summaryrefslogtreecommitdiffstats
path: root/src
diff options
context:
space:
mode:
Diffstat (limited to 'src')
-rw-r--r--src/mesa/main/nvvertexec.c104
-rw-r--r--src/mesa/main/nvvertprog.h20
2 files changed, 115 insertions, 9 deletions
diff --git a/src/mesa/main/nvvertexec.c b/src/mesa/main/nvvertexec.c
index 94507aba880..bacbea8343c 100644
--- a/src/mesa/main/nvvertexec.c
+++ b/src/mesa/main/nvvertexec.c
@@ -38,6 +38,9 @@
#include "math/m_matrix.h"
+static const GLfloat zeroVec[4] = { 0, 0, 0, 0 };
+
+
/**
* Load/initialize the vertex program registers.
* This needs to be done per vertex.
@@ -225,13 +228,12 @@ fetch_vector4( const struct vp_src_register *source,
const struct vp_machine *machine,
GLfloat result[4] )
{
- static const GLfloat zero[4] = { 0, 0, 0, 0 };
const GLfloat *src;
if (source->RelAddr) {
const GLint reg = source->Register + machine->AddressReg;
if (reg < 0 || reg > MAX_NV_VERTEX_PROGRAM_PARAMS)
- src = zero;
+ src = zeroVec;
else
src = machine->Registers[VP_PROG_REG_START + reg];
}
@@ -262,13 +264,12 @@ fetch_vector1( const struct vp_src_register *source,
const struct vp_machine *machine,
GLfloat result[4] )
{
- static const GLfloat zero[4] = { 0, 0, 0, 0 };
const GLfloat *src;
if (source->RelAddr) {
const GLint reg = source->Register + machine->AddressReg;
if (reg < 0 || reg > MAX_NV_VERTEX_PROGRAM_PARAMS)
- src = zero;
+ src = zeroVec;
else
src = machine->Registers[VP_PROG_REG_START + reg];
}
@@ -606,7 +607,7 @@ _mesa_exec_vertex_program(GLcontext *ctx, const struct vertex_program *program)
store_vector4( &inst->DstReg, machine, t );
}
break;
- case VP_OPCODE_SUB:
+ case VP_OPCODE_SUB: /* GL_NV_vertex_program1_1 */
{
GLfloat t[4], u[4], sum[4];
fetch_vector4( &inst->SrcReg[0], machine, t );
@@ -618,7 +619,7 @@ _mesa_exec_vertex_program(GLcontext *ctx, const struct vertex_program *program)
store_vector4( &inst->DstReg, machine, sum );
}
break;
- case VP_OPCODE_ABS:
+ case VP_OPCODE_ABS: /* GL_NV_vertex_program1_1 */
{
GLfloat t[4];
fetch_vector4( &inst->SrcReg[0], machine, t );
@@ -629,6 +630,97 @@ _mesa_exec_vertex_program(GLcontext *ctx, const struct vertex_program *program)
store_vector4( &inst->DstReg, machine, t );
}
break;
+ case VP_OPCODE_FLR: /* GL_ARB_vertex_program */
+ {
+ GLfloat t[4];
+ fetch_vector4( &inst->SrcReg[0], machine, t );
+ t[0] = FLOORF(t[0]);
+ t[1] = FLOORF(t[1]);
+ t[2] = FLOORF(t[2]);
+ t[3] = FLOORF(t[3]);
+ store_vector4( &inst->DstReg, machine, t );
+ }
+ break;
+ case VP_OPCODE_FRC: /* GL_ARB_vertex_program */
+ {
+ GLfloat t[4];
+ fetch_vector4( &inst->SrcReg[0], machine, t );
+ t[0] = t[0] - FLOORF(t[0]);
+ t[1] = t[1] - FLOORF(t[1]);
+ t[2] = t[2] - FLOORF(t[2]);
+ t[3] = t[3] - FLOORF(t[3]);
+ store_vector4( &inst->DstReg, machine, t );
+ }
+ break;
+ case VP_OPCODE_EX2: /* GL_ARB_vertex_program */
+ {
+ GLfloat t[4];
+ fetch_vector1( &inst->SrcReg[0], machine, t );
+ t[0] = t[1] = t[2] = t[3] = _mesa_pow(2.0, t[0]);
+ store_vector4( &inst->DstReg, machine, t );
+ }
+ break;
+ case VP_OPCODE_LG2: /* GL_ARB_vertex_program */
+ {
+ GLfloat t[4];
+ fetch_vector1( &inst->SrcReg[0], machine, t );
+ t[0] = t[1] = t[2] = t[3] = LOG2(t[0]);
+ store_vector4( &inst->DstReg, machine, t );
+ }
+ break;
+ case VP_OPCODE_POW: /* GL_ARB_vertex_program */
+ {
+ GLfloat t[4], u[4];
+ fetch_vector1( &inst->SrcReg[0], machine, t );
+ fetch_vector1( &inst->SrcReg[1], machine, u );
+ t[0] = t[1] = t[2] = t[3] = _mesa_pow(t[0], u[0]);
+ store_vector4( &inst->DstReg, machine, t );
+ }
+ break;
+ case VP_OPCODE_XPD: /* GL_ARB_vertex_program */
+ {
+ GLfloat t[4], u[4], cross[4];
+ fetch_vector4( &inst->SrcReg[0], machine, t );
+ fetch_vector4( &inst->SrcReg[1], machine, u );
+ cross[0] = t[1] * u[2] - t[2] * u[1];
+ cross[1] = t[2] * u[0] - t[0] * u[2];
+ cross[2] = t[0] * u[1] - t[1] * u[0];
+ store_vector4( &inst->DstReg, machine, cross );
+ }
+ break;
+ case VP_OPCODE_SWZ: /* GL_ARB_vertex_program */
+ {
+ const struct vp_src_register *source = &inst->SrcReg[0];
+ const GLfloat *src;
+ GLfloat result[4];
+ GLuint i;
+
+ /* Code similar to fetch_vector4() */
+ if (source->RelAddr) {
+ const GLint reg = source->Register + machine->AddressReg;
+ if (reg < 0 || reg > MAX_NV_VERTEX_PROGRAM_PARAMS)
+ src = zeroVec;
+ else
+ src = machine->Registers[VP_PROG_REG_START + reg];
+ }
+ else {
+ src = machine->Registers[source->Register];
+ }
+
+ /* extended swizzling here */
+ for (i = 0; i < 3; i++) {
+ if (source->Swizzle[i] == SWIZZLE_ZERO)
+ result[i] = 0.0;
+ else if (source->Swizzle[i] == SWIZZLE_ONE)
+ result[i] = -1.0;
+ else
+ result[i] = -src[source->Swizzle[i]];
+ if (source->Negate)
+ result[i] = -result[i];
+ }
+ store_vector4( &inst->DstReg, machine, result );
+ }
+ break;
case VP_OPCODE_END:
return;
diff --git a/src/mesa/main/nvvertprog.h b/src/mesa/main/nvvertprog.h
index b96d1133eb6..44207e2e4c8 100644
--- a/src/mesa/main/nvvertprog.h
+++ b/src/mesa/main/nvvertprog.h
@@ -48,6 +48,11 @@
#define VP_PROG_REG_END (VP_PROG_REG_START + VP_NUM_PROG_REGS - 1)
+/* for GL_ARB_v_p SWZ instruction */
+#define SWIZZLE_ZERO 100
+#define SWIZZLE_ONE 101
+
+
/* Vertex program opcodes */
enum vp_opcode
{
@@ -72,15 +77,24 @@ enum vp_opcode
VP_OPCODE_RCC,
VP_OPCODE_SUB,
VP_OPCODE_ABS,
- VP_OPCODE_END
+ VP_OPCODE_END,
+ /* Additional opcodes for GL_ARB_vertex_program */
+ VP_OPCODE_FLR,
+ VP_OPCODE_FRC,
+ VP_OPCODE_EX2,
+ VP_OPCODE_LG2,
+ VP_OPCODE_POW,
+ VP_OPCODE_XPD,
+ VP_OPCODE_SWZ
};
/* Instruction source register */
struct vp_src_register
{
- GLint Register; /* or the offset from the address register */
- GLuint Swizzle[4];
+ GLint Register; /* or the offset from the address register */
+ GLubyte Swizzle[4]; /* Each value is 0,1,2,3 for x,y,z,w or */
+ /* SWIZZLE_ZERO or SWIZZLE_ONE for VP_OPCODE_SWZ. */
GLboolean Negate;
GLboolean RelAddr;
};