[igt-dev] [PATCH i-g-t 2/2] tests/gem_exec_basic: Add gem create fallback

Petri Latvala petri.latvala at intel.com
Thu Dec 9 11:21:41 UTC 2021


On Wed, Dec 08, 2021 at 06:24:16PM +0100, Zbigniew Kempczyński wrote:
> If we're on stable kernels which still don't have local memory turned on
> provide a fallback to generic gem_create() call when requested region is
> the system memory.
> 
> Signed-off-by: Zbigniew Kempczyński <zbigniew.kempczynski at intel.com>
> Cc: Ashutosh Dixit <ashutosh.dixit at intel.com>
> ---
>  tests/i915/gem_exec_basic.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/tests/i915/gem_exec_basic.c b/tests/i915/gem_exec_basic.c
> index 008a35d0a..6307e929c 100644
> --- a/tests/i915/gem_exec_basic.c
> +++ b/tests/i915/gem_exec_basic.c
> @@ -28,12 +28,15 @@
>  
>  IGT_TEST_DESCRIPTION("Basic sanity check of execbuf-ioctl rings.");
>  
> -static uint32_t batch_create(int fd, uint32_t batch_size, uint32_t region)
> +static uint32_t batch_create(int fd, uint64_t batch_size, uint32_t region)
>  {
>  	const uint32_t bbe = MI_BATCH_BUFFER_END;
>  	uint32_t handle;
> +	int ret;
>  
> -	handle = gem_create_in_memory_regions(fd, batch_size, region);
> +	ret = __gem_create_in_memory_regions(fd, &handle, &batch_size, region);
> +	if (ret == -ENODEV && IS_SYSTEM_MEMORY_REGION(region))
> +		handle = gem_create(fd, batch_size);
>  	gem_write(fd, handle, 0, &bbe, sizeof(bbe));

Can't gem_create_in_memory_regions() do this fallback?

Anyway,
Acked-by: Petri Latvala <petri.latvala at intel.com>



>  
>  	return handle;
> -- 
> 2.26.0
> 


More information about the igt-dev mailing list