class ir_to_mesa_instruction : public exec_node {
public:
- /* Callers of this talloc-based new need not call delete. It's
- * easier to just talloc_free 'ctx' (or any of its ancestors). */
+ /* Callers of this ralloc-based new need not call delete. It's
+ * easier to just ralloc_free 'ctx' (or any of its ancestors). */
static void* operator new(size_t size, void *ctx)
{
void *node;
- node = talloc_zero_size(ctx, size);
+ node = rzalloc_size(ctx, size);
assert(node != NULL);
return node;
{
va_list args;
va_start(args, fmt);
- prog->InfoLog = talloc_vasprintf_append(prog->InfoLog, fmt, args);
+ ralloc_vasprintf_append(&prog->InfoLog, fmt, args);
va_end(args);
prog->LinkStatus = GL_FALSE;
this->result, src_reg_for_float(element_size));
}
- src_reg.reladdr = talloc(mem_ctx, ir_to_mesa_src_reg);
+ src_reg.reladdr = ralloc(mem_ctx, ir_to_mesa_src_reg);
memcpy(src_reg.reladdr, &index_reg, sizeof(index_reg));
}
return entry;
}
- entry = talloc(mem_ctx, function_entry);
+ entry = ralloc(mem_ctx, function_entry);
entry->sig = sig;
entry->sig_id = this->next_signature_id++;
entry->bgn_inst = NULL;
next_temp = 1;
next_signature_id = 1;
current_function = NULL;
- mem_ctx = talloc_new(NULL);
+ mem_ctx = ralloc_context(NULL);
}
ir_to_mesa_visitor::~ir_to_mesa_visitor()
{
- talloc_free(mem_ctx);
+ ralloc_free(mem_ctx);
}
static struct prog_src_register
}
}
- if_stack = talloc_zero_array(v->mem_ctx, int, if_count);
- loop_stack = talloc_zero_array(v->mem_ctx, int, loop_count);
+ if_stack = rzalloc_array(v->mem_ctx, int, if_count);
+ loop_stack = rzalloc_array(v->mem_ctx, int, loop_count);
for (i = 0; i < num_instructions; i++) {
switch (mesa_instructions[i].Opcode) {
unsigned int next_sampler = 0, num_uniforms = 0;
struct uniform_sort *sorted_uniforms;
- sorted_uniforms = talloc_array(NULL, struct uniform_sort,
+ sorted_uniforms = ralloc_array(NULL, struct uniform_sort,
shader_program->Uniforms->NumUniforms);
for (i = 0; i < shader_program->Uniforms->NumUniforms; i++) {
}
}
- talloc_free(sorted_uniforms);
+ ralloc_free(sorted_uniforms);
}
static void
for (unsigned int i = 0; i < type->length; i++) {
const glsl_type *field_type = type->fields.structure[i].type;
- const char *field_name = talloc_asprintf(mem_ctx, "%s.%s", name,
+ const char *field_name = ralloc_asprintf(mem_ctx, "%s.%s", name,
type->fields.structure[i].name);
set_uniform_initializer(ctx, mem_ctx, shader_program, field_name,
field_type, field_constant);
void *values;
if (element_type->base_type == GLSL_TYPE_BOOL) {
- int *conv = talloc_array(mem_ctx, int, element_type->components());
+ int *conv = ralloc_array(mem_ctx, int, element_type->components());
for (unsigned int j = 0; j < element_type->components(); j++) {
conv[j] = element->value.b[j];
}
continue;
if (!mem_ctx)
- mem_ctx = talloc_new(NULL);
+ mem_ctx = ralloc_context(NULL);
set_uniform_initializer(ctx, mem_ctx, shader_program, var->name,
var->type, var->constant_value);
}
}
- talloc_free(mem_ctx);
+ ralloc_free(mem_ctx);
}
/*
void
ir_to_mesa_visitor::copy_propagate(void)
{
- ir_to_mesa_instruction **acp = talloc_zero_array(mem_ctx,
+ ir_to_mesa_instruction **acp = rzalloc_array(mem_ctx,
ir_to_mesa_instruction *,
this->next_temp * 4);
+ int *acp_level = rzalloc_array(mem_ctx, int, this->next_temp * 4);
+ int level = 0;
foreach_iter(exec_list_iterator, iter, this->instructions) {
ir_to_mesa_instruction *inst = (ir_to_mesa_instruction *)iter.get();
break;
}
+ assert(acp_level[acp_base + src_chan] <= level);
+
if (!first) {
first = copy_chan;
} else {
switch (inst->op) {
case OPCODE_BGNLOOP:
case OPCODE_ENDLOOP:
- case OPCODE_ELSE:
- case OPCODE_ENDIF:
/* End of a basic block, clear the ACP entirely. */
memset(acp, 0, sizeof(*acp) * this->next_temp * 4);
break;
+ case OPCODE_IF:
+ ++level;
+ break;
+
+ case OPCODE_ENDIF:
+ case OPCODE_ELSE:
+ /* Clear all channels written inside the block from the ACP, but
+ * leaving those that were not touched.
+ */
+ for (int r = 0; r < this->next_temp; r++) {
+ for (int c = 0; c < 4; c++) {
+ if (!acp[4 * r + c])
+ continue;
+
+ if (acp_level[4 * r + c] >= level)
+ acp[4 * r + c] = NULL;
+ }
+ }
+ if (inst->op == OPCODE_ENDIF)
+ --level;
+ break;
+
default:
/* Continuing the block, clear any written channels from
* the ACP.
*/
- if (inst->dst_reg.file == PROGRAM_TEMPORARY) {
- if (inst->dst_reg.reladdr) {
- memset(acp, 0, sizeof(*acp) * this->next_temp * 4);
- } else {
- for (int i = 0; i < 4; i++) {
- if (inst->dst_reg.writemask & (1 << i)) {
- acp[4 * inst->dst_reg.index + i] = NULL;
+ if (inst->dst_reg.file == PROGRAM_TEMPORARY && inst->dst_reg.reladdr) {
+ /* Any temporary might be written, so no copy propagation
+ * across this instruction.
+ */
+ memset(acp, 0, sizeof(*acp) * this->next_temp * 4);
+ } else if (inst->dst_reg.file == PROGRAM_OUTPUT &&
+ inst->dst_reg.reladdr) {
+ /* Any output might be written, so no copy propagation
+ * from outputs across this instruction.
+ */
+ for (int r = 0; r < this->next_temp; r++) {
+ for (int c = 0; c < 4; c++) {
+ if (!acp[4 * r + c])
+ continue;
+
+ if (acp[4 * r + c]->src_reg[0].file == PROGRAM_OUTPUT)
+ acp[4 * r + c] = NULL;
+ }
+ }
+ } else if (inst->dst_reg.file == PROGRAM_TEMPORARY ||
+ inst->dst_reg.file == PROGRAM_OUTPUT) {
+ /* Clear where it's used as dst. */
+ if (inst->dst_reg.file == PROGRAM_TEMPORARY) {
+ for (int c = 0; c < 4; c++) {
+ if (inst->dst_reg.writemask & (1 << c)) {
+ acp[4 * inst->dst_reg.index + c] = NULL;
+ }
+ }
+ }
+
+ /* Clear where it's used as src. */
+ for (int r = 0; r < this->next_temp; r++) {
+ for (int c = 0; c < 4; c++) {
+ if (!acp[4 * r + c])
+ continue;
+
+ int src_chan = GET_SWZ(acp[4 * r + c]->src_reg[0].swizzle, c);
+
+ if (acp[4 * r + c]->src_reg[0].file == inst->dst_reg.file &&
+ acp[4 * r + c]->src_reg[0].index == inst->dst_reg.index &&
+ inst->dst_reg.writemask & (1 << src_chan))
+ {
+ acp[4 * r + c] = NULL;
}
}
}
for (int i = 0; i < 4; i++) {
if (inst->dst_reg.writemask & (1 << i)) {
acp[4 * inst->dst_reg.index + i] = inst;
+ acp_level[4 * inst->dst_reg.index + i] = level;
}
}
}
}
- talloc_free(acp);
+ ralloc_free(acp_level);
+ ralloc_free(acp);
}
mesa_instructions =
(struct prog_instruction *)calloc(num_instructions,
sizeof(*mesa_instructions));
- mesa_instruction_annotation = talloc_array(v.mem_ctx, ir_instruction *,
+ mesa_instruction_annotation = ralloc_array(v.mem_ctx, ir_instruction *,
num_instructions);
v.copy_propagate();
_mesa_glsl_lexer_dtor(state);
}
- talloc_free(shader->ir);
+ ralloc_free(shader->ir);
shader->ir = new(shader) exec_list;
if (!state->error && !state->translation_unit.is_empty())
_mesa_ast_to_hir(shader->ir, state);
/* Retain any live IR, but trash the rest. */
reparent_ir(shader->ir, shader->ir);
- talloc_free(state);
+ ralloc_free(state);
if (shader->CompileStatus) {
if (!ctx->Driver.CompileShader(ctx, shader))