GCC 13-specific option until the next release.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Change from using a patched, intermediate SHA to the latest.
Unfortunately, the latest stable mbedtls doesn't boot on tc1.
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update TF-A and fiptool (which is part of tf-a) to the latest stable version.
Also, use the tf-a tests lts branch (which is still at version 2.8.0).
Signed-off-by: Jon Mason <jon.mason@arm.com>
gcc-arm-none-eabi and gcc-aarch64-none-elf both fails when packaged as
RPM for sdk with
nothing provides libcrypt.so.1()(64bit) needed by nativesdk-gcc-arm-none-eabi-12.2.rel1-r0.x86_64_nativesdk
As we don't control the dependencies for prebuilt libraries we create a
workaround by skipping the FILEDEPS
Signed-off-by: Claus Stovgaard <claus.stovgaard@gmail.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
TF-A has LTS releases, which are prefixed with lts- for some reason.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add a link to the FVP EULA in LICENSE_FLAGS_DETAILS, so the user has
some context as to what they're agreeing to upfront.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
armcompiler was removed from meta-arm-toolchain (24c4cfa) so this isn't
needed.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Even though corstone1000 platform does not support the entire PSCI APIs, it
relys on PSCI reset interface for system reset. The name of this config
changed in the new version of u-boot. This enables PSCI reset, so
the system can be resetted in u-boot again.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Split trusted-services.xml into qemuarm64-secureboot-ts.yml and
n1sdp-ts.yml as collection of Trusted Services which can be tested on
each platform has diverged.
Signed-off-by: Adam Johnston <adam.johnston@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Make sure we setup the new variable for the configuration
of the SE-Proxy service for our machine. This will trigger
the right configuration building trusted services and all
psa-arch test pass as before.
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Update Trusted Services and backport an OP-TEE update which allows
interrupting the SPs by NWd interrupts. This solves the kernel stall
problems which are due to long cryptographic operations being executed
in the SWd.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The nanopb build step randomly fails in the yocto CI due to a race condition.
This change adds a patch file to disable parallel build for nanopb. This is a
temporary workaround and a proper fix will be up-streamed int he future.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Patch related with the changes to support the in/out_vec modifications
in TF-M v1.7 was merged in upstream trusted-services integration branch.
So, drop this 3 out of tree patches not needed to be applied any more.
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
TF-A v2.8 does not support measured boot and FF-A which is mandatory for
PSA Initial Attestation SP to work correctly.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add information related to SPMC tests and fix stale links.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Remove already merged patches in trusted services integration
branch to avoid clash during apply patch stage and rebase the
remaining patches.
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Run the ffa_spmc test group of xtest if the optee-spmc-test machine
feature is enabled.
Signed-off-by: Anton Antonov <Anton.Antonov@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add ta-devkit and optee-test. Change configuration to enable building
and deploying OP-TEE SPMC tests.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Anton Antonov <Anton.Antonov@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
The ABI used by the arm-ffa-user driver to call into the SWd changed.
The change was driven by the MM over FF-A ABI implementation which is
used by SmmGW SP and uefi-test. uefi-test uses the same arm-ffa-user
driver as xtest hence xtest needs to be updated to use the new driver.
This xtest change is already merged up-stream but after v3.20, which is
used here.
This change adds backported xtest changes as carried patches.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This change:
- cherry-picks TF-A changes from master which implement passing
TOS_FW_CONFIG DTB from the FIP package to the trusted OS.
- add an OP-TEE SPMC specific SPMC manifest file
- configures TF-A to build the manifest, add it to the FIP package
and pass it to OP-TEE as a boot argument.
This functionality needs matching changes in OPTEE (OP-TEE v3.21
or v3.20 + carried patches.)
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
OP-TEE SPMC v3.20 and TF-A v2.8 is incompatible on qemu, and OP-TEE
panics during boot because having an SPMC manifest passed to the SPMC is
mandatory since v3.20. TF-A and OP-TEE upstream already fixed this issue
by modifying the ABI between the SPMD and SPMC. Moreover qemu support in
TF-A has been extended to allow building an SPMC manifest DTS file, and
loading it from the FIP package.
This change adds the needed OP-TEE fixes as carried patches. The TF-A
change will be added in the next commit.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Split tests to groups, and enable groups based on machine features set.
This allows limiting tests to testing deployed SPs only.
Signed-off-by: Anton Antonov <Anton.Antonov@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
To enable up-to date version of Trusted Services op-tee v3.20 or newer
is needed.
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This change updates to latest available version of Trusted Services.
List of changes:
- adapt SP recipes to file structure changes and support for
"configurations". In TS each SP can be built in various different
setups to allow adapting to platform and integration specific
differences.
- MbedTLS dependency has been updated to v3.3.0.
- This needs new python dependencies are required in the build
environment.
- psa-acs was updated to a matching version.
- do_patch() has been updated to support the MbedTLS patch added
in TS.
- Update TS dependency patching method to use git instead of patch.
- Downgrade nanopb to match up-stream dependency version.
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
Signed-off-by: Gyorgy Szing <Gyorgy.Szing@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Issue when building the kernel on FVP (and probably all aarch64
platforms) with GCC 13 on the 6.1 kernel (and possibly others).
Backport the upstream fix.
Signed-off-by: Jon Mason <jon.mason@arm.com>
There are cases where a developer might want to enable things, like
debug-tweaks, which are useful in their testing but not something we
would want in a production environment. Create a file where these can
be added without affecting other things.
Signed-off-by: Jon Mason <jon.mason@arm.com>
virtual/arm-none-eabi-gcc-native is still at version 12 which
doesn't support it. poky comes with gcc 13 already.
Signed-off-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Add musl testing to have qemuarm-secureboot match qemuarm64-secureboot.
Since the Arm GCC binary toolchain cannot work with musl, move that out
into it's own testing.
Signed-off-by: Jon Mason <jon.mason@arm.com>
debug-tweaks is useful in testing and internal usage, but is a massive
security hole (as it allow password-less root login). Remove the
default enablement on machine files and in kas base yml.
Signed-off-by: Jon Mason <jon.mason@arm.com>
It is only necessary to accept the FVP usage license when using FVPs.
So, move that to the fvp.yml file from the base.yml file to make things
a little cleaner.
Signed-off-by: Jon Mason <jon.mason@arm.com>
The fvp-base machine only has minimal patches, so should be good to
always track the latest release of u-boot.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
Using absolute path in fvpconf will leak the host machine path.
This is a bit annoying when the builder and the runner doesn't use
the same filepath hierachy.
Switch to relative path instead of absolute.
Signed-off-by: Clément Péron <peron.clem@gmail.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
In Order to be able to have filepath relative to fvpconf, execute the
fvp process in the same working directory.
Signed-off-by: Clément Péron <peron.clem@gmail.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
At the moment the config is load and pass to FVPRunner.
Change the ownership to FVPRunner.
Signed-off-by: Clément Péron <peron.clem@gmail.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
We access the dictionnary element that doesn't exist.
Use the get() method instead that will default the element to None if it
doesn't exist.
Signed-off-by: Clément Péron <peron.clem@gmail.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
When running an FVP machine the model executable need to be found
in the PATH environement.
At the moment the script doesn't provide any PATH to the subprocess.
Add PATH to the allowed environement variable to be forwaded.
Signed-off-by: Clément Péron <peron.clem@gmail.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
It fixes and limits the partition size to fix capsule update feature
after the GPT changes.
The partitions in the second bank needs to have correct size and
the partitions in first bank should have a fixed size since corstone1000 does
not support partial update and has a limited flash to support variable size.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
This patch aligns capsule update feature in tfm with GPT/BL1 changes.
Adjusts BL2 flash and data size and adds missing CRC checks.
Signed-off-by: Emekcan Aras <emekcan.aras@arm.com>
Signed-off-by: Jon Mason <jon.mason@arm.com>
2023-05-16 10:42:55 -04:00
752 changed files with 35064 additions and 55441 deletions
@@ -4,32 +4,36 @@ OE-Core's [oeqa][OEQA] framework provides a method of performing runtime tests o
Tests can be configured to run automatically post-build by setting the variable `TESTIMAGE_AUTO="1"`, e.g. in your Kas file or local.conf.
There are two main methods of testing, using different test "targets". Both test targets generate an additional log file with the prefix 'fvp_log' in the image recipe's `${WORKDIR}/testimage` containing the FVP's stdout.
## OEFVPTarget
This runs test cases on a machine using SSH. It therefore requires that an SSH server is installed in the image.
In test cases, the primary interface with the target is, e.g:
```
(status, output) = self.target.run('uname -a')
```
which runs a single command on the target (using `ssh -c`) and returns the status code and the output. It is therefore useful for running tests in a Linux environment.
For examples of test cases, see meta/lib/oeqa/runtime/cases in OE-Core. The majority of test cases depend on `ssh.SSHTest.test_ssh`, which first validates that the SSH connection is functioning.
Example machine configuration:
meta-arm provides the OEFVPTarget which must be set up in the machine configuration:
The test target also generates a log file with the prefix 'fvp_log' in the image recipe's `${WORKDIR}/testimage` containing the FVP's stdout.
This runs tests against one or more serial consoles on the FVP. It is more flexible than OEFVPTarget, but test cases written for this test target do not support the test cases in OE-core. As it does not require an SSH server, it is suitable for machines with performance or memory limitations.
OEFVPTarget supports two different test interfaces - SSH and pexpect.
## SSH
As in OEQA in OE-core, tests cases can run commands on the machine using SSH. It therefore requires that an SSH server is installed in the image.
This uses the `run` method on the target, e.g:
```
(status, output) = self.target.run('uname -a')
```
which executes a single command on the target (using `ssh -c`) and returns the status code and the output. It is therefore useful for running tests in a Linux environment.
For examples of test cases, see meta/lib/oeqa/runtime/cases in OE-Core. The majority of test cases depend on `ssh.SSHTest.test_ssh`, which first validates that the SSH connection is functioning.
## pexpect
To support firmware and baremetal testing, OEFVPTarget also allows test cases to make assertions against one or more consoles using the pexpect library.
Internally, this test target launches a [Pexpect][PEXPECT] instance for each entry in FVP_CONSOLES which can be used with the provided alias. The whole Pexpect API is exposed on the target, where the alias is always passed as the first argument, e.g.:
For an example of a full test case, see meta-arm/lib/oeqa/runtime/cases/linuxboot.py This test case can be used to minimally verify that a machine boots to a Linux shell. The default timeout is 10 minutes, but this can be configured with the variable TEST_FVP_LINUX_BOOT_TIMEOUT, which expects a value in seconds.
The SSH interface described above is also available on OEFVPSerialTarget to support writing a set of hybrid test suites that use a combination of serial and SSH access. Note however that this test target does not guarantee that Linux has booted to shell prior to running any tests, so the test cases in OE-core are not supported.
@@ -55,7 +55,7 @@ The name of the FVP binary itself, for example `fvp-base` uses `FVP_Base_RevC-2x
The name of the recipe that provides the FVP executable set in `FVP_EXE`, for example `fvp-base` uses `fvp-base-a-aem-native`. This *must* be a `-native` recipe as the binary will be executed on the build host.
There are recipes for common FVPs in meta-arm already, and writing new recipes is trivial. For FVPs which are free to download `fvp-base-a-aem.bb` is a good example. Some FVPs must be downloaded separately as they need an account on Arm's website, `fvp-base-r-aem.bb` is a good example of those.
There are recipes for common FVPs in meta-arm already, and writing new recipes is trivial. For FVPs which are free to download `fvp-base-a-aem.bb` is a good example. Some FVPs must be downloaded separately as they need an account on Arm's website.
If `FVP_PROVIDER` is not set then it is assumed that `FVP_EXE` is installed on the host already.
@@ -64,7 +64,7 @@ If `FVP_PROVIDER` is not set then it is assumed that `FVP_EXE` is installed on t
Parameters passed to the FVP with the `--parameter`/`-C` option. These are expressed as variable flags so individual parameters can be altered easily. For example:
Specify raw data to load at the specified address, passed to the FVP with the `--data` option. This is a space-separated list of parameters in the format `[INST=]FILE@[MEMSPACE:]ADDRESS`. For example:
Applications to load on the cores, passed to the FVP with the `--application` option. These are expressed as variable flags with the flag name being the instance and flag value the filename, for example:
Note that symbols are not allowed in flag names, so if you need to use a wildcard in the instance then you'll need to use `FVP_EXTRA_ARGS` and `--application` directly.
@@ -19,6 +19,46 @@ intended for safety-critical applications. Should Your Software or Your Hardware
prove defective, you assume the entire cost of all necessary servicing, repair
or correction.
***********************
Release notes - 2024.06
***********************
Known Issues or Limitations
---------------------------
- Use Ethernet over VirtIO due to lan91c111 Ethernet driver support dropped from U-Boot.
- Due to the performance uplimit of MPS3 FPGA and FVP, some Linux distros like Fedora Rawhide can not boot on Corstone-1000 (i.e. user may experience timeouts or boot hang).
- Corstone-1000 SoC on FVP doesn't have a secure debug peripheral. It does on the MPS3.
- See previous release notes for the known limitations regarding ACS tests.
Platform Support
-----------------
- This software release is tested on Corstone-1000 FPGA version AN550_v2
- Use Ethernet over VirtIO due to lan91c111 Ethernet driver support dropped from U-Boot.
- Temporally removing the External system support in Linux due to it using multiple custom devicetree bindings that caused problems with SystemReady IR 2.0 certification. For External system support please refer to the version 2023.06. We are aiming to restore it in a more standardised manner in our next release.
- Due to the performance uplimit of MPS3 FPGA and FVP, some Linux distros like Fedora Rawhide can not boot on Corstone-1000 (i.e. user may experience timeouts or boot hang).
- PSA Crypto tests (psa-crypto-api-test command) approximately take 30 minutes to complete for FVP and MPS3.
- Corstone-1000 SoC on FVP doesn't have a secure debug peripheral. It does on the MPS3.
- See previous release notes for the known limitations regarding ACS tests.
Platform Support
-----------------
- This software release is tested on Corstone-1000 FPGA version AN550_v2
For technical support email: support-subsystem-iot@arm.com
For all security issues, contact Arm by email at arm-security@arm.com.
For all security issues, contact Arm by email at psirt@arm.com.
--------------
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
James McGregor
Ross Burton
Luca Fancellu
Amr Mohamed
Ziad Elhanafy
Abdellatif El Khlifi
Bence Balogh
Jon Mason
Delane Brandy
Ben Cownley
Harsimran Singh Tungal
Ali Can Ozaslan
Emekcan Aras
Drew Reed
Mikko Rapeli
Gyorgy Szing
Adam Johnston
Amr Mohamed
Debbie Martin
Anusmita Dutta Mazumder
Mathieu Poirier
Alexander Sverdlin
Khem Raj
Vincent Stehlé
Vikas Katariya
Mariam Elshakfy
Xueliang Zhong
Javier Tia
Divin Raj
Peter Hoyes
Robbie Cao
Rouven Czerwinski
Tomás González
Sumit Garg
Denys Dmytriyenko
Claus Stovgaard
Rui Miguel Silva
Anton Antonov
Clément Péron
