- case OPCODE_NOT: /* bitwise NOT */
- {
- GLuint a[4], result[4];
- fetch_vector4ui(&inst->SrcReg[0], machine, a);
- result[0] = ~a[0];
- result[1] = ~a[1];
- result[2] = ~a[2];
- result[3] = ~a[3];
- store_vector4ui(inst, machine, result);
- }
- break;
- case OPCODE_NRM3: /* 3-component normalization */
- {
- GLfloat a[4], result[4];
- GLfloat tmp;
- fetch_vector4(&inst->SrcReg[0], machine, a);
- tmp = a[0] * a[0] + a[1] * a[1] + a[2] * a[2];
- if (tmp != 0.0F)
- tmp = INV_SQRTF(tmp);
- result[0] = tmp * a[0];
- result[1] = tmp * a[1];
- result[2] = tmp * a[2];
- result[3] = 0.0; /* undefined, but prevent valgrind warnings */
- store_vector4(inst, machine, result);
- }
- break;
- case OPCODE_NRM4: /* 4-component normalization */
- {
- GLfloat a[4], result[4];
- GLfloat tmp;
- fetch_vector4(&inst->SrcReg[0], machine, a);
- tmp = a[0] * a[0] + a[1] * a[1] + a[2] * a[2] + a[3] * a[3];
- if (tmp != 0.0F)
- tmp = INV_SQRTF(tmp);
- result[0] = tmp * a[0];
- result[1] = tmp * a[1];
- result[2] = tmp * a[2];
- result[3] = tmp * a[3];
- store_vector4(inst, machine, result);
- }
- break;
- case OPCODE_OR: /* bitwise OR */
- {
- GLuint a[4], b[4], result[4];
- fetch_vector4ui(&inst->SrcReg[0], machine, a);
- fetch_vector4ui(&inst->SrcReg[1], machine, 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_vector4ui(inst, machine, result);
- }
- break;
- case OPCODE_PK2H: /* pack two 16-bit floats in one 32-bit float */
- {
- GLfloat a[4];
- GLuint result[4];
- GLhalfNV hx, hy;
- fetch_vector4(&inst->SrcReg[0], machine, a);
- hx = _mesa_float_to_half(a[0]);
- hy = _mesa_float_to_half(a[1]);
- result[0] =
- result[1] =
- result[2] =
- result[3] = hx | (hy << 16);
- store_vector4ui(inst, machine, result);
- }
- break;
- case OPCODE_PK2US: /* pack two GLushorts into one 32-bit float */
- {
- GLfloat a[4];
- GLuint result[4], usx, usy;
- fetch_vector4(&inst->SrcReg[0], machine, 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);
- result[0] =
- result[1] =
- result[2] =
- result[3] = usx | (usy << 16);
- store_vector4ui(inst, machine, result);
- }
- break;
- case OPCODE_PK4B: /* pack four GLbytes into one 32-bit float */
- {
- GLfloat a[4];
- GLuint result[4], ubx, uby, ubz, ubw;
- fetch_vector4(&inst->SrcReg[0], machine, 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);
- result[0] =
- result[1] =
- result[2] =
- result[3] = ubx | (uby << 8) | (ubz << 16) | (ubw << 24);
- store_vector4ui(inst, machine, result);
- }
- break;
- case OPCODE_PK4UB: /* pack four GLubytes into one 32-bit float */
- {
- GLfloat a[4];
- GLuint result[4], ubx, uby, ubz, ubw;
- fetch_vector4(&inst->SrcReg[0], machine, 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]);
- result[0] =
- result[1] =
- result[2] =
- result[3] = ubx | (uby << 8) | (ubz << 16) | (ubw << 24);
- store_vector4ui(inst, machine, result);
- }
- break;