From c264fdbc073a0dfc393f53a8be880f535fd4b988 Mon Sep 17 00:00:00 2001 From: Kenneth Graunke Date: Mon, 20 Jun 2016 11:20:51 -0700 Subject: glsl: Split arrays even in the presence of whole-array copies. Previously, we failed to split constant arrays. Code such as int[2] numbers = int[](1, 2); would generates a whole-array assignment: (assign () (var_ref numbers) (constant (array int 4) (constant int 1) (constant int 2))) opt_array_splitting generally tried to visit ir_dereference_array nodes, and avoid recursing into the inner ir_dereference_variable. So if it ever saw a ir_dereference_variable, it assumed this was a whole-array read and bailed. However, in the above case, there's no array deref, and we can totally handle it - we just have to "unroll" the assignment, creating assignments for each element. This was mitigated by the fact that we constant propagate whole arrays, so a dereference of a single component would usually get the desired single value anyway. However, I plan to stop doing that shortly; early experiments with disabling constant propagation of arrays revealed this shortcoming. This patch causes some arrays in Gl32GSCloth's geometry shaders to be split, which allows other optimizations to eliminate unused GS inputs. The VS then doesn't have to write them, which eliminates the entire VS (5 -> 2 instructions). It still renders correctly. No other change in shader-db. v2: Drop !AOA check and improve a comment (feedback from Tim Arceri). Cc: mesa-stable@lists.freedesktop.org Signed-off-by: Kenneth Graunke Reviewed-by: Timothy Arceri --- src/compiler/glsl/opt_array_splitting.cpp | 55 +++++++++++++++++++++++++++++++ 1 file changed, 55 insertions(+) (limited to 'src/compiler') diff --git a/src/compiler/glsl/opt_array_splitting.cpp b/src/compiler/glsl/opt_array_splitting.cpp index a294da56616..e3073b022ca 100644 --- a/src/compiler/glsl/opt_array_splitting.cpp +++ b/src/compiler/glsl/opt_array_splitting.cpp @@ -93,6 +93,7 @@ public: { this->mem_ctx = ralloc_context(NULL); this->variable_list.make_empty(); + this->in_whole_array_copy = false; } ~ir_array_reference_visitor(void) @@ -104,6 +105,8 @@ public: virtual ir_visitor_status visit(ir_variable *); virtual ir_visitor_status visit(ir_dereference_variable *); + virtual ir_visitor_status visit_enter(ir_assignment *); + virtual ir_visitor_status visit_leave(ir_assignment *); virtual ir_visitor_status visit_enter(ir_dereference_array *); virtual ir_visitor_status visit_enter(ir_function_signature *); @@ -113,6 +116,8 @@ public: exec_list variable_list; void *mem_ctx; + + bool in_whole_array_copy; }; } /* namespace */ @@ -157,11 +162,34 @@ ir_array_reference_visitor::visit(ir_variable *ir) return visit_continue; } +ir_visitor_status +ir_array_reference_visitor::visit_enter(ir_assignment *ir) +{ + in_whole_array_copy = + ir->lhs->type->is_array() && ir->whole_variable_written(); + + return visit_continue; +} + +ir_visitor_status +ir_array_reference_visitor::visit_leave(ir_assignment *ir) +{ + in_whole_array_copy = false; + + return visit_continue; +} + ir_visitor_status ir_array_reference_visitor::visit(ir_dereference_variable *ir) { variable_entry *entry = this->get_variable_entry(ir->var); + /* Allow whole-array assignments on the LHS. We can split those + * by "unrolling" the assignment into component-wise assignments. + */ + if (in_assignee && in_whole_array_copy) + return visit_continue; + /* If we made it to here without seeing an ir_dereference_array, * then the dereference of this array didn't have a constant index * (see the visit_continue_with_parent below), so we can't split @@ -350,6 +378,33 @@ ir_array_splitting_visitor::visit_leave(ir_assignment *ir) */ ir_rvalue *lhs = ir->lhs; + /* "Unroll" any whole array assignments, creating assignments for + * each array element. Then, do splitting on each new assignment. + */ + if (lhs->type->is_array() && ir->whole_variable_written() && + get_splitting_entry(ir->whole_variable_written())) { + void *mem_ctx = ralloc_parent(ir); + + for (unsigned i = 0; i < lhs->type->length; i++) { + ir_rvalue *lhs_i = + new(mem_ctx) ir_dereference_array(ir->lhs->clone(mem_ctx, NULL), + new(mem_ctx) ir_constant(i)); + ir_rvalue *rhs_i = + new(mem_ctx) ir_dereference_array(ir->rhs->clone(mem_ctx, NULL), + new(mem_ctx) ir_constant(i)); + ir_rvalue *condition_i = + ir->condition ? ir->condition->clone(mem_ctx, NULL) : NULL; + + ir_assignment *assign_i = + new(mem_ctx) ir_assignment(lhs_i, rhs_i, condition_i); + + ir->insert_before(assign_i); + assign_i->accept(this); + } + ir->remove(); + return visit_continue; + } + handle_rvalue(&lhs); ir->lhs = lhs->as_dereference(); -- cgit v1.2.3