src/mesa/drivers/dri/i965/brw_fs_peephole_predicated_break.cpp

   1 /*
   2  * Copyright © 2013 Intel Corporation
   3  *
   4  * Permission is hereby granted, free of charge, to any person obtaining a
   5  * copy of this software and associated documentation files (the "Software"),
   6  * to deal in the Software without restriction, including without limitation
   7  * the rights to use, copy, modify, merge, publish, distribute, sublicense,
   8  * and/or sell copies of the Software, and to permit persons to whom the
   9  * Software is furnished to do so, subject to the following conditions:
  10  *
  11  * The above copyright notice and this permission notice (including the next
  12  * paragraph) shall be included in all copies or substantial portions of the
  13  * Software.
  14  *
  15  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  16  * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  17  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
  18  * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  19  * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
  20  * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
  21  * IN THE SOFTWARE.
  22  */
  23
  24 #include "brw_fs.h"
  25 #include "brw_cfg.h"
  26
  27 /** @file brw_fs_peephole_predicated_break.cpp
  28  *
  29  * Loops are often structured as
  30  *
  31  * loop:
  32  *    CMP.f0
  33  *    (+f0) IF
  34  *    BREAK
  35  *    ENDIF
  36  *    ...
  37  *    WHILE loop
  38  *
  39  * This peephole pass removes the IF and ENDIF instructions and predicates the
  40  * BREAK, dropping two instructions from the loop body.
  41  */
  42
  43 bool
  44 fs_visitor::opt_peephole_predicated_break()
  45 {
  46    bool progress = false;
  47
  48    foreach_block (block, cfg) {
  49       if (block->start_ip != block->end_ip)
  50          continue;
  51
  52       /* BREAK and CONTINUE instructions, by definition, can only be found at
  53        * the ends of basic blocks.
  54        */
  55       fs_inst *jump_inst = (fs_inst *)block->end;
  56       if (jump_inst->opcode != BRW_OPCODE_BREAK &&
  57           jump_inst->opcode != BRW_OPCODE_CONTINUE)
  58          continue;
  59
  60       fs_inst *if_inst = (fs_inst *)((bblock_t *)block->link.prev)->end;
  61       if (if_inst->opcode != BRW_OPCODE_IF)
  62          continue;
  63
  64       fs_inst *endif_inst = (fs_inst *)((bblock_t *)block->link.next)->start;
  65       if (endif_inst->opcode != BRW_OPCODE_ENDIF)
  66          continue;
  67
  68       bblock_t *jump_block = block;
  69       bblock_t *if_block = (bblock_t *)jump_block->link.prev;
  70       bblock_t *endif_block = (bblock_t *)jump_block->link.next;
  71
  72       /* For Sandybridge with IF with embedded comparison we need to emit an
  73        * instruction to set the flag register.
  74        */
  75       if (brw->gen == 6 && if_inst->conditional_mod) {
  76          fs_inst *cmp_inst = CMP(reg_null_d, if_inst->src[0], if_inst->src[1],
  77                                  if_inst->conditional_mod);
  78          if_inst->insert_before(if_block, cmp_inst);
  79          jump_inst->predicate = BRW_PREDICATE_NORMAL;
  80       } else {
  81          jump_inst->predicate = if_inst->predicate;
  82          jump_inst->predicate_inverse = if_inst->predicate_inverse;
  83       }
  84
  85       bblock_t *earlier_block = if_block;
  86       if (if_block->start_ip == if_block->end_ip) {
  87          earlier_block = (bblock_t *)if_block->link.prev;
  88       }
  89
  90       if_inst->remove(if_block);
  91
  92       bblock_t *later_block = endif_block;
  93       if (endif_block->start_ip == endif_block->end_ip) {
  94          later_block = (bblock_t *)endif_block->link.next;
  95       }
  96       endif_inst->remove(endif_block);
  97
  98       earlier_block->children.make_empty();
  99       later_block->parents.make_empty();
 100
 101       earlier_block->add_successor(cfg->mem_ctx, jump_block);
 102       jump_block->add_successor(cfg->mem_ctx, later_block);
 103
 104       if (earlier_block->can_combine_with(jump_block)) {
 105          earlier_block->combine_with(jump_block);
 106
 107          block = earlier_block;
 108       }
 109
 110       progress = true;
 111    }
 112
 113    if (progress)
 114       invalidate_live_intervals();
 115
 116    return progress;
 117 }