[PATCH] drm/xe/ring_ops_types: Increase MAX_JOB_SIZE_DW

[Upadhyay, Tejas]

> -----Original Message-----
> From: Intel-xe <intel-xe-bounces at lists.freedesktop.org> On Behalf Of
> Francois Dugast
> Sent: 15 April 2025 16:45
> To: intel-xe at lists.freedesktop.org
> Cc: Dugast, Francois <francois.dugast at intel.com>
> Subject: [PATCH] drm/xe/ring_ops_types: Increase MAX_JOB_SIZE_DW
> 
> This limit can be hit, for example when running IGT compute-square on MTL:
> 
>     [   63.582385] xe 0000:00:02.0: [drm] Assertion `i <= 48` failed!
>                    platform: METEORLAKE subplatform: 1
>                    graphics: Xe_LPG 12.71 step B0
>                    media: Xe_LPM+ 13.00 step C0
>                    tile: 0 VRAM 0 B
>                    GT: 0 type 1
>     [   63.582408] WARNING: CPU: 0 PID: 2439 at
> drivers/gpu/drm/xe/xe_ring_ops.c:395
> __emit_job_gen12_render_compute+0xa7c/0xc70 [xe]
> 
> Increase the value to allow this test to pass.
> 
> Signed-off-by: Francois Dugast <francois.dugast at intel.com>
> ---
>  drivers/gpu/drm/xe/xe_ring_ops_types.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_ring_ops_types.h
> b/drivers/gpu/drm/xe/xe_ring_ops_types.h
> index 1ae56e2ee7b4..121279c826eb 100644
> --- a/drivers/gpu/drm/xe/xe_ring_ops_types.h
> +++ b/drivers/gpu/drm/xe/xe_ring_ops_types.h
> @@ -8,7 +8,7 @@
> 
>  struct xe_sched_job;
> 
> -#define MAX_JOB_SIZE_DW 48
> +#define MAX_JOB_SIZE_DW 64

To me it looks like this limit should be handled by KMD itself. Any test should not be able to break it. May be your test is triggering some optional bracket in driver to add those extra DWs? DWs used by your test is not counted against above define as far as I know.

Tejas

>  #define MAX_JOB_SIZE_BYTES (MAX_JOB_SIZE_DW * 4)
> 
>  /**
> --
> 2.43.0