[igt-dev] [PATCH i-g-t v2 2/2] tests/intel/xe_exec_reset: Fix cm-gt-reset

[Kamil Konieczny]

Hi Brian,
On 2023-12-04 at 11:43:06 -0800, Brian Welty wrote:
> With use of GT_RESET and forcing GT reset, it is expected for the
> xe_wait_ufence to expire. Move the assertion checking into the IGT
> by using __xe_wait_ufence().
> In addition, we cannot validate the exec operation was successful
> for GT_RESET case.
> 

+cc Francois

> Signed-off-by: Brian Welty <brian.welty at intel.com>

LGTM,
Reviewed-by: Kamil Konieczny <kamil.konieczny at linux.intel.com>

> ---
>  tests/intel/xe_exec_reset.c | 21 ++++++++++++++++-----
>  1 file changed, 16 insertions(+), 5 deletions(-)
> 
> diff --git a/tests/intel/xe_exec_reset.c b/tests/intel/xe_exec_reset.c
> index 195e62911f..d322cf4bdd 100644
> --- a/tests/intel/xe_exec_reset.c
> +++ b/tests/intel/xe_exec_reset.c
> @@ -612,16 +612,27 @@ test_compute_mode(int fd, struct drm_xe_engine_class_instance *eci,
>  		return;
>  	}
>  
> -	for (i = 1; i < n_execs; i++)
> -		xe_wait_ufence(fd, &data[i].exec_sync, USER_FENCE_VALUE,
> -			       NULL, THREE_SEC);
> +	for (i = 1; i < n_execs; i++) {
> +		int64_t timeout = THREE_SEC;
> +		int err;
> +
> +		err = __xe_wait_ufence(fd, &data[i].exec_sync, USER_FENCE_VALUE,
> +				       NULL, &timeout);
> +		if (flags & GT_RESET)
> +			/* exec races with reset: may timeout or complete */
> +			igt_assert(err == -ETIME || !err);
> +		else
> +			igt_assert_eq(err, 0);
> +	}
>  
>  	sync[0].addr = to_user_pointer(&data[0].vm_sync);
>  	xe_vm_unbind_async(fd, vm, 0, 0, addr, bo_size, sync, 1);
>  	xe_wait_ufence(fd, &data[0].vm_sync, USER_FENCE_VALUE, NULL, THREE_SEC);
>  
> -	for (i = 1; i < n_execs; i++)
> -		igt_assert_eq(data[i].data, 0xc0ffee);
> +	if (!(flags & GT_RESET)) {
> +		for (i = 1; i < n_execs; i++)
> +			igt_assert_eq(data[i].data, 0xc0ffee);
> +	}
>  
>  	for (i = 0; i < n_exec_queues; i++)
>  		xe_exec_queue_destroy(fd, exec_queues[i]);
> -- 
> 2.38.0
>