ci: QEMUv8: check Firmware Handoff

[raymo200915]

Add a check entry for Arm Firmware Handoff.

This patch depends on:
OP-TEE/build#821
OP-TEE/manifest#321
OP-TEE/manifest#325

[jforissier]

This PR supposedly depends on OP-TEE/build#821, yet CI passed when that PR was not merged yet. So something is clearly wrong (CI is not testing what it should?)

[jforissier]

Oh I suppose this needs #7352 and OP-TEE/optee_test#788 to be meaningful and trigger the tests.

[raymo200915]

Oh I suppose this needs #7352 and OP-TEE/optee_test#788 to be meaningful and trigger the tests.

Yea, this PR should be merged after all related ones are merged, and then trigger the tests.

[jforissier]

Acked-by: Jerome Forissier <jerome.forissier@linaro.org>

@raymo200915 please apply the tag and push again so we can check CI. Thanks.

[raymo200915]

Acked-by: Jerome Forissier <jerome.forissier@linaro.org>

@raymo200915 please apply the tag and push again so we can check CI. Thanks.

Done.

[raymo200915]

@jforissier This patch seems trigger a bug in 'qemu-check.exp': $star in proc check_test_result arg is not initialized before it is used.
I think below two lines should be added at the beginning of the precedure:

set star 0
set ncases 0

If you feel good I can submit the patch.

[raymo200915]

I found that the TF-A rev v2.12 does not fullfil the Firmware Handoff requirement. The manifest has to be updated to point to my previously merged commit cc58f08fe6a53ab375322dbdec6edbe5707110b8 or current main branch.
Please see my PR at: OP-TEE/manifest#325

[jforissier]

@jforissier This patch seems trigger a bug in 'qemu-check.exp': $star in proc check_test_result arg is not initialized before it is used. I think below two lines should be added at the beginning of the precedure:

set star 0
set ncases 0

If you feel good I can submit the patch.

Sounds good to me, thanks.

[jforissier]

@raymo200915 there is now a merge conflict, can you please resolve it? Thanks.

[raymo200915]

Fixed the procedure check_test_result un-initialized variables at: OP-TEE/build#832

[raymo200915]

@raymo200915 there is now a merge conflict, can you please resolve it? Thanks.

Rebased.

[raymo200915]

The CI has to be re-triggered after OP-TEE/manifest#325 is merged.

[jforissier]

@raymo200915 can you please rebase onto master and force push?

[jforissier]

The CI failure in job "make check (QEMUv8, Xen FF-A)" is a known issue and is being dealt with. See #7394.
The CI failure in job "make check (QEMUv8, BTI+MTE+PAC)" is addressed in #7395.

[raymo200915]

The CI failure in job "make check (QEMUv8, Xen FF-A)" is a known issue and is being dealt with. See #7394. The CI failure in job "make check (QEMUv8, BTI+MTE+PAC)" is addressed in #7395.

Is it able to get the xtest log of QEMUv8 1/2 ? I ran it locally with ARM_FIRMWARE_HANDOFF=y but didn't find issues.

[jforissier]

The CI failure in job "make check (QEMUv8, Xen FF-A)" is a known issue and is being dealt with. See #7394. The CI failure in job "make check (QEMUv8, BTI+MTE+PAC)" is addressed in #7395.

Is it able to get the xtest log of QEMUv8 1/2 ? I ran it locally with ARM_FIRMWARE_HANDOFF=y but didn't find issues.

Unfortunately, no. It would be nice to be able to obtain the file out/bin/serial0.log but it is not available once the CI job has exited.

I too have not been able to reproduce locally. That's weird. I will take a closer look.

[jforissier]

The CI failure in job "make check (QEMUv8, Xen FF-A)" is a known issue and is being dealt with. See #7394. The CI failure in job "make check (QEMUv8, BTI+MTE+PAC)" is addressed in #7395.

Is it able to get the xtest log of QEMUv8 1/2 ? I ran it locally with ARM_FIRMWARE_HANDOFF=y but didn't find issues.

Unfortunately, no. It would be nice to be able to obtain the file out/bin/serial0.log but it is not available once the CI job has exited.

I too have not been able to reproduce locally. That's weird. I will take a closer look.

I found the problem. Please use:

make arm-tf-clean && make -j$(nproc) check ARM_FIRMWARE_HANDOFF=y

...and CI should pass. Thanks.

[raymo200915]

The CI failure in job "make check (QEMUv8, Xen FF-A)" is a known issue and is being dealt with. See #7394. The CI failure in job "make check (QEMUv8, BTI+MTE+PAC)" is addressed in #7395.

Is it able to get the xtest log of QEMUv8 1/2 ? I ran it locally with ARM_FIRMWARE_HANDOFF=y but didn't find issues.

Unfortunately, no. It would be nice to be able to obtain the file out/bin/serial0.log but it is not available once the CI job has exited.
I too have not been able to reproduce locally. That's weird. I will take a closer look.

I found the problem. Please use:

make arm-tf-clean && make -j$(nproc) check ARM_FIRMWARE_HANDOFF=y

...and CI should pass. Thanks.

Ah, I supposed make check will do the clean-up...
Done.