ps部分

ps_1_1,ps_1_2,ps_1_3,ps_1_4

总览

部分函数细节

bem

语法

bem dst.rg, src0, src1

算法

(Given n == dest register #)

dest.r = src0.r + D3DTSS_BUMPENVMAT00(stage n) * src1.r

                +D3DTSS_BUMPENVMAT10(stage n) * src1.g

dest.g = src0.g + D3DTSS_BUMPENVMAT01(stage n) * src1.r

                +D3DTSS_BUMPENVMAT11(stage n) * src1.g

cmp

语法

cmp dst, src0, src1, src2

如果src0>=0返回src1否则src2

算法

ps_1_4

def c0, -0.6, 0.6, 0, 0.6

def c1  0,0,0,0

def c2  1,1,1,1

mov r1, c1

mov r2, c2

cmp r0, c0, r1, r2   // r0 is assigned 1,0,0,0 based on the following:

// r0.x = c2.x because c0.x < 0

// r0.y = c1.y because c0.y >= 0

// r0.z = c1.z because c0.z >= 0

// r0.w = c1.w because c0.w >= 0

cnd

语法

cmp dst, src0, src1, src2

如果src0>0.5 返回src1否则src2

算法

在1_1到1_3版本，src0必须为r0.a(单通道)

// Version 1_1 to 1_3

if (r0.a > 0.5)

  dest = src1

else

  dest = src2

在1_4版本就可以每个通道分别比较值

for each component in src0

{

   if (src0.component > 0.5)

     dest.component = src1.component

   else

     dest.component = src2.component

}

示例

ps_1_4

def c0, -0.5, 0.5, 0, 0.6

def c1,  0,0,0,0

def c2,  1,1,1,1

cnd r1, c0, c1, c2   // r0 contains 1,1,1,0 because

// r1.x = c2.x because c0.x <= 0.5

// r1.y = c2.y because c0.y <= 0.5

// r1.z = c2.z because c0.z <= 0.5

// r1.w = c1.w because c0.w >  0.5

dp3

语法

dp3 dst, src0, src1

计算三个分量的点积

算法

dest.x = dest.y = dest.z = dest.w = 

  (src0.x * src1.x) + (src0.y * src1.y) + (src0.z * src1.z);

dp4

语法

dp4 dst, src0, src1

计算四个分量的点积

算法

dest.x = dest.y = dest.z = dest.w = 

    (src0.x * src1.x) + (src0.y * src1.y) + 

    (src0.z * src1.z) + (src0.w * src1.w);

lrp

语法

lrp dst, src0, src1, src2

基于src0对src1与src2做线性插值运算

算法

dest = src0 * src1 + (1-src0) * src2

// which is the same as

dest = src2 + src0 * (src1 - src2)

mad

语法

mad dst, src0, src1, src2

做(src0 * src1) + src2处理

算法

dest.x = src0.x * src1.x + src2.x;

dest.y = src0.y * src1.y + src2.y;

dest.z = src0.z * src1.z + src2.z;

dest.w = src0.w * src1.w + src2.w;

mov

语法

mov dst, src

转移值处理

mul

语法

mul dst, src0, src1

乘法

算法

dest.x = src0.x * src1.x;

dest.y = src0.y * src1.y;

dest.z = src0.z * src1.z;

dest.w = src0.w * src1.w;

nop

语法

nop

执行无运算

sub

语法

sub dst, src0, src1

减法运算

算法

dest = src0 - src1

 vs部分

 vs1

总览

 部分函数细节

dst

语法

dst dest, src0, src1

计算距离向量

src0为(ignored, d*d, d*d,ignored)

src1为(ignored, 1/d,ignored, 1/d)

最终得到的结果为(1, d, d*d, 1/d)

算法

dest.x = 1;

dest.y = src0.y * src1.y;

dest.z = src0.z;

dest.w = src1.w;

exp

语法

exp dst, src

算法

dest.x = dest.y = dest.z = dest.w = (float)pow(2, src.replicateSwizzleComponent);

frc

语法

frc dst, src

算法

dest.x = src.x - (float)floor(src.x);

dest.y = src.y - (float)floor(src.y);

dest.z = src.z - (float)floor(src.z);

dest.w = src.w - (float)floor(src.w);

lit

语法

lit dst, src

src的各部分为

src.x = N*L        ; The dot product between normal and direction to light

src.y = N*H        ; The dot product between normal and half vector

src.z = ignored    ; This value is ignored

src.w = exponent   ; The value must be between -128.0 and 128.0

算法

dest.x = 1;

dest.y = 0;

dest.z = 0;

dest.w = 1;

float power = src.w;

const float MAXPOWER = 127.9961f;

if (power < -MAXPOWER)

    power = -MAXPOWER;          // Fits into 8.8 fixed point format

else if (power > MAXPOWER)

    power = MAXPOWER;          // Fits into 8.8 fixed point format

if (src.x > 0)

{

    dest.y = src.x;//diffuse

    if (src.y > 0)

    {

        // Allowed approximation is EXP(power * LOG(src.y))

        dest.z = (float)(pow(src.y, power));//specular

    }

}

log

语法

log dst, src

算法

float v = abs(src);

if (v != 0)

{

    dest.x = dest.y = dest.z = dest.w = 

        (float)(log(v)/log(2));  

}

else

{

    dest.x = dest.y = dest.z = dest.w = -FLT_MAX;

}

logp

语法

logp dst, src

算法

float f = abs(src);

if (f != 0)

    dest.x = dest.y = dest.z = dest.w = (float)(log(f)/log(2));

else

    dest.x = dest.y = dest.z = dest.w = -FLT_MAX;   

m3x2

语法

m3x2 dst, src0, src1

算法

dest.x = (src0.x * src1.x) + (src0.x * src1.y) + (src0.x * src1.z);

dest.y = (src0.x * src2.x) + (src0.y * src2.y) + (src0.z * src2.z);

m3x3

语法

m3x3 dst,src0, src1

算法

dest.x = (src0.x * src1.x) + (src0.y * src1.y) + (src0.z * src1.z);

dest.y = (src0.x * src2.x) + (src0.y * src2.y) + (src0.z * src2.z);

dest.z = (src0.x * src3.x) + (src0.y * src3.y) + (src0.z * src3.z);

m3x4

语法

m3x4 dst, src0, src1

算法

dest.x = (src0.x * src1.x) + (src0.y * src1.y) + (src0.z * src1.z);

dest.y = (src0.x * src2.x) + (src0.y * src2.y) + (src0.z * src2.z);

dest.z = (src0.x * src3.x) + (src0.y * src3.y) + (src0.z * src3.z);

dest.w = (src0.x * src4.x) + (src0.y * src4.y) + (src0.z * src4.z);

m4x3

语法

m4x3dst, src0, src1

算法

dest.x = (src0.x * src1.x) + (src0.y * src1.y) + (src0.z * src1.z) + (src0.w * src1.w);

dest.y = (src0.x * src2.x) + (src0.y * src2.y) + (src0.z * src2.z) + (src0.w * src2.w);

dest.z = (src0.x * src3.x) + (src0.y * src3.y) + (src0.z * src3.z) + (src0.w * src3.w);

m4x4

语法

m4x4 dst, src0, src1

算法

dest.x = (src0.x * src1.x) + (src0.y * src1.y) + (src0.z * src1.z) + 

               (src0.w * src1.w);

dest.y = (src0.x * src2.x) + (src0.y * src2.y) + (src0.z * src2.z) + 

               (src0.w * src2.w);

dest.z = (src0.x * src3.x) + (src0.y * src3.y) + (src0.z * src3.z) + 

               (src0.w * src3.w);

dest.w = (src0.x * src4.x) + (src0.y * src4.y) + (src0.z * src4.z) + 

               (src0.w * src4.w);

max

语法

max dst, src0, src1

算法

dest.x=(src0.x >= src1.x) ? src0.x : src1.x;

dest.y=(src0.y >= src1.y) ? src0.y : src1.y;

dest.z=(src0.z >= src1.z) ? src0.z : src1.z;

dest.w=(src0.w >= src1.w) ? src0.w : src1.w;

min

语法

min dst, src0, src1

算法

dest.x=(src0.x < src1.x) ? src0.x : src1.x;

dest.y=(src0.y < src1.y) ? src0.y : src1.y;

dest.z=(src0.z < src1.z) ? src0.z : src1.z;

dest.w=(src0.w < src1.w) ? src0.w : src1.w;

rcp

语法

rcp dst, src

算法

float f = src0;

if(f == 0.0f)

{

    f = FLT_MAX;

}

else 

{

    if(f != 1.0)

    {

        f = 1/f;

    }

}

dest = f;

rsq

语法

rsq dst, src

算法

float f = abs(src0);

if (f == 0)

    f = FLT_MAX

else

{

    if (f != 1.0)

        f = 1.0/(float)sqrt(f);

}

dest.z = dest.y = dest.z = dest.w = f;

sge

语法

sge dst, src0, src1

src0大于等于 src1返回1否则为0

算法

dest.x = (src0.x >= src1.x) ? 1.0f : 0.0f;

dest.y = (src0.y >= src1.y) ? 1.0f : 0.0f;

dest.z = (src0.z >= src1.z) ? 1.0f : 0.0f;

dest.w = (src0.w >= src1.w) ? 1.0f : 0.0f;

slt

语法

slt dst, src0, src1

src0小于 src1返回1否则为0

算法

dest.x = (src0.x < src1.x) ? 1.0f : 0.0f;

dest.y = (src0.y < src1.y) ? 1.0f : 0.0f;

dest.z = (src0.z < src1.z) ? 1.0f : 0.0f;

dest.w = (src0.w < src1.w) ? 1.0f : 0.0f;

库： https://msdn.microsoft.com/en-us/library/bb219840(v=vs.85).aspx

 --wolf96 2017/1/1

Asm Shader Reference --- Shader Model 1 part