the code below generated from intel GPA tool and i have been studying these HLSL code and i have trouble understanding in what is l register used in the code and what values it holds? how do we know the values in registers what values they are using?
dcl_constantbuffer CB2[16], immediateIndexed
dcl_constantbuffer CB12[13], immediateIndexed
dcl_sampler s6, mode_default
dcl_resource_texture2d (float,float,float,float) t6
dcl_input_ps linear v1.xy
dcl_output o0.xyzw
dcl_temps 6
mov r0.x, cb2[15].z
mov r0.yw, l(0,0,0,0)
add r0.xy, r0.xyxx, v1.xyxx
mad r1.xy, r0.xyxx, l(2.000000, -2.000000, 0.000000, 0.000000), l(-1.000000, 1.000000, 0.000000, 0.000000)
sample_l r2.xyzw, r0.xyxx, t6.xyzw, s6, l(0.000000)
mul r1.xy, r1.xyxx, cb12[0].xyxx
mov r1.z, l(1.000000)
mul r1.xyz, r2.xxxx, r1.xyzx
mov r0.z, -cb2[15].z
add r0.xy, r0.zwzz, v1.xyxx
mad r0.zw, r0.xxxy, l(0.000000, 0.000000, 2.000000, -2.000000), l(0.000000, 0.000000, -1.000000, 1.000000)
sample_l r2.xyzw, r0.xyxx, t6.xyzw, s6, l(0.000000)
mul r0.xy, r0.zwzz, cb12[0].xyxx
mov r0.z, l(1.000000)
mul r0.xyz, r2.xxxx, r0.xyzx
mad r2.xy, v1.xyxx, l(2.000000, -2.000000, 0.000000, 0.000000), l(-1.000000, 1.000000, 0.000000, 0.000000)
mul r2.xy, r2.xyxx, cb12[0].xyxx
sample_l r3.xyzw, v1.xyxx, t6.xyzw, s6, l(0.000000)
mov r2.z, l(1.000000)
mad r3.yzw, r2.xxyz, r3.xxxx, -r0.xxyz
dp3 r0.w, r3.yzwy, r3.yzwy
sqrt r0.w, r0.w
mad r3.yzw, r2.xxyz, r3.xxxx, -r1.xxyz
dp3 r1.w, r3.yzwy, r3.yzwy
sqrt r1.w, r1.w
lt r0.w, r0.w, r1.w
movc r0.xyz, r0.wwww, r0.xyzx, r1.xyzx
mad r0.xyz, -r2.xyzx, r3.xxxx, r0.xyzx
dp3 r0.x, r0.xyzx, r0.xyzx
sqrt r0.x, r0.x
mov r1.y, cb2[15].w
mov r1.xz, l(0,0,0,0)
add r0.yz, r1.xxyx, v1.xxyx
mad r1.xy, r0.yzyy, l(2.000000, -2.000000, 0.000000, 0.000000), l(-1.000000, 1.000000, 0.000000, 0.000000)
sample_l r4.xyzw, r0.yzyy, t6.xyzw, s6, l(0.000000)
mul r5.xy, r1.xyxx, cb12[0].xyxx
mov r5.z, l(1.000000)
mul r0.yzw, r4.xxxx, r5.xxyz
mad r3.yzw, r2.xxyz, r3.xxxx, -r0.yyzw
dp3 r1.x, r3.yzwy, r3.yzwy
sqrt r1.x, r1.x
mov r1.w, -cb2[15].w
add r1.yz, r1.zzwz, v1.xxyx
mad r3.yz, r1.yyzy, l(0.000000, 2.000000, -2.000000, 0.000000), l(0.000000, -1.000000, 1.000000, 0.000000)
sample_l r4.xyzw, r1.yzyy, t6.xyzw, s6, l(0.000000)
mul r5.xy, r3.yzyy, cb12[0].xyxx
mov r5.z, l(1.000000)
mul r1.yzw, r4.xxxx, r5.xxyz
mad r3.yzw, r2.xxyz, r3.xxxx, -r1.yyzw
dp3 r2.w, r3.yzwy, r3.yzwy
sqrt r2.w, r2.w
lt r1.x, r1.x, r2.w
movc r0.yzw, r1.xxxx, r0.yyzw, r1.yyzw
mad r0.yzw, -r2.xxyz, r3.xxxx, r0.yyzw
dp3 r0.y, r0.yzwy, r0.yzwy
sqrt r0.y, r0.y
lt r0.xy, r0.xyxx, cb12[12].wwww
and r0.x, r0.y, r0.x
and o0.xyzw, r0.xxxx, l(0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000)
ret
we can see l is being used many times in above code with different values in brackets like
mad r1.xy, r0.xyxx, l(2.000000, -2.000000, 0.000000, 0.000000), l(-1.000000, 1.000000, 0.000000, 0.000000)
and
and o0.xyzw, r0.xxxx, l(0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000)
i know about and operation but how l getting these values and what could be the final output i'm very new to this please pardon if it seams very dumb thanks in advance
Reading shader bytecode is really not an easy task, for such a complex shader it needs a little patience. To get a deeper understanding I'll recommend to read the documentation, so you can easily lookup operators or registers.
To your questions:
xy and a texture is declared I would assume that it is describing some kind of texture coordinate passed to the shadermad r1.xy, r0.xyxx, l(2.000000, -2.000000, 0.000000, 0.000000), l(-1.000000, 1.000000, 0.000000, 0.000000) =>
r1.xy = r0.xy * float2(2,-2) + float2(-1, 1)and o0.xyzw, r0.xxxx, l(0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000) is a bit special to understand. It and's the registers componentwise, two lines above there is a lt, which means that now in r0.x is a boolean in the form 0xFFFFFFFF (true) or 0x00000000(false). By an and with 0x3f800000 (which is the floating point representation of 1.0), it essentially transforms the boolean to a float either 1.0 or 0.0. So the output of the shader is 1.0 or 0.0 in each channel regarding the boolean value of r0.