#include "st_program.h"
#include "st_mesa_to_tgsi.h"
#include "st_format.h"
-#include "st_glsl_types.h"
#include "st_nir.h"
#include "st_shader_cache.h"
return swizzle;
}
+static unsigned is_precise(const ir_variable *ir)
+{
+ if (!ir)
+ return 0;
+ return ir->data.precise || ir->data.invariant;
+}
+
/**
* This struct is a corresponding struct to TGSI ureg_src.
*/
st_dst_reg dst[2];
st_src_reg src[4];
- st_src_reg resource; /**< sampler or buffer register */
+ st_src_reg resource; /**< sampler, image or buffer register */
st_src_reg *tex_offsets;
/** Pointer to the ir source this tree came from for debugging */
ir_instruction *ir;
unsigned op:8; /**< TGSI opcode */
+ unsigned precise:1;
unsigned saturate:1;
unsigned is_64bit_expanded:1;
unsigned sampler_base:5;
}
struct rename_reg_pair {
- bool valid;
+ int old_reg;
int new_reg;
};
bool have_fma;
bool use_shared_memory;
bool has_tex_txf_lz;
+ bool precise;
variable_storage *find_variable_storage(ir_variable *var);
void simplify_cmp(void);
- void rename_temp_registers(struct rename_reg_pair *renames);
+ void rename_temp_registers(int num_renames, struct rename_reg_pair *renames);
void get_first_temp_read(int *first_reads);
void get_first_temp_write(int *first_writes);
void get_last_temp_read_first_temp_write(int *last_reads, int *first_writes);
STATIC_ASSERT(TGSI_OPCODE_LAST <= 255);
inst->op = op;
+ inst->precise = this->precise;
inst->info = tgsi_get_opcode_info(op);
inst->dst[0] = dst;
inst->dst[1] = dst1;
static int
attrib_type_size(const struct glsl_type *type, bool is_vs_input)
{
- return st_glsl_attrib_type_size(type, is_vs_input);
+ return type->count_attribute_slots(is_vs_input);
}
static int
type_size(const struct glsl_type *type)
{
- return st_glsl_type_size(type);
+ return type->count_attribute_slots(false);
}
/**
/* Quick peephole: Emit MAD(a, b, c) instead of ADD(MUL(a, b), c)
*/
- if (ir->operation == ir_binop_add) {
+ if (!this->precise && ir->operation == ir_binop_add) {
if (try_emit_mad(ir, 1))
return;
if (try_emit_mad(ir, 0))
case ir_unop_pack_int_2x32:
case ir_unop_unpack_uint_2x32:
case ir_unop_pack_uint_2x32:
+ case ir_unop_unpack_sampler_2x32:
+ case ir_unop_pack_sampler_2x32:
+ case ir_unop_unpack_image_2x32:
+ case ir_unop_pack_image_2x32:
emit_asm(ir, TGSI_OPCODE_MOV, result_dst, op[0]);
break;
case ir_unop_unpack_snorm_4x8:
case ir_unop_unpack_unorm_4x8:
- case ir_unop_unpack_sampler_2x32:
- case ir_unop_pack_sampler_2x32:
- case ir_unop_unpack_image_2x32:
- case ir_unop_pack_image_2x32:
-
case ir_quadop_vector:
case ir_binop_vector_extract:
case ir_triop_vector_insert:
st_dst_reg l;
st_src_reg r;
+ /* all generated instructions need to be flaged as precise */
+ this->precise = is_precise(ir->lhs->variable_referenced());
ir->rhs->accept(this);
r = this->result;
} else {
emit_block_mov(ir, ir->rhs->type, &l, &r, NULL, false);
}
+ this->precise = 0;
}
}
}
+static void
+get_image_qualifiers(ir_dereference *ir, const glsl_type **type,
+ bool *memory_coherent, bool *memory_volatile,
+ bool *memory_restrict, unsigned *image_format)
+{
+
+ switch (ir->ir_type) {
+ case ir_type_dereference_record: {
+ ir_dereference_record *deref_record = ir->as_dereference_record();
+ const glsl_type *struct_type = deref_record->record->type;
+
+ for (unsigned i = 0; i < struct_type->length; i++) {
+ if (!strcmp(struct_type->fields.structure[i].name,
+ deref_record->field)) {
+ *type = struct_type->fields.structure[i].type->without_array();
+ *memory_coherent =
+ struct_type->fields.structure[i].memory_coherent;
+ *memory_volatile =
+ struct_type->fields.structure[i].memory_volatile;
+ *memory_restrict =
+ struct_type->fields.structure[i].memory_restrict;
+ *image_format =
+ struct_type->fields.structure[i].image_format;
+ break;
+ }
+ }
+ break;
+ }
+
+ case ir_type_dereference_array: {
+ ir_dereference_array *deref_arr = ir->as_dereference_array();
+ get_image_qualifiers((ir_dereference *)deref_arr->array, type,
+ memory_coherent, memory_volatile, memory_restrict,
+ image_format);
+ break;
+ }
+
+ case ir_type_dereference_variable: {
+ ir_variable *var = ir->variable_referenced();
+
+ *type = var->type->without_array();
+ *memory_coherent = var->data.memory_coherent;
+ *memory_volatile = var->data.memory_volatile;
+ *memory_restrict = var->data.memory_restrict;
+ *image_format = var->data.image_format;
+ break;
+ }
+
+ default:
+ break;
+ }
+}
+
void
glsl_to_tgsi_visitor::visit_image_intrinsic(ir_call *ir)
{
ir_dereference *img = (ir_dereference *)param;
const ir_variable *imgvar = img->variable_referenced();
- const glsl_type *type = imgvar->type->without_array();
unsigned sampler_array_size = 1, sampler_base = 0;
+ bool memory_coherent = false, memory_volatile = false, memory_restrict = false;
+ unsigned image_format = 0;
+ const glsl_type *type = NULL;
+
+ get_image_qualifiers(img, &type, &memory_coherent, &memory_volatile,
+ &memory_restrict, &image_format);
st_src_reg reladdr;
st_src_reg image(PROGRAM_IMAGE, 0, GLSL_TYPE_UINT);
uint16_t index = 0;
get_deref_offsets(img, &sampler_array_size, &sampler_base,
- &index, &reladdr, true);
+ &index, &reladdr, !imgvar->contains_bindless());
image.index = index;
if (reladdr.file != PROGRAM_UNDEFINED) {
inst->dst[0].writemask = WRITEMASK_XYZW;
}
- inst->resource = image;
- inst->sampler_array_size = sampler_array_size;
- inst->sampler_base = sampler_base;
+ if (imgvar->contains_bindless()) {
+ img->accept(this);
+ inst->resource = this->result;
+ inst->resource.swizzle = MAKE_SWIZZLE4(SWIZZLE_X, SWIZZLE_Y,
+ SWIZZLE_X, SWIZZLE_Y);
+ } else {
+ inst->resource = image;
+ inst->sampler_array_size = sampler_array_size;
+ inst->sampler_base = sampler_base;
+ }
inst->tex_target = type->sampler_index();
inst->image_format = st_mesa_format_to_pipe_format(st_context(ctx),
- _mesa_get_shader_image_format(imgvar->data.image_format));
+ _mesa_get_shader_image_format(image_format));
- if (imgvar->data.memory_coherent)
+ if (memory_coherent)
inst->buffer_access |= TGSI_MEMORY_COHERENT;
- if (imgvar->data.memory_restrict)
+ if (memory_restrict)
inst->buffer_access |= TGSI_MEMORY_RESTRICT;
- if (imgvar->data.memory_volatile)
+ if (memory_volatile)
inst->buffer_access |= TGSI_MEMORY_VOLATILE;
}
ir->return_deref->accept(this);
st_dst_reg dst = st_dst_reg(this->result);
+ dst.writemask = u_bit_consecutive(0, ir->return_deref->var->type->vector_elements);
+
st_src_reg src[4] = { undef_src, undef_src, undef_src, undef_src };
unsigned num_src = 0;
foreach_in_list(ir_rvalue, param, &ir->actual_parameters) {
mem_ctx = ralloc_context(NULL);
ctx = NULL;
prog = NULL;
+ precise = 0;
shader_program = NULL;
shader = NULL;
options = NULL;
/* Replaces all references to a temporary register index with another index. */
void
-glsl_to_tgsi_visitor::rename_temp_registers(struct rename_reg_pair *renames)
+glsl_to_tgsi_visitor::rename_temp_registers(int num_renames, struct rename_reg_pair *renames)
{
foreach_in_list(glsl_to_tgsi_instruction, inst, &this->instructions) {
unsigned j;
+ int k;
for (j = 0; j < num_inst_src_regs(inst); j++) {
- if (inst->src[j].file == PROGRAM_TEMPORARY) {
- int old_idx = inst->src[j].index;
- if (renames[old_idx].valid)
- inst->src[j].index = renames[old_idx].new_reg;
- }
+ if (inst->src[j].file == PROGRAM_TEMPORARY)
+ for (k = 0; k < num_renames; k++)
+ if (inst->src[j].index == renames[k].old_reg)
+ inst->src[j].index = renames[k].new_reg;
}
for (j = 0; j < inst->tex_offset_num_offset; j++) {
- if (inst->tex_offsets[j].file == PROGRAM_TEMPORARY) {
- int old_idx = inst->tex_offsets[j].index;
- if (renames[old_idx].valid)
- inst->tex_offsets[j].index = renames[old_idx].new_reg;
- }
+ if (inst->tex_offsets[j].file == PROGRAM_TEMPORARY)
+ for (k = 0; k < num_renames; k++)
+ if (inst->tex_offsets[j].index == renames[k].old_reg)
+ inst->tex_offsets[j].index = renames[k].new_reg;
+ }
+
+ if (inst->resource.file == PROGRAM_TEMPORARY) {
+ for (k = 0; k < num_renames; k++)
+ if (inst->resource.index == renames[k].old_reg)
+ inst->resource.index = renames[k].new_reg;
}
for (j = 0; j < num_inst_dst_regs(inst); j++) {
- if (inst->dst[j].file == PROGRAM_TEMPORARY) {
- int old_idx = inst->dst[j].index;
- if (renames[old_idx].valid)
- inst->dst[j].index = renames[old_idx].new_reg;}
+ if (inst->dst[j].file == PROGRAM_TEMPORARY)
+ for (k = 0; k < num_renames; k++)
+ if (inst->dst[j].index == renames[k].old_reg)
+ inst->dst[j].index = renames[k].new_reg;
}
}
}
}
}
}
+
+ if (inst->resource.file == PROGRAM_TEMPORARY) {
+ int src_chans;
+
+ src_chans = 1 << GET_SWZ(inst->resource.swizzle, 0);
+ src_chans |= 1 << GET_SWZ(inst->resource.swizzle, 1);
+ src_chans |= 1 << GET_SWZ(inst->resource.swizzle, 2);
+ src_chans |= 1 << GET_SWZ(inst->resource.swizzle, 3);
+
+ for (int c = 0; c < 4; c++) {
+ if (src_chans & (1 << c))
+ writes[4 * inst->resource.index + c] = NULL;
+ }
+ }
+
break;
}
int *first_writes = ralloc_array(mem_ctx, int, this->next_temp);
struct rename_reg_pair *renames = rzalloc_array(mem_ctx, struct rename_reg_pair, this->next_temp);
int i, j;
+ int num_renames = 0;
/* Read the indices of the last read and first write to each temp register
* into an array so that we don't have to traverse the instruction list as
* as the register at index j. */
if (first_writes[i] <= first_writes[j] &&
last_reads[i] <= first_writes[j]) {
- renames[j].new_reg = i;
- renames[j].valid = true;
+ renames[num_renames].old_reg = j;
+ renames[num_renames].new_reg = i;
+ num_renames++;
/* Update the first_writes and last_reads arrays with the new
* values for the merged register index, and mark the newly unused
}
}
- rename_temp_registers(renames);
+ rename_temp_registers(num_renames, renames);
ralloc_free(renames);
ralloc_free(last_reads);
ralloc_free(first_writes);
int new_index = 0;
int *first_writes = ralloc_array(mem_ctx, int, this->next_temp);
struct rename_reg_pair *renames = rzalloc_array(mem_ctx, struct rename_reg_pair, this->next_temp);
+ int num_renames = 0;
for (i = 0; i < this->next_temp; i++) {
first_writes[i] = -1;
for (i = 0; i < this->next_temp; i++) {
if (first_writes[i] < 0) continue;
if (i != new_index) {
- renames[i].new_reg = new_index;
- renames[i].valid = true;
+ renames[num_renames].old_reg = i;
+ renames[num_renames].new_reg = new_index;
+ num_renames++;
}
new_index++;
}
- rename_temp_registers(renames);
+ rename_temp_registers(num_renames, renames);
this->next_temp = new_index;
ralloc_free(renames);
ralloc_free(first_writes);
case TGSI_OPCODE_IF:
case TGSI_OPCODE_UIF:
assert(num_dst == 0);
- ureg_insn(ureg, inst->op, NULL, 0, src, num_src);
+ ureg_insn(ureg, inst->op, NULL, 0, src, num_src, inst->precise);
return;
case TGSI_OPCODE_TEX:
} else if (inst->resource.file == PROGRAM_BUFFER) {
src[0] = t->buffers[inst->resource.index];
} else {
- src[0] = t->images[inst->resource.index];
+ if (inst->resource.file == PROGRAM_IMAGE) {
+ src[0] = t->images[inst->resource.index];
+ } else {
+ /* Bindless images. */
+ src[0] = translate_src(t, &inst->resource);
+ }
tex_target = st_translate_texture_target(inst->tex_target, inst->tex_shadow);
}
if (inst->resource.reladdr)
} else if (inst->resource.file == PROGRAM_BUFFER) {
dst[0] = ureg_dst(t->buffers[inst->resource.index]);
} else {
- dst[0] = ureg_dst(t->images[inst->resource.index]);
+ if (inst->resource.file == PROGRAM_IMAGE) {
+ dst[0] = ureg_dst(t->images[inst->resource.index]);
+ } else {
+ /* Bindless images. */
+ dst[0] = ureg_dst(translate_src(t, &inst->resource));
+ }
tex_target = st_translate_texture_target(inst->tex_target, inst->tex_shadow);
}
dst[0] = ureg_writemask(dst[0], inst->dst[0].writemask);
case TGSI_OPCODE_SCS:
dst[0] = ureg_writemask(dst[0], TGSI_WRITEMASK_XY);
- ureg_insn(ureg, inst->op, dst, num_dst, src, num_src);
+ ureg_insn(ureg, inst->op, dst, num_dst, src, num_src, inst->precise);
break;
default:
ureg_insn(ureg,
inst->op,
dst, num_dst,
- src, num_src);
+ src, num_src, inst->precise);
break;
}
}
for (i = program->shader->Stage+1; i <= MESA_SHADER_FRAGMENT; i++) {
if (program->shader_program->_LinkedShaders[i]) {
- unsigned next;
-
- switch (i) {
- case MESA_SHADER_TESS_CTRL:
- next = PIPE_SHADER_TESS_CTRL;
- break;
- case MESA_SHADER_TESS_EVAL:
- next = PIPE_SHADER_TESS_EVAL;
- break;
- case MESA_SHADER_GEOMETRY:
- next = PIPE_SHADER_GEOMETRY;
- break;
- case MESA_SHADER_FRAGMENT:
- next = PIPE_SHADER_FRAGMENT;
- break;
- default:
- assert(0);
- continue;
- }
-
- ureg_set_next_shader_processor(ureg, next);
+ ureg_set_next_shader_processor(
+ ureg, pipe_shader_type_from_mesa((gl_shader_stage)i));
break;
}
}
struct gl_shader_compiler_options *options =
&ctx->Const.ShaderCompilerOptions[shader->Stage];
struct pipe_screen *pscreen = ctx->st->pipe->screen;
- enum pipe_shader_type ptarget = st_shader_stage_to_ptarget(shader->Stage);
+ enum pipe_shader_type ptarget = pipe_shader_type_from_mesa(shader->Stage);
unsigned skip_merge_registers;
validate_ir_tree(shader->ir);
gl_shader_stage stage = shader->Stage;
const struct gl_shader_compiler_options *options =
&ctx->Const.ShaderCompilerOptions[stage];
- enum pipe_shader_type ptarget = st_shader_stage_to_ptarget(stage);
+ enum pipe_shader_type ptarget = pipe_shader_type_from_mesa(stage);
bool have_dround = pscreen->get_shader_param(pscreen, ptarget,
PIPE_SHADER_CAP_TGSI_DROUND_SUPPORTED);
bool have_dfrexp = pscreen->get_shader_param(pscreen, ptarget,
continue;
enum pipe_shader_type ptarget =
- st_shader_stage_to_ptarget(shader->Stage);
+ pipe_shader_type_from_mesa(shader->Stage);
enum pipe_shader_ir preferred_ir = (enum pipe_shader_ir)
pscreen->get_shader_param(pscreen, ptarget,
PIPE_SHADER_CAP_PREFERRED_IR);