gcc/tree-ssa-threadupdate.c

   1 /* Thread edges through blocks and update the control flow and SSA graphs.
   2    Copyright (C) 2004 Free Software Foundation, Inc.
   3
   4 This file is part of GCC.
   5
   6 GCC is free software; you can redistribute it and/or modify
   7 it under the terms of the GNU General Public License as published by
   8 the Free Software Foundation; either version 2, or (at your option)
   9 any later version.
  10
  11 GCC is distributed in the hope that it will be useful,
  12 but WITHOUT ANY WARRANTY; without even the implied warranty of
  13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14 GNU General Public License for more details.
  15
  16 You should have received a copy of the GNU General Public License
  17 along with GCC; see the file COPYING.  If not, write to
  18 the Free Software Foundation, 59 Temple Place - Suite 330,
  19 Boston, MA 02111-1307, USA.  */
  20
  21 #include "config.h"
  22 #include "system.h"
  23 #include "coretypes.h"
  24 #include "tm.h"
  25 #include "tree.h"
  26 #include "flags.h"
  27 #include "rtl.h"
  28 #include "tm_p.h"
  29 #include "ggc.h"
  30 #include "basic-block.h"
  31 #include "output.h"
  32 #include "errors.h"
  33 #include "expr.h"
  34 #include "function.h"
  35 #include "diagnostic.h"
  36 #include "tree-flow.h"
  37 #include "tree-dump.h"
  38 #include "tree-pass.h"
  39
  40 /* Given a block B, update the CFG and SSA graph to reflect redirecting
  41    one or more in-edges to B to instead reach the destination of an
  42    out-edge from B while preserving any side effects in B.
  43
  44    i.e., given A->B and B->C, change A->B to be A->C yet still preserve the
  45    side effects of executing B.
  46
  47      1. Make a copy of B (including its outgoing edges and statements).  Call
  48         the copy B'.  Note B' has no incoming edges or PHIs at this time.
  49
  50      2. Remove the control statement at the end of B' and all outgoing edges
  51         except B'->C.
  52
  53      3. Add a new argument to each PHI in C with the same value as the existing
  54         argument associated with edge B->C.  Associate the new PHI arguments
  55         with the edge B'->C.
  56
  57      4. For each PHI in B, find or create a PHI in B' with an identical
  58         PHI_RESULT.  Add an argument to the PHI in B' which has the same
  59         value as the PHI in B associated with the edge A->B.  Associate
  60         the new argument in the PHI in B' with the edge A->B.
  61
  62      5. Change the edge A->B to A->B'.
  63
  64         5a. This automatically deletes any PHI arguments associated with the
  65             edge A->B in B.
  66
  67         5b. This automatically associates each new argument added in step 4
  68             with the edge A->B'.
  69
  70      6. Repeat for other incoming edges into B.
  71
  72      7. Put the duplicated resources in B and all the B' blocks into SSA form.
  73
  74    Note that block duplication can be minimized by first collecting the
  75    the set of unique destination blocks that the incoming edges should
  76    be threaded to.  Block duplication can be further minimized by using
  77    B instead of creating B' for one destination if all edges into B are
  78    going to be threaded to a successor of B.  */
  79
  80
  81 /* Main data structure recording information regarding B's duplicate
  82    blocks.  */
  83
  84 struct redirection_data
  85 {
  86   /* A duplicate of B with the trailing control statement removed and which
  87      targets a single successor of B.  */
  88   basic_block dup_block;
  89
  90   /* An outgoing edge from B.  DUP_BLOCK will have OUTGOING_EDGE->dest as
  91      its single successor.  */
  92   edge outgoing_edge;
  93 };
  94
  95 /* Main data structure to hold information for duplicates of BB.  */
  96 static varray_type redirection_data;
  97
  98 /* For each PHI node in BB, find or create a PHI node in NEW_BB for the
  99    same PHI_RESULT.  Add an argument to the PHI node in NEW_BB which
 100    corresponds to the same PHI argument associated with edge E in BB.  */
 101
 102 static void
 103 copy_phis_to_block (basic_block new_bb, basic_block bb, edge e)
 104 {
 105   tree phi, arg;
 106
 107   /* Walk over every PHI in BB.  */
 108   for (phi = phi_nodes (bb); phi; phi = PHI_CHAIN (phi))
 109     {
 110       tree new_phi;
 111
 112       /* First try to find a PHI node in NEW_BB which has the same
 113          PHI_RESULT as the PHI from BB we are currently processing.  */
 114       for (new_phi = phi_nodes (new_bb); new_phi;
 115            new_phi = PHI_CHAIN (new_phi))
 116         if (PHI_RESULT (new_phi) == PHI_RESULT (phi))
 117           break;
 118
 119       /* If we did not find a suitable PHI in NEW_BB, create one.  */
 120       if (!new_phi)
 121         new_phi = create_phi_node (PHI_RESULT (phi), new_bb);
 122
 123       /* Extract the argument corresponding to E from the current PHI
 124          node in BB.  */
 125       arg = PHI_ARG_DEF_TREE (phi, phi_arg_from_edge (phi, e));
 126
 127       /* Now add that same argument to the new PHI node in block NEW_BB.  */
 128       add_phi_arg (&new_phi, arg, e);
 129     }
 130 }
 131
 132 /* Remove the last statement in block BB which must be a COND_EXPR or
 133    SWITCH_EXPR.  Also remove all outgoing edges except the edge which
 134    reaches DEST_BB.
 135
 136    This is only used by jump threading which knows the last statement in
 137    BB should be a COND_EXPR or SWITCH_EXPR.  If the block ends with any other
 138    statement, then we abort.  */
 139
 140 static void
 141 remove_last_stmt_and_useless_edges (basic_block bb, basic_block dest_bb)
 142 {
 143   block_stmt_iterator bsi;
 144   edge e, next;
 145
 146   bsi = bsi_last (bb);
 147
 148   gcc_assert (TREE_CODE (bsi_stmt (bsi)) == COND_EXPR
 149               || TREE_CODE (bsi_stmt (bsi)) == SWITCH_EXPR);
 150
 151   bsi_remove (&bsi);
 152
 153   next = NULL;
 154   for (e = bb->succ; e; e = next)
 155     {
 156       next = e->succ_next;
 157
 158       if (e->dest != dest_bb)
 159         ssa_remove_edge (e);
 160     }
 161
 162   /* BB now has a single outgoing edge. We need to update the flags for
 163      that single outgoing edge.  */
 164   bb->succ->flags &= ~(EDGE_TRUE_VALUE | EDGE_FALSE_VALUE);
 165   bb->succ->flags |= EDGE_FALLTHRU;
 166 }
 167
 168 /* Create a duplicate of BB which only reaches the destination of the edge
 169    stored in RD.  Record the duplicate block in RD.  */
 170
 171 static void
 172 create_block_for_threading (basic_block bb, struct redirection_data *rd)
 173 {
 174   tree phi;
 175   edge e;
 176
 177   /* We can use the generic block duplication code and simply remove
 178      the stuff we do not need.  */
 179   rd->dup_block = duplicate_block (bb, NULL);
 180
 181   /* Zero out the profile, since the block is unreachable for now.  */
 182   rd->dup_block->frequency = 0;
 183   rd->dup_block->count = 0;
 184
 185   /* The call to duplicate_block will copy everything, including the
 186      useless COND_EXPR or SWITCH_EXPR at the end of the block.  We just remove
 187      the useless COND_EXPR or SWITCH_EXPR here rather than having a
 188      specialized block copier.  */
 189   remove_last_stmt_and_useless_edges (rd->dup_block, rd->outgoing_edge->dest);
 190
 191   for (e = rd->dup_block->succ; e; e = e->succ_next)
 192     e->count = 0;
 193
 194   /* If there are any PHI nodes at the destination of the outgoing edge
 195      from the duplicate block, then we will need to add a new argument
 196      to them.  The argument should have the same value as the argument
 197      associated with the outgoing edge stored in RD.  */
 198   for (phi = phi_nodes (rd->dup_block->succ->dest); phi;
 199        phi = PHI_CHAIN (phi))
 200     {
 201       int indx = phi_arg_from_edge (phi, rd->outgoing_edge);
 202       add_phi_arg (&phi, PHI_ARG_DEF_TREE (phi, indx), rd->dup_block->succ);
 203     }
 204 }
 205
 206 /* BB is a block which ends with a COND_EXPR or SWITCH_EXPR and when BB
 207    is reached via one or more specific incoming edges, we know which
 208    outgoing edge from BB will be traversed.
 209
 210    We want to redirect those incoming edges to the target of the
 211    appropriate outgoing edge.  Doing so avoids a conditional branch
 212    and may expose new optimization opportunities.  Note that we have
 213    to update dominator tree and SSA graph after such changes.
 214
 215    The key to keeping the SSA graph update manageable is to duplicate
 216    the side effects occurring in BB so that those side effects still
 217    occur on the paths which bypass BB after redirecting edges.
 218
 219    We accomplish this by creating duplicates of BB and arranging for
 220    the duplicates to unconditionally pass control to one specific
 221    successor of BB.  We then revector the incoming edges into BB to
 222    the appropriate duplicate of BB.
 223
 224    BB and its duplicates will have assignments to the same set of
 225    SSA_NAMEs.  Right now, we just call into rewrite_ssa_into_ssa
 226    to update the SSA graph for those names.
 227
 228    We are also going to experiment with a true incremental update
 229    scheme for the duplicated resources.  One of the interesting
 230    properties we can exploit here is that all the resources set
 231    in BB will have the same IDFS, so we have one IDFS computation
 232    per block with incoming threaded edges, which can lower the
 233    cost of the true incremental update algorithm.  */
 234
 235 static void
 236 thread_block (basic_block bb)
 237 {
 238   /* E is an incoming edge into BB that we may or may not want to
 239      redirect to a duplicate of BB.  */
 240   edge e;
 241
 242   /* The next edge in a predecessor list.  Used in loops where E->pred_next
 243      may change within the loop.  */
 244   edge next;
 245
 246   /* ALL indicates whether or not all incoming edges into BB should
 247      be threaded to a duplicate of BB.  */
 248   bool all = true;
 249
 250   unsigned int i;
 251
 252   VARRAY_GENERIC_PTR_INIT (redirection_data, 2, "redirection data");
 253
 254   /* Look at each incoming edge into BB.  Record each unique outgoing
 255      edge that we want to thread an incoming edge to.  Also note if
 256      all incoming edges are threaded or not.  */
 257   for (e = bb->pred; e; e = e->pred_next)
 258     {
 259       if (!e->aux)
 260         {
 261           all = false;
 262         }
 263       else
 264         {
 265           unsigned int i;
 266
 267           /* See if we can find an entry for the destination of this
 268              threaded edge that has already been recorded.  */
 269           for (i = 0; i < VARRAY_ACTIVE_SIZE (redirection_data); i++)
 270             {
 271               struct redirection_data *rd;
 272               edge e2;
 273
 274               rd = VARRAY_GENERIC_PTR (redirection_data, i);
 275               e2 = e->aux;
 276
 277               if (e2->dest == rd->outgoing_edge->dest)
 278                 break;
 279             }
 280
 281           /* If the loop did not terminate early, then we have a new
 282              destination for the incoming threaded edges.  Record it.  */
 283           if (i == VARRAY_ACTIVE_SIZE (redirection_data))
 284             {
 285               struct redirection_data *rd;
 286
 287               rd = ggc_alloc_cleared (sizeof (struct redirection_data));
 288               rd->outgoing_edge = e->aux;
 289               VARRAY_PUSH_GENERIC_PTR (redirection_data, rd);
 290             }
 291         }
 292     }
 293
 294   /* Now create duplicates of BB.  Note that if all incoming edges are
 295      threaded, then BB is going to become unreachable.  In that case
 296      we use BB for one of the duplicates rather than wasting memory
 297      duplicating BB.  Thus the odd starting condition for the loop.  */
 298   for (i = (all ? 1 : 0); i < VARRAY_ACTIVE_SIZE (redirection_data); i++)
 299     {
 300       struct redirection_data *rd = VARRAY_GENERIC_PTR (redirection_data, i);
 301       create_block_for_threading (bb, rd);
 302     }
 303
 304   /* The loop above created the duplicate blocks (and the statements
 305      within the duplicate blocks).  This loop creates PHI nodes for the
 306      duplicated blocks and redirects the incoming edges into BB to reach
 307      the duplicates of BB.
 308
 309      Note that redirecting the edge will change e->pred_next, so we have
 310      to hold e->pred_next in a temporary.
 311
 312      If this turns out to be a performance problem, then we could create
 313      a list of incoming edges associated with each entry in
 314      REDIRECTION_DATA and walk over that list of edges instead.  */
 315   next = NULL;
 316   for (e = bb->pred; e; e = next)
 317     {
 318       edge new_dest = e->aux;
 319
 320       next = e->pred_next;
 321
 322       /* E was not threaded, then there is nothing to do.  */
 323       if (!new_dest)
 324         continue;
 325
 326       /* Go ahead and clear E->aux.  It's not needed anymore and failure
 327          to clear it will cause all kinds of unpleasant problems later.  */
 328       e->aux = NULL;
 329
 330       /* We know E is an edge we want to thread.  Find the entry associated
 331          with E's new destination in the REDIRECTION_DATA array.  */
 332       for (i = 0; i < VARRAY_ACTIVE_SIZE (redirection_data); i++)
 333         {
 334           struct redirection_data *rd;
 335
 336           rd = VARRAY_GENERIC_PTR (redirection_data, i);
 337
 338           /* We have found the right entry if the outgoing edge in this
 339              entry matches E's new destination.  Note that if we have not
 340              created a duplicate block (rd->dup_block is NULL), then we
 341              are going to re-use BB as a duplicate and we do not need
 342              to create PHI nodes or redirect the edge.  */
 343           if (rd->outgoing_edge == new_dest && rd->dup_block)
 344             {
 345               edge e2;
 346               copy_phis_to_block (rd->dup_block, bb, e);
 347
 348               if (dump_file && (dump_flags & TDF_DETAILS))
 349                 fprintf (dump_file, "  Threaded jump %d --> %d to %d\n",
 350                          e->src->index, e->dest->index, rd->dup_block->index);
 351
 352               e2 = redirect_edge_and_branch (e, rd->dup_block);
 353               PENDING_STMT (e2) = NULL;
 354
 355               if ((dump_file && (dump_flags & TDF_DETAILS))
 356                   && e->src != e2->src)
 357               fprintf (dump_file, "    basic block %d created\n",
 358                        e2->src->index);
 359               break;
 360             }
 361         }
 362     }
 363
 364   /* If all the incoming edges where threaded, then we used BB as one
 365      of the duplicate blocks.  We need to fixup BB in that case so that
 366      it no longer has a COND_EXPR or SWITCH_EXPR and reaches one destination
 367      unconditionally.  */
 368   if (all)
 369     {
 370       struct redirection_data *rd;
 371
 372       rd = VARRAY_GENERIC_PTR (redirection_data, 0);
 373
 374       if (dump_file && (dump_flags & TDF_DETAILS))
 375         fprintf (dump_file, "  Threaded jump %d --> %d to %d\n",
 376                  bb->pred->src->index, bb->index, bb->succ->dest->index);
 377
 378       remove_last_stmt_and_useless_edges (bb, rd->outgoing_edge->dest);
 379     }
 380
 381   /* Done with this block.  Clear REDIRECTION_DATA.  */
 382   VARRAY_CLEAR (redirection_data);
 383 }
 384
 385 /* Walk through all blocks and thread incoming edges to the block's
 386    destinations as requested.  This is the only entry point into this
 387    file.
 388
 389    Blocks which have one or more incoming edges have INCOMING_EDGE_THREADED
 390    set in the block's annotation.
 391    this routine.
 392
 393    Each edge that should be threaded has the new destination edge stored in
 394    the original edge's AUX field.
 395
 396    This routine (or one of its callees) will clear INCOMING_EDGE_THREADED
 397    in the block annotations and the AUX field in the edges.
 398
 399    It is the caller's responsibility to fix the dominance information
 400    and rewrite duplicated SSA_NAMEs back into SSA form.
 401
 402    Returns true if one or more edges were threaded, false otherwise.   */
 403
 404 bool
 405 thread_through_all_blocks (void)
 406 {
 407   basic_block bb;
 408   bool retval = false;
 409
 410   FOR_EACH_BB (bb)
 411     {
 412       if (bb_ann (bb)->incoming_edge_threaded)
 413         {
 414           thread_block (bb);
 415           retval = true;
 416           bb_ann (bb)->incoming_edge_threaded = false;
 417         }
 418     }
 419   return retval;
 420 }