[PATCH 4/5] tests/intel/xe_exec_system_allocator: Set data to non-zero value

[Ghimiray, Himal Prasad]

On 14-07-2025 22:33, Matthew Brost wrote:
> The compiler seems to optimize out memset of value 0, set to non-zero
> value to fault in the pages.
> 
> Signed-off-by: Matthew Brost <matthew.brost at intel.com>
> ---
>   tests/intel/xe_exec_system_allocator.c | 16 ++++++++--------
>   1 file changed, 8 insertions(+), 8 deletions(-)
> 
> diff --git a/tests/intel/xe_exec_system_allocator.c b/tests/intel/xe_exec_system_allocator.c
> index 006a3dc920..dd5303855d 100644
> --- a/tests/intel/xe_exec_system_allocator.c
> +++ b/tests/intel/xe_exec_system_allocator.c
> @@ -294,7 +294,7 @@ static void touch_all_pages(int fd, uint32_t exec_queue, void *ptr,
>   			   PROT_WRITE, MAP_SHARED |
>   			   MAP_ANONYMOUS, -1, 0);
>   	igt_assert(exec_ufence != MAP_FAILED);
> -	memset(exec_ufence, 0, SZ_4K);
> +	memset(exec_ufence, 5, SZ_4K);
>   	sync[0].addr = to_user_pointer(exec_ufence);
>   
>   	for (i = 0; i < n_writes; ++i, addr += stride) {
> @@ -678,7 +678,7 @@ partial(int fd, struct drm_xe_engine_class_instance *eci, unsigned int flags)
>   	data = mmap(alloc.ptr, bo_size, PROT_READ | PROT_WRITE,
>   		    MAP_PRIVATE | MAP_ANONYMOUS | MAP_FIXED, -1, 0);
>   	igt_assert(data != MAP_FAILED);
> -	memset(data, 0, bo_size);
> +	memset(data, 5, bo_size);
>   	old = data;
>   
>   	exec_queue = xe_exec_queue_create(fd, vm, eci, 0);
> @@ -692,7 +692,7 @@ partial(int fd, struct drm_xe_engine_class_instance *eci, unsigned int flags)
>   			   PROT_WRITE, MAP_SHARED |
>   			   MAP_ANONYMOUS, -1, 0);
>   	igt_assert(exec_ufence != MAP_FAILED);
> -	memset(exec_ufence, 0, SZ_4K);
> +	memset(exec_ufence, 5, SZ_4K);
>   
>   	for (i = 0; i < 2; i++) {
>   		uint64_t addr = to_user_pointer(data);
> @@ -1109,7 +1109,7 @@ test_exec(int fd, struct drm_xe_engine_class_instance *eci,
>   			igt_assert(data);
>   		}
>   		if (!(flags & SKIP_MEMSET))
> -			memset(data, 0, bo_size);
> +			memset(data, 5, bo_size);
>   		if (flags & LOCK) {
>   			igt_assert(!(flags & NEW));
>   			mlock(data, bo_size);
> @@ -1151,7 +1151,7 @@ test_exec(int fd, struct drm_xe_engine_class_instance *eci,
>   				   PROT_WRITE, MAP_SHARED |
>   				   MAP_ANONYMOUS, -1, 0);
>   		igt_assert(exec_ufence != MAP_FAILED);
> -		memset(exec_ufence, 0, SZ_4K);
> +		memset(exec_ufence, 5, SZ_4K);
>   	}
>   
>   	aligned_alloc_type = __aligned_alloc(SZ_4K, SZ_4K);
> @@ -1351,7 +1351,7 @@ test_exec(int fd, struct drm_xe_engine_class_instance *eci,
>   					igt_assert_eq(data[idx].data,
>   						      READ_VALUE(&data[idx]));
>   					if (flags & PREFETCH_BENCHMARK)
> -						memset(data, 0, bo_size);
> +						memset(data, 5, bo_size);
>   
>   					if (flags & MULTI_FAULT) {
>   						for (j = 1; j < N_MULTI_FAULT; ++j) {
> @@ -1432,7 +1432,7 @@ test_exec(int fd, struct drm_xe_engine_class_instance *eci,
>   			}
>   			addr = to_user_pointer(data);
>   			if (!(flags & SKIP_MEMSET))
> -				memset(data, 0, bo_size);
> +				memset(data, 5, bo_size);
>   		}
>   
>   		prev_idx = idx;
> @@ -1586,7 +1586,7 @@ threads(int fd, int n_exec_queues, int n_execs, size_t bo_size,
>   		alloc = aligned_alloc(SZ_2M, alloc_size);
>   		igt_assert(alloc);
>   
> -		memset(alloc, 0, alloc_size);
> +		memset(alloc, 5, alloc_size);

LGTM
Reviewed-by: Himal Prasad Ghimiray <himal.prasad.ghimiray at intel.com>


>   		flags &= ~SHARED_ALLOC;
>   	}
>