void
vec4_tes_visitor::nir_emit_intrinsic(nir_intrinsic_instr *instr)
{
const struct brw_tes_prog_data *tes_prog_data =
(const struct brw_tes_prog_data *) prog_data;
switch (instr->intrinsic) {
case nir_intrinsic_load_tess_coord:
/* gl_TessCoord is part of the payload in g1 channels 0-2 and 4-6. */
emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F),
src_reg(brw_vec8_grf(1, 0))));
break;
case nir_intrinsic_load_tess_level_outer:
if (tes_prog_data->domain == BRW_TESS_DOMAIN_ISOLINE) {
emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F),
swizzle(src_reg(ATTR, 1, glsl_type::vec4_type),
BRW_SWIZZLE_ZWZW)));
} else {
emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F),
swizzle(src_reg(ATTR, 1, glsl_type::vec4_type),
BRW_SWIZZLE_WZYX)));
}
break;
case nir_intrinsic_load_tess_level_inner:
if (tes_prog_data->domain == BRW_TESS_DOMAIN_QUAD) {
emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F),
swizzle(src_reg(ATTR, 0, glsl_type::vec4_type),
BRW_SWIZZLE_WZYX)));
} else {
emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F),
src_reg(ATTR, 1, glsl_type::float_type)));
}
break;
case nir_intrinsic_load_primitive_id:
emit(TES_OPCODE_GET_PRIMITIVE_ID,
get_nir_dest(instr->dest, BRW_REGISTER_TYPE_UD));
break;
case nir_intrinsic_load_input:
case nir_intrinsic_load_per_vertex_input: {
src_reg indirect_offset = get_indirect_offset(instr);
unsigned imm_offset = instr->const_index[0];
src_reg header = input_read_header;
bool is_64bit = nir_dest_bit_size(instr->dest) == 64;
unsigned first_component = nir_intrinsic_component(instr);
if (is_64bit)
first_component /= 2;
if (indirect_offset.file != BAD_FILE) {
header = src_reg(this, glsl_type::uvec4_type);
emit(TES_OPCODE_ADD_INDIRECT_URB_OFFSET, dst_reg(header),
input_read_header, indirect_offset);
} else {
/* Arbitrarily only push up to 24 vec4 slots worth of data,
* which is 12 registers (since each holds 2 vec4 slots).
*/
const unsigned max_push_slots = 24;
if (imm_offset < max_push_slots) {
const glsl_type *src_glsl_type =
is_64bit ? glsl_type::dvec4_type : glsl_type::ivec4_type;
src_reg src = src_reg(ATTR, imm_offset, src_glsl_type);
src.swizzle = BRW_SWZ_COMP_INPUT(first_component);
const brw_reg_type dst_reg_type =
is_64bit ? BRW_REGISTER_TYPE_DF : BRW_REGISTER_TYPE_D;
emit(MOV(get_nir_dest(instr->dest, dst_reg_type), src));
prog_data->urb_read_length =
MAX2(prog_data->urb_read_length,
DIV_ROUND_UP(imm_offset + (is_64bit ? 2 : 1), 2));
break;
}
}
if (!is_64bit) {
dst_reg temp(this, glsl_type::ivec4_type);
vec4_instruction *read =
emit(VEC4_OPCODE_URB_READ, temp, src_reg(header));
read->offset = imm_offset;
read->urb_write_flags = BRW_URB_WRITE_PER_SLOT_OFFSET;
src_reg src = src_reg(temp);
src.swizzle = BRW_SWZ_COMP_INPUT(first_component);
/* Copy to target. We might end up with some funky writemasks landing
* in here, but we really don't want them in the above pseudo-ops.
*/
dst_reg dst = get_nir_dest(instr->dest, BRW_REGISTER_TYPE_D);
dst.writemask = brw_writemask_for_size(instr->num_components);
emit(MOV(dst, src));
} else {
/* For 64-bit we need to load twice as many 32-bit components, and for
* dvec3/4 we need to emit 2 URB Read messages
*/
dst_reg temp(this, glsl_type::dvec4_type);
dst_reg temp_d = retype(temp, BRW_REGISTER_TYPE_D);
vec4_instruction *read =
emit(VEC4_OPCODE_URB_READ, temp_d, src_reg(header));
read->offset = imm_offset;
read->urb_write_flags = BRW_URB_WRITE_PER_SLOT_OFFSET;
if (instr->num_components > 2) {
read = emit(VEC4_OPCODE_URB_READ, byte_offset(temp_d, REG_SIZE),
src_reg(header));
read->offset = imm_offset + 1;
read->urb_write_flags = BRW_URB_WRITE_PER_SLOT_OFFSET;
}
src_reg temp_as_src = src_reg(temp);
temp_as_src.swizzle = BRW_SWZ_COMP_INPUT(first_component);
dst_reg shuffled(this, glsl_type::dvec4_type);
shuffle_64bit_data(shuffled, temp_as_src, false);
dst_reg dst = get_nir_dest(instr->dest, BRW_REGISTER_TYPE_DF);
dst.writemask = brw_writemask_for_size(instr->num_components);
emit(MOV(dst, src_reg(shuffled)));
}
break;
}
default:
vec4_visitor::nir_emit_intrinsic(instr);
}
}
static void
validate_alu_instr(nir_alu_instr *instr, validate_state *state)
{
validate_assert(state, instr->op < nir_num_opcodes);
unsigned instr_bit_size = 0;
for (unsigned i = 0; i < nir_op_infos[instr->op].num_inputs; i++) {
nir_alu_type src_type = nir_op_infos[instr->op].input_types[i];
unsigned src_bit_size = nir_src_bit_size(instr->src[i].src);
if (nir_alu_type_get_type_size(src_type)) {
validate_assert(state, src_bit_size == nir_alu_type_get_type_size(src_type));
} else if (instr_bit_size) {
validate_assert(state, src_bit_size == instr_bit_size);
} else {
instr_bit_size = src_bit_size;
}
if (nir_alu_type_get_base_type(src_type) == nir_type_float) {
/* 8-bit float isn't a thing */
validate_assert(state, src_bit_size == 16 || src_bit_size == 32 ||
src_bit_size == 64);
}
validate_alu_src(instr, i, state);
}
nir_alu_type dest_type = nir_op_infos[instr->op].output_type;
unsigned dest_bit_size = nir_dest_bit_size(instr->dest.dest);
if (nir_alu_type_get_type_size(dest_type)) {
validate_assert(state, dest_bit_size == nir_alu_type_get_type_size(dest_type));
} else if (instr_bit_size) {
validate_assert(state, dest_bit_size == instr_bit_size);
} else {
/* The only unsized thing is the destination so it's vacuously valid */
}
if (nir_alu_type_get_base_type(dest_type) == nir_type_float) {
/* 8-bit float isn't a thing */
validate_assert(state, dest_bit_size == 16 || dest_bit_size == 32 ||
dest_bit_size == 64);
}
validate_alu_dest(instr, state);
}
