s_nvfragprog.c

winNT技术操作系统,国外开放的原代码和LIUX一样

            }
            break;
         case FP_OPCODE_EX2: /* Exponential base 2 */
            {
               GLfloat a[4], result[4];
               fetch_vector1( ctx, &inst->SrcReg[0], machine, program, a );
               result[0] = result[1] = result[2] = result[3] =
                  (GLfloat) _mesa_pow(2.0, a[0]);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_FLR:
            {
               GLfloat a[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               result[0] = FLOORF(a[0]);
               result[1] = FLOORF(a[1]);
               result[2] = FLOORF(a[2]);
               result[3] = FLOORF(a[3]);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_FRC:
            {
               GLfloat a[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               result[0] = a[0] - FLOORF(a[0]);
               result[1] = a[1] - FLOORF(a[1]);
               result[2] = a[2] - FLOORF(a[2]);
               result[3] = a[3] - FLOORF(a[3]);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_KIL_NV: /* NV_f_p only */
            {
               const GLuint swizzle = inst->DstReg.CondSwizzle;
               const GLuint condMask = inst->DstReg.CondMask;
               if (test_cc(machine->CondCodes[GET_SWZ(swizzle, 0)], condMask) ||
                   test_cc(machine->CondCodes[GET_SWZ(swizzle, 1)], condMask) ||
                   test_cc(machine->CondCodes[GET_SWZ(swizzle, 2)], condMask) ||
                   test_cc(machine->CondCodes[GET_SWZ(swizzle, 3)], condMask)) {
                  return GL_FALSE;
               }
            }
            break;
         case FP_OPCODE_KIL: /* ARB_f_p only */
            {
               GLfloat a[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               if (a[0] < 0.0F || a[1] < 0.0F || a[2] < 0.0F || a[3] < 0.0F) {
                  return GL_FALSE;
               }
            }
            break;
         case FP_OPCODE_LG2:  /* log base 2 */
            {
               GLfloat a[4], result[4];
               fetch_vector1( ctx, &inst->SrcReg[0], machine, program, a );
               result[0] = result[1] = result[2] = result[3]
                  = LOG2(a[0]);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_LIT:
            {
               const GLfloat epsilon = 1.0F / 256.0F; /* from NV VP spec */
               GLfloat a[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               a[0] = MAX2(a[0], 0.0F);
               a[1] = MAX2(a[1], 0.0F);
               /* XXX ARB version clamps a[3], NV version doesn't */
               a[3] = CLAMP(a[3], -(128.0F - epsilon), (128.0F - epsilon));
               result[0] = 1.0F;
               result[1] = a[0];
               /* XXX we could probably just use pow() here */
               result[2] = (a[0] > 0.0F) ? (GLfloat) exp(a[3] * log(a[1])) : 0.0F;
               result[3] = 1.0F;
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_LRP:
            {
               GLfloat a[4], b[4], c[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               fetch_vector4( ctx, &inst->SrcReg[2], machine, program, c );
               result[0] = a[0] * b[0] + (1.0F - a[0]) * c[0];
               result[1] = a[1] * b[1] + (1.0F - a[1]) * c[1];
               result[2] = a[2] * b[2] + (1.0F - a[2]) * c[2];
               result[3] = a[3] * b[3] + (1.0F - a[3]) * c[3];
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_MAD:
            {
               GLfloat a[4], b[4], c[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               fetch_vector4( ctx, &inst->SrcReg[2], machine, program, c );
               result[0] = a[0] * b[0] + c[0];
               result[1] = a[1] * b[1] + c[1];
               result[2] = a[2] * b[2] + c[2];
               result[3] = a[3] * b[3] + c[3];
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_MAX:
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               result[0] = MAX2(a[0], b[0]);
               result[1] = MAX2(a[1], b[1]);
               result[2] = MAX2(a[2], b[2]);
               result[3] = MAX2(a[3], b[3]);
               store_vector4( inst, machine, result );
#if DEBUG_FRAG
               printf("MAX (%g %g %g %g) = (%g %g %g %g), (%g %g %g %g)\n",
                      result[0], result[1], result[2], result[3], 
                      a[0], a[1], a[2], a[3],
                      b[0], b[1], b[2], b[3]);
#endif
            }
            break;
         case FP_OPCODE_MIN:
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               result[0] = MIN2(a[0], b[0]);
               result[1] = MIN2(a[1], b[1]);
               result[2] = MIN2(a[2], b[2]);
               result[3] = MIN2(a[3], b[3]);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_MOV:
            {
               GLfloat result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, result );
               store_vector4( inst, machine, result );
#if DEBUG_FRAG
               printf("MOV (%g %g %g %g)\n",
                      result[0], result[1], result[2], result[3]);
#endif
            }
            break;
         case FP_OPCODE_MUL:
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               result[0] = a[0] * b[0];
               result[1] = a[1] * b[1];
               result[2] = a[2] * b[2];
               result[3] = a[3] * b[3];
               store_vector4( inst, machine, result );
#if DEBUG_FRAG
               printf("MUL (%g %g %g %g) = (%g %g %g %g) * (%g %g %g %g)\n",
                      result[0], result[1], result[2], result[3], 
                      a[0], a[1], a[2], a[3],
                      b[0], b[1], b[2], b[3]);
#endif
            }
            break;
         case FP_OPCODE_PK2H: /* pack two 16-bit floats in one 32-bit float */
            {
               GLfloat a[4], result[4];
               GLhalfNV hx, hy;
               GLuint *rawResult = (GLuint *) result;
               GLuint twoHalves;
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               hx = _mesa_float_to_half(a[0]);
               hy = _mesa_float_to_half(a[1]);
               twoHalves = hx | (hy << 16);
               rawResult[0] = rawResult[1] = rawResult[2] = rawResult[3]
                  = twoHalves;
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_PK2US: /* pack two GLushorts into one 32-bit float */
            {
               GLfloat a[4], result[4];
               GLuint usx, usy, *rawResult = (GLuint *) result;
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               a[0] = CLAMP(a[0], 0.0F, 1.0F);
               a[1] = CLAMP(a[1], 0.0F, 1.0F);
               usx = IROUND(a[0] * 65535.0F);
               usy = IROUND(a[1] * 65535.0F);
               rawResult[0] = rawResult[1] = rawResult[2] = rawResult[3]
                  = usx | (usy << 16);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_PK4B: /* pack four GLbytes into one 32-bit float */
            {
               GLfloat a[4], result[4];
               GLuint ubx, uby, ubz, ubw, *rawResult = (GLuint *) result;
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               a[0] = CLAMP(a[0], -128.0F / 127.0F, 1.0F);
               a[1] = CLAMP(a[1], -128.0F / 127.0F, 1.0F);
               a[2] = CLAMP(a[2], -128.0F / 127.0F, 1.0F);
               a[3] = CLAMP(a[3], -128.0F / 127.0F, 1.0F);
               ubx = IROUND(127.0F * a[0] + 128.0F);
               uby = IROUND(127.0F * a[1] + 128.0F);
               ubz = IROUND(127.0F * a[2] + 128.0F);
               ubw = IROUND(127.0F * a[3] + 128.0F);
               rawResult[0] = rawResult[1] = rawResult[2] = rawResult[3]
                  = ubx | (uby << 8) | (ubz << 16) | (ubw << 24);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_PK4UB: /* pack four GLubytes into one 32-bit float */
            {
               GLfloat a[4], result[4];
               GLuint ubx, uby, ubz, ubw, *rawResult = (GLuint *) result;
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               a[0] = CLAMP(a[0], 0.0F, 1.0F);
               a[1] = CLAMP(a[1], 0.0F, 1.0F);
               a[2] = CLAMP(a[2], 0.0F, 1.0F);
               a[3] = CLAMP(a[3], 0.0F, 1.0F);
               ubx = IROUND(255.0F * a[0]);
               uby = IROUND(255.0F * a[1]);
               ubz = IROUND(255.0F * a[2]);
               ubw = IROUND(255.0F * a[3]);
               rawResult[0] = rawResult[1] = rawResult[2] = rawResult[3]
                  = ubx | (uby << 8) | (ubz << 16) | (ubw << 24);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_POW:
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector1( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector1( ctx, &inst->SrcReg[1], machine, program, b );
               result[0] = result[1] = result[2] = result[3]
                  = (GLfloat)_mesa_pow(a[0], b[0]);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_RCP:
            {
               GLfloat a[4], result[4];
               fetch_vector1( ctx, &inst->SrcReg[0], machine, program, a );
#if DEBUG_FRAG
               if (a[0] == 0)
                  printf("RCP(0)\n");
               else if (IS_INF_OR_NAN(a[0]))
                  printf("RCP(inf)\n");
#endif
               result[0] = result[1] = result[2] = result[3]
                  = 1.0F / a[0];
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_RFL:
            {
               GLfloat axis[4], dir[4], result[4], tmp[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, axis );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, dir );
               tmp[3] = axis[0] * axis[0]
                      + axis[1] * axis[1]
                      + axis[2] * axis[2];
               tmp[0] = (2.0F * (axis[0] * dir[0] +
                                 axis[1] * dir[1] +
                                 axis[2] * dir[2])) / tmp[3];
               result[0] = tmp[0] * axis[0] - dir[0];
               result[1] = tmp[0] * axis[1] - dir[1];
               result[2] = tmp[0] * axis[2] - dir[2];
               /* result[3] is never written! XXX enforce in parser! */
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_RSQ: /* 1 / sqrt() */
            {
               GLfloat a[4], result[4];
               fetch_vector1( ctx, &inst->SrcReg[0], machine, program, a );
               result[0] = result[1] = result[2] = result[3] = INV_SQRTF(a[0]);
               store_vector4( inst, machine, result );
#if DEBUG_FRAG
               printf("RSQ %g = 1/sqrt(%g)\n", result[0], a[0]);
#endif
            }
            break;
         case FP_OPCODE_SCS: /* sine and cos */
            {
               GLfloat a[4], result[4];
               fetch_vector1( ctx, &inst->SrcReg[0], machine, program, a );
               result[0] = (GLfloat)cos(a[0]);
               result[1] = (GLfloat)sin(a[0]);
               result[2] = 0.0;  /* undefined! */
               result[3] = 0.0;  /* undefined! */
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_SEQ: /* set on equal */
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               result[0] = (a[0] == b[0]) ? 1.0F : 0.0F;
               result[1] = (a[1] == b[1]) ? 1.0F : 0.0F;
               result[2] = (a[2] == b[2]) ? 1.0F : 0.0F;
               result[3] = (a[3] == b[3]) ? 1.0F : 0.0F;
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_SFL: /* set false, operands ignored */
            {
               static const GLfloat result[4] = { 0.0F, 0.0F, 0.0F, 0.0F };
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_SGE: /* set on greater or equal */
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               result[0] = (a[0] >= b[0]) ? 1.0F : 0.0F;
               result[1] = (a[1] >= b[1]) ? 1.0F : 0.0F;
               result[2] = (a[2] >= b[2]) ? 1.0F : 0.0F;
               result[3] = (a[3] >= b[3]) ? 1.0F : 0.0F;
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_SGT: /* set on greater */
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               result[0] = (a[0] > b[0]) ? 1.0F : 0.0F;
               result[1] = (a[1] > b[1]) ? 1.0F : 0.0F;
               result[2] = (a[2] > b[2]) ? 1.0F : 0.0F;
               result[3] = (a[3] > b[3]) ? 1.0F : 0.0F;
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_SIN:
            {
               GLfloat a[4], result[4];
               fetch_vector1( ctx, &inst->SrcReg[0], machine, program, a );
               result[0] = result[1] = result[2] = 
		       result[3] = (GLfloat)_mesa_sin(a[0]);
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_SLE: /* set on less or equal */
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               result[0] = (a[0] <= b[0]) ? 1.0F : 0.0F;
               result[1] = (a[1] <= b[1]) ? 1.0F : 0.0F;
               result[2] = (a[2] <= b[2]) ? 1.0F : 0.0F;
               result[3] = (a[3] <= b[3]) ? 1.0F : 0.0F;
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_SLT: /* set on less */
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );
               result[0] = (a[0] < b[0]) ? 1.0F : 0.0F;
               result[1] = (a[1] < b[1]) ? 1.0F : 0.0F;
               result[2] = (a[2] < b[2]) ? 1.0F : 0.0F;
               result[3] = (a[3] < b[3]) ? 1.0F : 0.0F;
               store_vector4( inst, machine, result );
            }
            break;
         case FP_OPCODE_SNE: /* set on not equal */
            {
               GLfloat a[4], b[4], result[4];
               fetch_vector4( ctx, &inst->SrcReg[0], machine, program, a );
               fetch_vector4( ctx, &inst->SrcReg[1], machine, program, b );