Re: [PATCH] drm/xe/xe_guc_ct: Drop smp_mb in parse_g2h_response

["Summers, Stuart" <stuart.summers@intel.com>]

On Thu, 2025-12-18 at 16:58 +0000, Jonathan Cavitt wrote:
> smp_mb before wake_up_all in parse_g2h_response is superfluous.
> Drop it.
> 
> Suggested-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
> Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
> ---
>  drivers/gpu/drm/xe/xe_guc_ct.c | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_guc_ct.c
> b/drivers/gpu/drm/xe/xe_guc_ct.c
> index c3df9b3f1b4d..d101b37106da 100644
> --- a/drivers/gpu/drm/xe/xe_guc_ct.c
> +++ b/drivers/gpu/drm/xe/xe_guc_ct.c
> @@ -1499,7 +1499,6 @@ static int parse_g2h_response(struct xe_guc_ct
> *ct, u32 *msg, u32 len)
>         g2h_release_space(ct, GUC_CTB_HXG_MSG_MAX_LEN);
>  
>         g2h_fence->done = true;
> -       smp_mb();

I might be missing something here, but where is this getting called in
the current context? I see if we use woken_wake_function() this is
called there, but it looks like we would need to explicitly add that
like what is being done in the i915 driver:
DEFINE_WAIT_FUNC(wait, woken_wake_function);

Which I don't see us doing here. Should we use this wait function
instead?

Thanks,
Stuart

>  
>         wake_up_all(&ct->g2h_fence_wq);
>