Hi Heinrich, On 2026-05-21T18:50:22, Heinrich Schuchardt <[email protected]> wrote: > efi_selftest: add ACPI configuration table test > > The new unit tests checks if ACPI tables are available. > > If yes, it checks if the RSDT correctly points to the RSDT or XSDT. > > Signed-off-by: Heinrich Schuchardt <[email protected]> > > lib/efi_selftest/Makefile | 1 + > lib/efi_selftest/efi_selftest_acpi.c | 75 > ++++++++++++++++++++++++++++++++++++ > 2 files changed, 76 insertions(+)
Reviewed-by: Simon Glass <[email protected]> > efi_selftest: add ACPI configuration table test > > The new unit tests checks if ACPI tables are available. > > If yes, it checks if the RSDT correctly points to the RSDT or XSDT. the first 'RSDT' should be 'RSDP', since it's the pointer. Also 'unit tests checks' should be 'unit test checks'. > diff --git a/lib/efi_selftest/efi_selftest_acpi.c > b/lib/efi_selftest/efi_selftest_acpi.c > @@ -0,0 +1,75 @@ > + if (memcmp(rsdp->signature, RSDP_SIG, sizeof(rsdp->signature))) { > + efi_st_error("Wrong RSDP signature\n"); > + return EFI_ST_FAILURE; > + } I think Ilias says the same thing, but please also verify the RSDP checksum (first 20 bytes for revision 0, full struct for revision 2). > diff --git a/lib/efi_selftest/efi_selftest_acpi.c > b/lib/efi_selftest/efi_selftest_acpi.c > @@ -0,0 +1,75 @@ > + rsdp = efi_st_get_config_table(&acpi2_guid); > + if (!rsdp) > + rsdp = efi_st_get_config_table(&acpi1_guid); > + if (!rsdp) > + return EFI_ST_SUCCESS; Silently returning success when no ACPI table is installed makes this a no-op on every non-ACPI build, with no log indication. The fdt selftest uses setup() returning EFI_ST_FAILURE for the unexpected combination. Could you at least emit an efi_st_printf() saying the test was skipped? Regards, Simon
