Hi Tom,
On Tue, 16 Sept 2025 at 14:27, Tom Rini <tr...@konsulko.com> wrote:
>
> On Tue, Sep 16, 2025 at 09:14:19AM -0700, Raymond Mao wrote:
> > Add test cases to validate FDT and TPM eventlog handoff from TF-A
> > and OP-TEE via bloblist.
> >
> > For FDT, the nodes 'reserved-memory' and 'firmware' appended by
> > OP-TEE indicates a successful handoff.
> >
> > For TPM eventlog, the events 'SECURE_RT_EL3', 'SECURE_RT_EL1_OPTEE'
> > and 'SECURE_RT_EL1_OPTEE_EXTRA1' created by TF-A indicates a
> > successful handoff.
> >
> > Signed-off-by: Raymond Mao <raymond....@linaro.org>
> > ---
> > test/py/tests/test_fw_handoff.py | 94 ++++++++++++++++++++++++++++++++
> > 1 file changed, 94 insertions(+)
> > create mode 100644 test/py/tests/test_fw_handoff.py
> >
> > diff --git a/test/py/tests/test_fw_handoff.py
> > b/test/py/tests/test_fw_handoff.py
> > new file mode 100644
> > index 00000000000..4cdf51a4375
> > --- /dev/null
> > +++ b/test/py/tests/test_fw_handoff.py
> > @@ -0,0 +1,94 @@
> > +# SPDX-License-Identifier: GPL-2.0+
> > +#
> > +# Copyright (c) 2025 Linaro Limited
> > +# Author: Raymond Mao <raymond....@linaro.org>
> > +#
> > +# Validate Firmware Handoff from TF-A and OP-TEE
> > +
> > +import pytest
> > +import re
> > +
> > +def _norm_ws(s: str) -> str:
> > + """Normalize whitespace for robust comparisons."""
> > + return re.sub(r"\s+", " ", s).strip()
> > +
> > +@pytest.mark.buildconfigspec("bloblist")
> > +@pytest.mark.buildconfigspec("cmd_bloblist")
> > +@pytest.mark.buildconfigspec("of_control")
> > +@pytest.mark.buildconfigspec("cmd_fdt")
> > +def test_fw_handoff_dt(ubman):
> > + """Validate FDT handoff via bloblist."""
> > +
> > + if (ubman.config.board_type != 'qemu_arm64' or
> > + ubman.config.board_identity != 'fw_handoff_tfa_optee'):
> > + pytest.skip("Test only for qemu_arm64 with ID
> > fw_handoff_tfa_optee")
>
> This (and the other tests) should be more pytest.mark.buildconfigspec
> but for TARGET_QEMU_ARM_64BIT and then relevant TF-A/OP-TEE options. And
> even then, possibly just the relevant TF-A/OP-TEE options as it should
> work on other platforms too, yes? Thanks.
>
Just like what you commented to my test-hooks patches, if we can merge
and use a general conf qemu_arm64_na, the check for the board ID is no
longer needed.
But still this is for qemu_arm64 only, since the TF-A driver is
platform-dependent, thus the feature currently only works on a few
platforms.
Raymond
> --
> Tom
