[PATCH i-g-t v6 3/3] tests/intel/xe_fault_injection: Do not assert for probe_guc_fail_* functions.

[Bernatowicz, Marcin]

Hi,

On 4/15/2025 10:27 AM, Satyanarayana K V P wrote:
> In the current implementation, test asserts if the captured error is not same as
> injected error. It is possible that the error received is translated to other
> error which can be returned to application and in some scenarios driver retries
> in case of failure and so, no error might be captured.
> 
> Considering above cases, added flags to control the assertion after injecting
> error. Test does not assert for probe_guc_fail_* functions as driver some times
> retries in case of failure. The main idea of injecting errors for these guc
> functions is to check the robustness of the driver.
> 
> Signed-off-by: Satyanarayana K V P <satyanarayana.k.v.p at intel.com>
> Cc: Michał Wajdeczko <michal.wajdeczko at intel.com>
> Cc: Francois Dugast <francois.dugast at intel.com>
> Reviewed-by: Francois Dugast <francois.dugast at intel.com>
> ---
>   tests/intel/xe_fault_injection.c | 14 ++++++++++----
>   1 file changed, 10 insertions(+), 4 deletions(-)
> 
> diff --git a/tests/intel/xe_fault_injection.c b/tests/intel/xe_fault_injection.c
> index 252209308..3c389a268 100644
> --- a/tests/intel/xe_fault_injection.c
> +++ b/tests/intel/xe_fault_injection.c
> @@ -28,6 +28,11 @@
>   #define BO_SIZE		(1024*1024)
>   #define INJECT_ITERATIONS	100
>   
> +enum assert_flags {
> +	INJECT_ASSERT_ON_FAIL,
> +	INJECT_NOASSERT_ON_FAIL,
> +};
> +
>   int32_t inject_iters_raw;
>   struct fault_injection_params {
>   	/* @probability: Likelihood of failure injection, in percent. */
> @@ -223,7 +228,7 @@ static void set_retval(const char function_name[], long long retval)
>    * @xe_wopcm_init:			xe_wopcm_init
>    */
>   static void
> -inject_fault_probe(int fd, char pci_slot[], const char function_name[])
> +inject_fault_probe(int fd, int flags, char pci_slot[], const char function_name[])
>   {
>   	igt_info("Injecting error \"%s\" (%d) in function \"%s\"\n",
>   		 strerror(-INJECT_ERRNO), INJECT_ERRNO, function_name);
> @@ -232,7 +237,8 @@ inject_fault_probe(int fd, char pci_slot[], const char function_name[])
>   	injection_list_add(function_name);
>   	set_retval(function_name, INJECT_ERRNO);
>   	xe_sysfs_driver_do(fd, pci_slot, XE_SYSFS_DRIVER_TRY_BIND);
> -	igt_assert_eq(-errno, INJECT_ERRNO);
> +	if (flags & INJECT_ASSERT_ON_FAIL)
> +		igt_assert_eq(-errno, INJECT_ERRNO);
>   	injection_list_remove(function_name);
>   }

Would it make sense to return the probe result from 
inject_fault_probe(...) and decide to assert/not assert outside?

>   
> @@ -263,7 +269,7 @@ static void probe_fail_guc(int fd, char pci_slot[], const char function_name[],
>   	for (int i = iter_start; i < iter_end; i++) {
>   		fault_params->space = i;
>   		setup_injection_fault(fault_params);
> -		inject_fault_probe(fd, pci_slot, function_name);
> +		inject_fault_probe(fd, INJECT_NOASSERT_ON_FAIL, pci_slot, function_name);

		inject_fault_probe(fd, pci_slot, function_name);

>   		igt_kmod_unbind("xe", pci_slot);
>   	}
>   }
> @@ -575,7 +581,7 @@ igt_main_args("I:", NULL, help_str, opt_handler, NULL)
>   
>   	for (const struct section *s = probe_fail_functions; s->name; s++)
>   		igt_subtest_f("inject-fault-probe-function-%s", s->name)
> -			inject_fault_probe(fd, pci_slot, s->name);
> +			inject_fault_probe(fd, INJECT_ASSERT_ON_FAIL, pci_slot,	s->name);

	igt_assert_eq(INJECT_ERRNO, inject_fault_probe(fd, pci_slot, 
function_name));

>   
>   	for (const struct section *s = guc_fail_functions; s->name; s++)
>   		igt_subtest_f("probe-fail-guc-%s", s->name) {