Linux Certif

Toute la documentation sur la certification Linux LPI

step.3Cg

NAME

step - implement a step function returning either zero or one

SYNOPSIS

   float  step(float  a, float  x);
   float1 step(float1 a, float1 x);
   float2 step(float2 a, float2 x);
   float3 step(float3 a, float3 x);
   float4 step(float4 a, float4 x);
 
 
 

   half   step(half  a, half  x);
   half1  step(half1 a, half1 x);
   half2  step(half2 a, half2 x);
   half3  step(half3 a, half3 x);
   half4  step(half4 a, half4 x);
 
 
 

   fixed  step(fixed  a, fixed  x);
   fixed1 step(fixed1 a, fixed1 x);
   fixed2 step(fixed2 a, fixed2 x);
   fixed3 step(fixed3 a, fixed3 x);
   fixed4 step(fixed4 a, fixed4 x);
 
 
 

PARAMETERS

a

Scalar or vector reference value.

x

Scalar or vector.

DESCRIPTION

Implements a step function returning one for each component of x that is greater than or equal to the corresponding component in the reference vector a, and zero otherwise.

REFERENCE IMPLEMENTATION

step for float3 vectors could be implemented this way:

   float3 step(float3 a, float3 x)
   {
     return x >= a;
   }
 
 
 

PROFILE SUPPORT

step is supported in all profiles.

Support in the fp20 is limited.

SEE ALSO

max, min, saturate, smoothstep