|
|
Computes the three-component dot product of the source registers.
| dp3 dst, src0, src1 |
|---|
where
| Pixel shader versions | 1_1 | 1_2 | 1_3 | 1_4 | 2_0 | 2_x | 2_sw | 3_0 | 3_sw |
|---|---|---|---|---|---|---|---|---|---|
| dp3 | x | x | x | x | x | x | x | x | x |
The following code snippet shows the operations performed:
dest.x = dest.y = dest.z = dest.w = (src0.x * src1.x) + (src0.y * src1.y) + (src0.z * src1.z);
This instruction runs in the vector pipeline, always writing out to the color channels. For version 1_4, this instruction still uses the vector pipeline but may write to any channel.
An instruction with a destination register .rgb (.xyz) write mask may be co-issued with dp3 As shown below.
dp3 r0.rgb, t0, v0 // Copy scalar result to color components +mov r2.a, t0 // Copy alpha component from t0 in parallel
The dp3 instruction can be modified using the Source Register Signed Scaling input argument modifier (_bx2) applied to its input arguments if they are not already expanded to signed dynamic range. For a lighting shader, the saturate instruction modifier (_sat) is often used to clamp the negative values to black, as shown in the following example.
dp3_sat r0, t0_bx2, v0_bx2 // t0 is a bump map, v0 contains the light direction