Re: [PATCH] drm/i915/display: Do not check crtc_state when initializing BW limits

[Imre Deak <imre.deak@intel.com>]

On Tue, Sep 22, 2026 at 02:46:17AM +0800, Jonathan Cavitt wrote:
> In intel_link_bw_init_limits, we call intel_atomic_get_new_crtc_state to
> grab an intel_crtc_state pointer.  This pointer is later used to set the
> max_bpp_x16 value for the given pipe.  There is a check here for if the
> return value of the intel_atomic_get_new_crtc_state function returns
> NULL, but this is not checked in any other place where this function is
> used.

The other places using intel_atomic_get_new_crtc_state() can omit the
check, since at those places it's guaranteed that the CRTC state for the
crtc passed to the function is part of the atomic state, hence
intel_atomic_get_new_crtc_state() is guaranteed to return a non-NULL
pointer.

> Removing this check makes the code more consistent and prevents
> confusion from static analyzers.
> 
> Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
> Cc: Imre Deak <imre.deak@intel.com>
> ---
>  drivers/gpu/drm/i915/display/intel_link_bw.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/display/intel_link_bw.c b/drivers/gpu/drm/i915/display/intel_link_bw.c
> index e71e76d6fd3e0..2a20a12de9e0a 100644
> --- a/drivers/gpu/drm/i915/display/intel_link_bw.c
> +++ b/drivers/gpu/drm/i915/display/intel_link_bw.c
> @@ -64,8 +64,7 @@ void intel_link_bw_init_limits(struct intel_atomic_state *state,
>  			intel_atomic_get_new_crtc_state(state, crtc);
>  		int forced_bpp_x16 = get_forced_link_bpp_x16(state, crtc);
>  
> -		if ((state->base.duplicated && crtc_state) ||
> -		    intel_dp_mst_stream_disconnected(state, crtc)) {
> +		if (state->base.duplicated || intel_dp_mst_stream_disconnected(state, crtc)) {

The loop goes through all pipes (i.e. CRTCs) and it's not guaranteed
that all CRTCs are part of the atomic state. Hence the crtc_state==NULL
is required here unless intel_dp_mst_stream_disconnected() returns true,
in which case that return value also guarantees that the CRTC is in the
atomic state and hence crtc_state is guaranteed to be non-NULL.

So the above condition in the existing code is correct.

>  			limits->max_bpp_x16[pipe] = crtc_state->max_link_bpp_x16;
>  			if (intel_dsc_enabled_on_link(crtc_state))
>  				limits->link_dsc_pipes |= BIT(pipe);
> -- 
> 2.53.0
>