Shader Assembly Language (ARB/NV)
Quick Reference Guide for OpenGL®
Describes OpenGL v2.0+ Assembly Language features as specified in the specs
ARB_vertex_program, ARB_fragment_program, NV_vertex_program1-4,
NV_fragment_program1-4, NV_geometry_program4, and NV_gpu_program4.
OpenGL® SUPPORT FUNCTIONS
glGenProgramsARB() or glGenProgramsNV()
glBindProgramARB() or glBindProgramNV()
glProgramStringARB() or glLoadProgramNV()
glDeleteProgramsARB() or glDeleteProgramsNV()
glVertexAttrib*ARB() - sets per-vertex attribute value
glProgramEnvParameter*ARB() - sets global environment constants
glProgramLocalParameter*ARB() - sets per-program constants
DATA TYPES
All ARB assembly variables are of type float4 vectors.
ARB registers are scalar variables were only one float element may be addressed.
PARAMETER TYPES
Attrib per-vertex attributes such as vertex normals
Local parameters applied across a program's given shader pass.
Env parameters are applied across all programs.
DATA TYPE QUALIFIERS
ADDRESS variables are registers.
ATTRIB per-vertex attributes.
PARAM uniform properties - constants, Env or Local.
TEMP temporary variables
ALIAS provides alternate names for variables.
OUTPUT designates variables that are passed back to the pipeline
VARIABLE TYPE MODIFIERS
SHORT, LONG, INT, UINT, FLOAT e.g. INT TEMP myInteger;
VECTOR COMPONENTS
f
ull swizzling supported, component names may not be mixed across sets
x, y, z, w
e.g. ADD R0.xy, V1.zw, V2.zw; # adds zw components of V1,V2 and
r, g, b, a # stores result in xy of R0
MATRIX MODIFIERS
inverse e.g. state.matrix.modelview.invtrans
transpose
invtrans
PROGRAM SYNTAX
Comment line: # some comment here
Start program line: !!ARBvp*, !!NVvp*, !!ARBfp*, !!NVfp*, !!NVgp4.0, etc…
End program line: END
*=1.0-4.0
ARB_vertex/fragment_program INSTRUCTIONS (section 2.14.4 / 3.11.4)
Instruction Output Input Description
----------- ------ ----- -----------
ABS v v absolute value
ADD v v,v add
ARL a s address register load
CMP v v,v,v compare
COS ssss s cosine with reduction to [-PI,PI]
DP3 ssss v,v 3-component dot product
DP4 ssss v,v 4-component dot product
DPH ssss v,v homogeneous dot product
DST v v,v distance vector
EX2 ssss s exponential base 2
EXP v s exponential base 2 (approximate)
FLR v v floor
FRC v v fraction
KIL v v kill fragment
LG2 ssss s logarithm base 2
LIT v v compute light coefficients
LOG v s logarithm base 2 (approximate)
LRP v v,v,v linear interpolation
MAD v v,v,v multiply and add
MAX v v,v maximum
MIN v v,v minimum
MOV v v move
MUL v v,v multiply
POW ssss s,s exponentiate
RCP ssss s reciprocal
RSQ ssss s reciprocal square root
SCS ss-- s sine/cosine without reduction
SGE v v,v set on greater than or equal
SIN ssss s sine with reduction to [-PI,PI]
SLT v v,v set on less than
SUB v v,v subtract
SWZ v v extended swizzle
TEX v v,u,t texture sample
TXB v v,u,t texture sample with bias
TXP v v,u,t texture sample with projection
XPD v v,v cross product
Key:
v = indicates a floating-point vector input or output
s = a floating-point scalar input
ssss = indicates a scalar output replicated across a 4-component result vector
ss-- = indicates two scalar outputs in the first two components
a = indicates a single address register component.
u = indicates a texture image unit identifier
t = indicates a texture tar
g
et
RELATIVE ADDRESSING (NV_gpu_program4)
This extension allows users to declare attribute, result, and temporary arrays such as:
PARAM lookup[] = { program.env[100..104] }; # declare an array of 5 variables (float4’s) to program Env’s
INT TEMP idx; # declare an integer variable for array indexing
ROUND.U idx.x, someFloat.x; # rounds a float variable to nearest unsigned integer index
MOV R0.x, lookup[idx.x]; # index into our array and store the variables x component in R0
资源评论
lfrdreamman19862013-03-24很有用，解决了大问题
