Hi,
On Fri, Apr 12, 2024 at 11:33:49AM +0100, Suzuki K Poulose wrote:
> From: Alexandru Elisei <alexandru.elisei@xxxxxxx>
>
> The realm manager treats instruction aborts as fatal errors, skip this
> test.
I wrote the commit message, and now I don't remember if this was a
limitation of the (at the time) implementation for the realm manager that I
used for developing the patch or if this was specified in the architecture
:(
Would you mind clearing that up?
Thanks,
Alex
>
> Signed-off-by: Alexandru Elisei <alexandru.elisei@xxxxxxx>
> Signed-off-by: Joey Gouly <joey.gouly@xxxxxxx>
> Signed-off-by: Suzuki K Poulose <suzuki.poulose@xxxxxxx>
> ---
> arm/selftest.c | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/arm/selftest.c b/arm/selftest.c
> index 1553ed8e..8caadad3 100644
> --- a/arm/selftest.c
> +++ b/arm/selftest.c
> @@ -19,6 +19,7 @@
> #include <asm/smp.h>
> #include <asm/mmu.h>
> #include <asm/barrier.h>
> +#include <asm/rsi.h>
>
> static cpumask_t ready, valid;
>
> @@ -393,11 +394,17 @@ static void check_vectors(void *arg __unused)
> user_psci_system_off);
> #endif
> } else {
> + if (is_realm()) {
> + report_skip("pabt test not supported in a realm");
> + goto out;
> + }
> +
> if (!check_pabt_init())
> report_skip("Couldn't guess an invalid physical address");
> else
> report(check_pabt(), "pabt");
> }
> +out:
> exit(report_summary());
> }
>
> --
> 2.34.1
>
