Mercurial > hg > CbC > CbC_gcc
annotate gcc/bb-reorder.c @ 145:1830386684a0
gcc-9.2.0
| author | anatofuz |
|---|---|
| date | Thu, 13 Feb 2020 11:34:05 +0900 |
| parents | 84e7813d76e9 |
| children |
| rev | line source |
|---|---|
| 0 | 1 /* Basic block reordering routines for the GNU compiler. |
| 145 | 2 Copyright (C) 2000-2020 Free Software Foundation, Inc. |
| 0 | 3 |
| 4 This file is part of GCC. | |
| 5 | |
| 6 GCC is free software; you can redistribute it and/or modify it | |
| 7 under the terms of the GNU General Public License as published by | |
| 8 the Free Software Foundation; either version 3, or (at your option) | |
| 9 any later version. | |
| 10 | |
| 11 GCC is distributed in the hope that it will be useful, but WITHOUT | |
| 12 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY | |
| 13 or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public | |
| 14 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 COPYING3. If not see | |
| 18 <http://www.gnu.org/licenses/>. */ | |
| 19 | |
| 111 | 20 /* This file contains the "reorder blocks" pass, which changes the control |
| 21 flow of a function to encounter fewer branches; the "partition blocks" | |
| 22 pass, which divides the basic blocks into "hot" and "cold" partitions, | |
| 23 which are kept separate; and the "duplicate computed gotos" pass, which | |
| 24 duplicates blocks ending in an indirect jump. | |
| 25 | |
| 26 There are two algorithms for "reorder blocks": the "simple" algorithm, | |
| 27 which just rearranges blocks, trying to minimize the number of executed | |
| 28 unconditional branches; and the "software trace cache" algorithm, which | |
| 29 also copies code, and in general tries a lot harder to have long linear | |
| 30 pieces of machine code executed. This algorithm is described next. */ | |
| 31 | |
| 0 | 32 /* This (greedy) algorithm constructs traces in several rounds. |
| 33 The construction starts from "seeds". The seed for the first round | |
| 111 | 34 is the entry point of the function. When there are more than one seed, |
| 35 the one with the lowest key in the heap is selected first (see bb_to_key). | |
| 36 Then the algorithm repeatedly adds the most probable successor to the end | |
| 37 of a trace. Finally it connects the traces. | |
| 0 | 38 |
| 39 There are two parameters: Branch Threshold and Exec Threshold. | |
| 111 | 40 If the probability of an edge to a successor of the current basic block is |
| 131 | 41 lower than Branch Threshold or its count is lower than Exec Threshold, |
| 111 | 42 then the successor will be the seed in one of the next rounds. |
| 0 | 43 Each round has these parameters lower than the previous one. |
| 111 | 44 The last round has to have these parameters set to zero so that the |
| 45 remaining blocks are picked up. | |
| 0 | 46 |
| 47 The algorithm selects the most probable successor from all unvisited | |
| 48 successors and successors that have been added to this trace. | |
| 49 The other successors (that has not been "sent" to the next round) will be | |
| 111 | 50 other seeds for this round and the secondary traces will start from them. |
| 51 If the successor has not been visited in this trace, it is added to the | |
| 52 trace (however, there is some heuristic for simple branches). | |
| 53 If the successor has been visited in this trace, a loop has been found. | |
| 54 If the loop has many iterations, the loop is rotated so that the source | |
| 55 block of the most probable edge going out of the loop is the last block | |
| 56 of the trace. | |
| 0 | 57 If the loop has few iterations and there is no edge from the last block of |
| 111 | 58 the loop going out of the loop, the loop header is duplicated. |
| 59 | |
| 60 When connecting traces, the algorithm first checks whether there is an edge | |
| 61 from the last block of a trace to the first block of another trace. | |
| 0 | 62 When there are still some unconnected traces it checks whether there exists |
| 111 | 63 a basic block BB such that BB is a successor of the last block of a trace |
| 64 and BB is a predecessor of the first block of another trace. In this case, | |
| 65 BB is duplicated, added at the end of the first trace and the traces are | |
| 66 connected through it. | |
| 0 | 67 The rest of traces are simply connected so there will be a jump to the |
| 111 | 68 beginning of the rest of traces. |
| 69 | |
| 70 The above description is for the full algorithm, which is used when the | |
| 71 function is optimized for speed. When the function is optimized for size, | |
| 72 in order to reduce long jumps and connect more fallthru edges, the | |
| 73 algorithm is modified as follows: | |
| 74 (1) Break long traces to short ones. A trace is broken at a block that has | |
| 75 multiple predecessors/ successors during trace discovery. When connecting | |
| 76 traces, only connect Trace n with Trace n + 1. This change reduces most | |
| 77 long jumps compared with the above algorithm. | |
| 131 | 78 (2) Ignore the edge probability and count for fallthru edges. |
| 111 | 79 (3) Keep the original order of blocks when there is no chance to fall |
| 80 through. We rely on the results of cfg_cleanup. | |
| 81 | |
| 82 To implement the change for code size optimization, block's index is | |
| 83 selected as the key and all traces are found in one round. | |
| 0 | 84 |
| 85 References: | |
| 86 | |
| 87 "Software Trace Cache" | |
| 88 A. Ramirez, J. Larriba-Pey, C. Navarro, J. Torrellas and M. Valero; 1999 | |
| 89 http://citeseer.nj.nec.com/15361.html | |
| 90 | |
| 91 */ | |
| 92 | |
| 93 #include "config.h" | |
| 94 #include "system.h" | |
| 95 #include "coretypes.h" | |
| 111 | 96 #include "backend.h" |
| 97 #include "target.h" | |
| 0 | 98 #include "rtl.h" |
| 111 | 99 #include "tree.h" |
| 100 #include "cfghooks.h" | |
| 101 #include "df.h" | |
| 102 #include "memmodel.h" | |
| 103 #include "optabs.h" | |
| 0 | 104 #include "regs.h" |
| 111 | 105 #include "emit-rtl.h" |
| 0 | 106 #include "output.h" |
| 107 #include "expr.h" | |
| 108 #include "tree-pass.h" | |
| 111 | 109 #include "cfgrtl.h" |
| 110 #include "cfganal.h" | |
| 111 #include "cfgbuild.h" | |
| 112 #include "cfgcleanup.h" | |
|
67
f6334be47118
update gcc from gcc-4.6-20100522 to gcc-4.6-20110318
nobuyasu <dimolto@cr.ie.u-ryukyu.ac.jp>
parents:
63
diff
changeset
|
113 #include "bb-reorder.h" |
| 111 | 114 #include "except.h" |
| 145 | 115 #include "alloc-pool.h" |
| 111 | 116 #include "fibonacci_heap.h" |
| 117 #include "stringpool.h" | |
| 118 #include "attribs.h" | |
| 131 | 119 #include "common/common-target.h" |
| 0 | 120 |
| 121 /* The number of rounds. In most cases there will only be 4 rounds, but | |
| 122 when partitioning hot and cold basic blocks into separate sections of | |
| 111 | 123 the object file there will be an extra round. */ |
| 0 | 124 #define N_ROUNDS 5 |
| 125 | |
|
67
f6334be47118
update gcc from gcc-4.6-20100522 to gcc-4.6-20110318
nobuyasu <dimolto@cr.ie.u-ryukyu.ac.jp>
parents:
63
diff
changeset
|
126 struct target_bb_reorder default_target_bb_reorder; |
|
f6334be47118
update gcc from gcc-4.6-20100522 to gcc-4.6-20110318
nobuyasu <dimolto@cr.ie.u-ryukyu.ac.jp>
parents:
63
diff
changeset
|
127 #if SWITCHABLE_TARGET |
|
f6334be47118
update gcc from gcc-4.6-20100522 to gcc-4.6-20110318
nobuyasu <dimolto@cr.ie.u-ryukyu.ac.jp>
parents:
63
diff
changeset
|
128 struct target_bb_reorder *this_target_bb_reorder = &default_target_bb_reorder; |
|
f6334be47118
update gcc from gcc-4.6-20100522 to gcc-4.6-20110318
nobuyasu <dimolto@cr.ie.u-ryukyu.ac.jp>
parents:
63
diff
changeset
|
129 #endif |
|
f6334be47118
update gcc from gcc-4.6-20100522 to gcc-4.6-20110318
nobuyasu <dimolto@cr.ie.u-ryukyu.ac.jp>
parents:
63
diff
changeset
|
130 |
|
f6334be47118
update gcc from gcc-4.6-20100522 to gcc-4.6-20110318
nobuyasu <dimolto@cr.ie.u-ryukyu.ac.jp>
parents:
63
diff
changeset
|
131 #define uncond_jump_length \ |
|
f6334be47118
update gcc from gcc-4.6-20100522 to gcc-4.6-20110318
nobuyasu <dimolto@cr.ie.u-ryukyu.ac.jp>
parents:
63
diff
changeset
|
132 (this_target_bb_reorder->x_uncond_jump_length) |
|
f6334be47118
update gcc from gcc-4.6-20100522 to gcc-4.6-20110318
nobuyasu <dimolto@cr.ie.u-ryukyu.ac.jp>
parents:
63
diff
changeset
|
133 |
| 0 | 134 /* Branch thresholds in thousandths (per mille) of the REG_BR_PROB_BASE. */ |
| 111 | 135 static const int branch_threshold[N_ROUNDS] = {400, 200, 100, 0, 0}; |
| 0 | 136 |
| 131 | 137 /* Exec thresholds in thousandths (per mille) of the count of bb 0. */ |
| 111 | 138 static const int exec_threshold[N_ROUNDS] = {500, 200, 50, 0, 0}; |
| 0 | 139 |
| 131 | 140 /* If edge count is lower than DUPLICATION_THRESHOLD per mille of entry |
| 0 | 141 block the edge destination is not duplicated while connecting traces. */ |
| 142 #define DUPLICATION_THRESHOLD 100 | |
| 143 | |
| 111 | 144 typedef fibonacci_heap <long, basic_block_def> bb_heap_t; |
| 145 typedef fibonacci_node <long, basic_block_def> bb_heap_node_t; | |
| 146 | |
| 0 | 147 /* Structure to hold needed information for each basic block. */ |
| 111 | 148 struct bbro_basic_block_data |
| 0 | 149 { |
| 111 | 150 /* Which trace is the bb start of (-1 means it is not a start of any). */ |
| 0 | 151 int start_of_trace; |
| 152 | |
| 111 | 153 /* Which trace is the bb end of (-1 means it is not an end of any). */ |
| 0 | 154 int end_of_trace; |
| 155 | |
| 156 /* Which trace is the bb in? */ | |
| 157 int in_trace; | |
| 158 | |
| 111 | 159 /* Which trace was this bb visited in? */ |
| 160 int visited; | |
| 161 | |
| 162 /* Cached maximum frequency of interesting incoming edges. | |
| 163 Minus one means not yet computed. */ | |
| 164 int priority; | |
| 165 | |
| 0 | 166 /* Which heap is BB in (if any)? */ |
| 111 | 167 bb_heap_t *heap; |
| 0 | 168 |
| 169 /* Which heap node is BB in (if any)? */ | |
| 111 | 170 bb_heap_node_t *node; |
| 171 }; | |
| 0 | 172 |
| 173 /* The current size of the following dynamic array. */ | |
| 174 static int array_size; | |
| 175 | |
| 176 /* The array which holds needed information for basic blocks. */ | |
| 177 static bbro_basic_block_data *bbd; | |
| 178 | |
| 179 /* To avoid frequent reallocation the size of arrays is greater than needed, | |
| 180 the number of elements is (not less than) 1.25 * size_wanted. */ | |
| 181 #define GET_ARRAY_SIZE(X) ((((X) / 4) + 1) * 5) | |
| 182 | |
| 183 /* Free the memory and set the pointer to NULL. */ | |
| 184 #define FREE(P) (gcc_assert (P), free (P), P = 0) | |
| 185 | |
| 186 /* Structure for holding information about a trace. */ | |
| 187 struct trace | |
| 188 { | |
| 189 /* First and last basic block of the trace. */ | |
| 190 basic_block first, last; | |
| 191 | |
| 192 /* The round of the STC creation which this trace was found in. */ | |
| 193 int round; | |
| 194 | |
| 195 /* The length (i.e. the number of basic blocks) of the trace. */ | |
| 196 int length; | |
| 197 }; | |
| 198 | |
| 131 | 199 /* Maximum count of one of the entry blocks. */ |
| 111 | 200 static profile_count max_entry_count; |
| 0 | 201 |
| 202 /* Local function prototypes. */ | |
| 131 | 203 static void find_traces_1_round (int, profile_count, struct trace *, int *, |
| 111 | 204 int, bb_heap_t **, int); |
| 0 | 205 static basic_block copy_bb (basic_block, edge, basic_block, int); |
| 111 | 206 static long bb_to_key (basic_block); |
| 207 static bool better_edge_p (const_basic_block, const_edge, profile_probability, | |
| 131 | 208 profile_count, profile_probability, profile_count, |
| 209 const_edge); | |
| 0 | 210 static bool copy_bb_p (const_basic_block, int); |
| 211 | |
| 111 | 212 /* Return the trace number in which BB was visited. */ |
| 213 | |
| 214 static int | |
| 215 bb_visited_trace (const_basic_block bb) | |
| 216 { | |
| 217 gcc_assert (bb->index < array_size); | |
| 218 return bbd[bb->index].visited; | |
| 219 } | |
| 220 | |
| 221 /* This function marks BB that it was visited in trace number TRACE. */ | |
| 222 | |
| 223 static void | |
| 224 mark_bb_visited (basic_block bb, int trace) | |
| 225 { | |
| 226 bbd[bb->index].visited = trace; | |
| 227 if (bbd[bb->index].heap) | |
| 228 { | |
| 229 bbd[bb->index].heap->delete_node (bbd[bb->index].node); | |
| 230 bbd[bb->index].heap = NULL; | |
| 231 bbd[bb->index].node = NULL; | |
| 232 } | |
| 233 } | |
| 234 | |
| 0 | 235 /* Check to see if bb should be pushed into the next round of trace |
| 236 collections or not. Reasons for pushing the block forward are 1). | |
| 237 If the block is cold, we are doing partitioning, and there will be | |
| 238 another round (cold partition blocks are not supposed to be | |
| 239 collected into traces until the very last round); or 2). There will | |
| 240 be another round, and the basic block is not "hot enough" for the | |
| 241 current round of trace collection. */ | |
| 242 | |
| 243 static bool | |
| 244 push_to_next_round_p (const_basic_block bb, int round, int number_of_rounds, | |
| 131 | 245 profile_count count_th) |
| 0 | 246 { |
| 247 bool there_exists_another_round; | |
| 248 bool block_not_hot_enough; | |
| 249 | |
| 250 there_exists_another_round = round < number_of_rounds - 1; | |
| 251 | |
| 131 | 252 block_not_hot_enough = (bb->count < count_th |
| 111 | 253 || probably_never_executed_bb_p (cfun, bb)); |
| 0 | 254 |
| 255 if (there_exists_another_round | |
| 256 && block_not_hot_enough) | |
| 257 return true; | |
| 258 else | |
| 259 return false; | |
| 260 } | |
| 261 | |
| 262 /* Find the traces for Software Trace Cache. Chain each trace through | |
| 263 RBI()->next. Store the number of traces to N_TRACES and description of | |
| 264 traces to TRACES. */ | |
| 265 | |
| 266 static void | |
| 267 find_traces (int *n_traces, struct trace *traces) | |
| 268 { | |
| 269 int i; | |
| 270 int number_of_rounds; | |
| 271 edge e; | |
| 272 edge_iterator ei; | |
| 111 | 273 bb_heap_t *heap = new bb_heap_t (LONG_MIN); |
| 0 | 274 |
| 275 /* Add one extra round of trace collection when partitioning hot/cold | |
| 276 basic blocks into separate sections. The last round is for all the | |
| 277 cold blocks (and ONLY the cold blocks). */ | |
| 278 | |
| 279 number_of_rounds = N_ROUNDS - 1; | |
| 280 | |
| 281 /* Insert entry points of function into heap. */ | |
| 111 | 282 max_entry_count = profile_count::zero (); |
| 283 FOR_EACH_EDGE (e, ei, ENTRY_BLOCK_PTR_FOR_FN (cfun)->succs) | |
| 0 | 284 { |
| 285 bbd[e->dest->index].heap = heap; | |
| 111 | 286 bbd[e->dest->index].node = heap->insert (bb_to_key (e->dest), e->dest); |
| 131 | 287 if (e->dest->count > max_entry_count) |
| 0 | 288 max_entry_count = e->dest->count; |
| 289 } | |
| 290 | |
| 291 /* Find the traces. */ | |
| 292 for (i = 0; i < number_of_rounds; i++) | |
| 293 { | |
| 131 | 294 profile_count count_threshold; |
| 0 | 295 |
| 296 if (dump_file) | |
| 297 fprintf (dump_file, "STC - round %d\n", i + 1); | |
| 298 | |
| 131 | 299 count_threshold = max_entry_count.apply_scale (exec_threshold[i], 1000); |
| 0 | 300 |
| 301 find_traces_1_round (REG_BR_PROB_BASE * branch_threshold[i] / 1000, | |
| 302 count_threshold, traces, n_traces, i, &heap, | |
| 303 number_of_rounds); | |
| 304 } | |
| 111 | 305 delete heap; |
| 0 | 306 |
| 307 if (dump_file) | |
| 308 { | |
| 309 for (i = 0; i < *n_traces; i++) | |
| 310 { | |
| 311 basic_block bb; | |
| 312 fprintf (dump_file, "Trace %d (round %d): ", i + 1, | |
| 313 traces[i].round + 1); | |
| 111 | 314 for (bb = traces[i].first; |
| 315 bb != traces[i].last; | |
| 316 bb = (basic_block) bb->aux) | |
| 131 | 317 { |
| 318 fprintf (dump_file, "%d [", bb->index); | |
| 319 bb->count.dump (dump_file); | |
| 320 fprintf (dump_file, "] "); | |
| 321 } | |
| 322 fprintf (dump_file, "%d [", bb->index); | |
| 323 bb->count.dump (dump_file); | |
| 324 fprintf (dump_file, "]\n"); | |
| 0 | 325 } |
| 326 fflush (dump_file); | |
| 327 } | |
| 328 } | |
| 329 | |
| 330 /* Rotate loop whose back edge is BACK_EDGE in the tail of trace TRACE | |
| 331 (with sequential number TRACE_N). */ | |
| 332 | |
| 333 static basic_block | |
| 334 rotate_loop (edge back_edge, struct trace *trace, int trace_n) | |
| 335 { | |
| 336 basic_block bb; | |
| 337 | |
| 338 /* Information about the best end (end after rotation) of the loop. */ | |
| 339 basic_block best_bb = NULL; | |
| 340 edge best_edge = NULL; | |
| 111 | 341 profile_count best_count = profile_count::uninitialized (); |
| 0 | 342 /* The best edge is preferred when its destination is not visited yet |
| 343 or is a start block of some trace. */ | |
| 344 bool is_preferred = false; | |
| 345 | |
| 346 /* Find the most frequent edge that goes out from current trace. */ | |
| 347 bb = back_edge->dest; | |
| 348 do | |
| 349 { | |
| 350 edge e; | |
| 351 edge_iterator ei; | |
| 352 | |
| 353 FOR_EACH_EDGE (e, ei, bb->succs) | |
| 111 | 354 if (e->dest != EXIT_BLOCK_PTR_FOR_FN (cfun) |
| 355 && bb_visited_trace (e->dest) != trace_n | |
| 0 | 356 && (e->flags & EDGE_CAN_FALLTHRU) |
| 357 && !(e->flags & EDGE_COMPLEX)) | |
| 358 { | |
| 359 if (is_preferred) | |
| 360 { | |
| 361 /* The best edge is preferred. */ | |
| 111 | 362 if (!bb_visited_trace (e->dest) |
| 0 | 363 || bbd[e->dest->index].start_of_trace >= 0) |
| 364 { | |
| 365 /* The current edge E is also preferred. */ | |
| 131 | 366 if (e->count () > best_count) |
| 0 | 367 { |
| 131 | 368 best_count = e->count (); |
| 0 | 369 best_edge = e; |
| 370 best_bb = bb; | |
| 371 } | |
| 372 } | |
| 373 } | |
| 374 else | |
| 375 { | |
| 111 | 376 if (!bb_visited_trace (e->dest) |
| 0 | 377 || bbd[e->dest->index].start_of_trace >= 0) |
| 378 { | |
| 379 /* The current edge E is preferred. */ | |
| 380 is_preferred = true; | |
| 111 | 381 best_count = e->count (); |
| 0 | 382 best_edge = e; |
| 383 best_bb = bb; | |
| 384 } | |
| 385 else | |
| 386 { | |
| 131 | 387 if (!best_edge || e->count () > best_count) |
| 0 | 388 { |
| 111 | 389 best_count = e->count (); |
| 0 | 390 best_edge = e; |
| 391 best_bb = bb; | |
| 392 } | |
| 393 } | |
| 394 } | |
| 395 } | |
| 396 bb = (basic_block) bb->aux; | |
| 397 } | |
| 398 while (bb != back_edge->dest); | |
| 399 | |
| 400 if (best_bb) | |
| 401 { | |
| 402 /* Rotate the loop so that the BEST_EDGE goes out from the last block of | |
| 403 the trace. */ | |
| 404 if (back_edge->dest == trace->first) | |
| 405 { | |
| 406 trace->first = (basic_block) best_bb->aux; | |
| 407 } | |
| 408 else | |
| 409 { | |
| 410 basic_block prev_bb; | |
| 411 | |
| 412 for (prev_bb = trace->first; | |
| 413 prev_bb->aux != back_edge->dest; | |
| 414 prev_bb = (basic_block) prev_bb->aux) | |
| 415 ; | |
| 416 prev_bb->aux = best_bb->aux; | |
| 417 | |
| 418 /* Try to get rid of uncond jump to cond jump. */ | |
| 419 if (single_succ_p (prev_bb)) | |
| 420 { | |
| 421 basic_block header = single_succ (prev_bb); | |
| 422 | |
| 423 /* Duplicate HEADER if it is a small block containing cond jump | |
| 424 in the end. */ | |
| 425 if (any_condjump_p (BB_END (header)) && copy_bb_p (header, 0) | |
| 111 | 426 && !CROSSING_JUMP_P (BB_END (header))) |
| 0 | 427 copy_bb (header, single_succ_edge (prev_bb), prev_bb, trace_n); |
| 428 } | |
| 429 } | |
| 430 } | |
| 431 else | |
| 432 { | |
| 433 /* We have not found suitable loop tail so do no rotation. */ | |
| 434 best_bb = back_edge->src; | |
| 435 } | |
| 436 best_bb->aux = NULL; | |
| 437 return best_bb; | |
| 438 } | |
| 439 | |
| 111 | 440 /* One round of finding traces. Find traces for BRANCH_TH and EXEC_TH i.e. do |
| 441 not include basic blocks whose probability is lower than BRANCH_TH or whose | |
| 131 | 442 count is lower than EXEC_TH into traces (or whose count is lower than |
| 111 | 443 COUNT_TH). Store the new traces into TRACES and modify the number of |
| 444 traces *N_TRACES. Set the round (which the trace belongs to) to ROUND. | |
| 445 The function expects starting basic blocks to be in *HEAP and will delete | |
| 446 *HEAP and store starting points for the next round into new *HEAP. */ | |
| 0 | 447 |
| 448 static void | |
| 131 | 449 find_traces_1_round (int branch_th, profile_count count_th, |
| 0 | 450 struct trace *traces, int *n_traces, int round, |
| 111 | 451 bb_heap_t **heap, int number_of_rounds) |
| 0 | 452 { |
| 453 /* Heap for discarded basic blocks which are possible starting points for | |
| 454 the next round. */ | |
| 111 | 455 bb_heap_t *new_heap = new bb_heap_t (LONG_MIN); |
| 456 bool for_size = optimize_function_for_size_p (cfun); | |
| 457 | |
| 458 while (!(*heap)->empty ()) | |
| 0 | 459 { |
| 460 basic_block bb; | |
| 461 struct trace *trace; | |
| 462 edge best_edge, e; | |
| 111 | 463 long key; |
| 0 | 464 edge_iterator ei; |
| 465 | |
| 111 | 466 bb = (*heap)->extract_min (); |
| 0 | 467 bbd[bb->index].heap = NULL; |
| 468 bbd[bb->index].node = NULL; | |
| 469 | |
| 470 if (dump_file) | |
| 471 fprintf (dump_file, "Getting bb %d\n", bb->index); | |
| 472 | |
| 131 | 473 /* If the BB's count is too low, send BB to the next round. When |
| 0 | 474 partitioning hot/cold blocks into separate sections, make sure all |
| 475 the cold blocks (and ONLY the cold blocks) go into the (extra) final | |
| 111 | 476 round. When optimizing for size, do not push to next round. */ |
| 477 | |
| 478 if (!for_size | |
| 131 | 479 && push_to_next_round_p (bb, round, number_of_rounds, |
| 111 | 480 count_th)) |
| 0 | 481 { |
| 482 int key = bb_to_key (bb); | |
| 483 bbd[bb->index].heap = new_heap; | |
| 111 | 484 bbd[bb->index].node = new_heap->insert (key, bb); |
| 0 | 485 |
| 486 if (dump_file) | |
| 487 fprintf (dump_file, | |
| 488 " Possible start point of next round: %d (key: %d)\n", | |
| 489 bb->index, key); | |
| 490 continue; | |
| 491 } | |
| 492 | |
| 493 trace = traces + *n_traces; | |
| 494 trace->first = bb; | |
| 495 trace->round = round; | |
| 496 trace->length = 0; | |
| 497 bbd[bb->index].in_trace = *n_traces; | |
| 498 (*n_traces)++; | |
| 499 | |
| 500 do | |
| 501 { | |
| 502 bool ends_in_call; | |
| 503 | |
| 131 | 504 /* The probability and count of the best edge. */ |
| 111 | 505 profile_probability best_prob = profile_probability::uninitialized (); |
| 131 | 506 profile_count best_count = profile_count::uninitialized (); |
| 0 | 507 |
| 508 best_edge = NULL; | |
| 509 mark_bb_visited (bb, *n_traces); | |
| 510 trace->length++; | |
| 511 | |
| 512 if (dump_file) | |
| 513 fprintf (dump_file, "Basic block %d was visited in trace %d\n", | |
| 131 | 514 bb->index, *n_traces); |
| 0 | 515 |
| 516 ends_in_call = block_ends_with_call_p (bb); | |
| 517 | |
| 518 /* Select the successor that will be placed after BB. */ | |
| 519 FOR_EACH_EDGE (e, ei, bb->succs) | |
| 520 { | |
| 521 gcc_assert (!(e->flags & EDGE_FAKE)); | |
| 522 | |
| 111 | 523 if (e->dest == EXIT_BLOCK_PTR_FOR_FN (cfun)) |
| 0 | 524 continue; |
| 525 | |
| 111 | 526 if (bb_visited_trace (e->dest) |
| 527 && bb_visited_trace (e->dest) != *n_traces) | |
| 0 | 528 continue; |
| 529 | |
| 131 | 530 /* If partitioning hot/cold basic blocks, don't consider edges |
| 531 that cross section boundaries. */ | |
| 0 | 532 if (BB_PARTITION (e->dest) != BB_PARTITION (bb)) |
| 533 continue; | |
| 534 | |
| 131 | 535 profile_probability prob = e->probability; |
| 536 profile_count count = e->dest->count; | |
| 0 | 537 |
| 538 /* The only sensible preference for a call instruction is the | |
| 539 fallthru edge. Don't bother selecting anything else. */ | |
| 540 if (ends_in_call) | |
| 541 { | |
| 542 if (e->flags & EDGE_CAN_FALLTHRU) | |
| 543 { | |
| 544 best_edge = e; | |
| 545 best_prob = prob; | |
| 131 | 546 best_count = count; |
| 0 | 547 } |
| 548 continue; | |
| 549 } | |
| 550 | |
| 551 /* Edge that cannot be fallthru or improbable or infrequent | |
| 111 | 552 successor (i.e. it is unsuitable successor). When optimizing |
| 131 | 553 for size, ignore the probability and count. */ |
| 0 | 554 if (!(e->flags & EDGE_CAN_FALLTHRU) || (e->flags & EDGE_COMPLEX) |
| 111 | 555 || !prob.initialized_p () |
| 556 || ((prob.to_reg_br_prob_base () < branch_th | |
| 557 || e->count () < count_th) && (!for_size))) | |
| 0 | 558 continue; |
| 559 | |
| 131 | 560 if (better_edge_p (bb, e, prob, count, best_prob, best_count, |
| 0 | 561 best_edge)) |
| 562 { | |
| 563 best_edge = e; | |
| 564 best_prob = prob; | |
| 131 | 565 best_count = count; |
| 0 | 566 } |
| 567 } | |
| 568 | |
| 131 | 569 /* If the best destination has multiple predecessors and can be |
| 570 duplicated cheaper than a jump, don't allow it to be added to | |
| 571 a trace; we'll duplicate it when connecting the traces later. | |
| 572 However, we need to check that this duplication wouldn't leave | |
| 573 the best destination with only crossing predecessors, because | |
| 574 this would change its effective partition from hot to cold. */ | |
| 575 if (best_edge | |
| 576 && EDGE_COUNT (best_edge->dest->preds) >= 2 | |
| 0 | 577 && copy_bb_p (best_edge->dest, 0)) |
| 131 | 578 { |
| 579 bool only_crossing_preds = true; | |
| 580 edge e; | |
| 581 edge_iterator ei; | |
| 582 FOR_EACH_EDGE (e, ei, best_edge->dest->preds) | |
| 583 if (e != best_edge && !(e->flags & EDGE_CROSSING)) | |
| 584 { | |
| 585 only_crossing_preds = false; | |
| 586 break; | |
| 587 } | |
| 588 if (!only_crossing_preds) | |
| 589 best_edge = NULL; | |
| 590 } | |
| 0 | 591 |
| 111 | 592 /* If the best destination has multiple successors or predecessors, |
| 593 don't allow it to be added when optimizing for size. This makes | |
| 594 sure predecessors with smaller index are handled before the best | |
| 595 destinarion. It breaks long trace and reduces long jumps. | |
| 596 | |
| 597 Take if-then-else as an example. | |
| 598 A | |
| 599 / \ | |
| 600 B C | |
| 601 \ / | |
| 602 D | |
| 603 If we do not remove the best edge B->D/C->D, the final order might | |
| 604 be A B D ... C. C is at the end of the program. If D's successors | |
| 605 and D are complicated, might need long jumps for A->C and C->D. | |
| 606 Similar issue for order: A C D ... B. | |
| 607 | |
| 608 After removing the best edge, the final result will be ABCD/ ACBD. | |
| 609 It does not add jump compared with the previous order. But it | |
| 610 reduces the possibility of long jumps. */ | |
| 611 if (best_edge && for_size | |
| 612 && (EDGE_COUNT (best_edge->dest->succs) > 1 | |
| 613 || EDGE_COUNT (best_edge->dest->preds) > 1)) | |
| 614 best_edge = NULL; | |
| 615 | |
| 0 | 616 /* Add all non-selected successors to the heaps. */ |
| 617 FOR_EACH_EDGE (e, ei, bb->succs) | |
| 618 { | |
| 619 if (e == best_edge | |
| 111 | 620 || e->dest == EXIT_BLOCK_PTR_FOR_FN (cfun) |
| 621 || bb_visited_trace (e->dest)) | |
| 0 | 622 continue; |
| 623 | |
| 624 key = bb_to_key (e->dest); | |
| 625 | |
| 626 if (bbd[e->dest->index].heap) | |
| 627 { | |
| 628 /* E->DEST is already in some heap. */ | |
| 111 | 629 if (key != bbd[e->dest->index].node->get_key ()) |
| 0 | 630 { |
| 631 if (dump_file) | |
| 632 { | |
| 633 fprintf (dump_file, | |
| 634 "Changing key for bb %d from %ld to %ld.\n", | |
| 635 e->dest->index, | |
| 111 | 636 (long) bbd[e->dest->index].node->get_key (), |
| 0 | 637 key); |
| 638 } | |
| 111 | 639 bbd[e->dest->index].heap->replace_key |
| 640 (bbd[e->dest->index].node, key); | |
| 0 | 641 } |
| 642 } | |
| 643 else | |
| 644 { | |
| 111 | 645 bb_heap_t *which_heap = *heap; |
| 0 | 646 |
| 131 | 647 profile_probability prob = e->probability; |
| 0 | 648 |
| 649 if (!(e->flags & EDGE_CAN_FALLTHRU) | |
| 650 || (e->flags & EDGE_COMPLEX) | |
| 111 | 651 || !prob.initialized_p () |
| 652 || prob.to_reg_br_prob_base () < branch_th | |
| 653 || e->count () < count_th) | |
| 0 | 654 { |
| 655 /* When partitioning hot/cold basic blocks, make sure | |
| 656 the cold blocks (and only the cold blocks) all get | |
| 111 | 657 pushed to the last round of trace collection. When |
| 658 optimizing for size, do not push to next round. */ | |
| 659 | |
| 660 if (!for_size && push_to_next_round_p (e->dest, round, | |
| 661 number_of_rounds, | |
| 131 | 662 count_th)) |
| 0 | 663 which_heap = new_heap; |
| 664 } | |
| 665 | |
| 666 bbd[e->dest->index].heap = which_heap; | |
| 111 | 667 bbd[e->dest->index].node = which_heap->insert (key, e->dest); |
| 0 | 668 |
| 669 if (dump_file) | |
| 670 { | |
| 671 fprintf (dump_file, | |
| 672 " Possible start of %s round: %d (key: %ld)\n", | |
| 673 (which_heap == new_heap) ? "next" : "this", | |
| 674 e->dest->index, (long) key); | |
| 675 } | |
| 676 | |
| 677 } | |
| 678 } | |
| 679 | |
| 680 if (best_edge) /* Suitable successor was found. */ | |
| 681 { | |
| 111 | 682 if (bb_visited_trace (best_edge->dest) == *n_traces) |
| 0 | 683 { |
| 684 /* We do nothing with one basic block loops. */ | |
| 685 if (best_edge->dest != bb) | |
| 686 { | |
| 131 | 687 if (best_edge->count () |
| 688 > best_edge->dest->count.apply_scale (4, 5)) | |
| 0 | 689 { |
| 690 /* The loop has at least 4 iterations. If the loop | |
| 691 header is not the first block of the function | |
| 692 we can rotate the loop. */ | |
| 693 | |
| 111 | 694 if (best_edge->dest |
| 695 != ENTRY_BLOCK_PTR_FOR_FN (cfun)->next_bb) | |
| 0 | 696 { |
| 697 if (dump_file) | |
| 698 { | |
| 699 fprintf (dump_file, | |
| 700 "Rotating loop %d - %d\n", | |
| 701 best_edge->dest->index, bb->index); | |
| 702 } | |
| 703 bb->aux = best_edge->dest; | |
| 704 bbd[best_edge->dest->index].in_trace = | |
| 705 (*n_traces) - 1; | |
| 706 bb = rotate_loop (best_edge, trace, *n_traces); | |
| 707 } | |
| 708 } | |
| 709 else | |
| 710 { | |
| 711 /* The loop has less than 4 iterations. */ | |
| 712 | |
| 713 if (single_succ_p (bb) | |
| 714 && copy_bb_p (best_edge->dest, | |
| 111 | 715 optimize_edge_for_speed_p |
| 716 (best_edge))) | |
| 0 | 717 { |
| 718 bb = copy_bb (best_edge->dest, best_edge, bb, | |
| 719 *n_traces); | |
| 720 trace->length++; | |
| 721 } | |
| 722 } | |
| 723 } | |
| 724 | |
| 725 /* Terminate the trace. */ | |
| 726 break; | |
| 727 } | |
| 728 else | |
| 729 { | |
| 730 /* Check for a situation | |
| 731 | |
| 732 A | |
| 733 /| | |
| 734 B | | |
| 735 \| | |
| 736 C | |
| 737 | |
| 738 where | |
| 131 | 739 AB->count () + BC->count () >= AC->count (). |
| 740 (i.e. 2 * B->count >= AC->count ) | |
| 0 | 741 Best ordering is then A B C. |
| 742 | |
| 111 | 743 When optimizing for size, A B C is always the best order. |
| 744 | |
| 0 | 745 This situation is created for example by: |
| 746 | |
| 747 if (A) B; | |
| 748 C; | |
| 749 | |
| 750 */ | |
| 751 | |
| 752 FOR_EACH_EDGE (e, ei, bb->succs) | |
| 753 if (e != best_edge | |
| 754 && (e->flags & EDGE_CAN_FALLTHRU) | |
| 755 && !(e->flags & EDGE_COMPLEX) | |
| 111 | 756 && !bb_visited_trace (e->dest) |
| 0 | 757 && single_pred_p (e->dest) |
| 758 && !(e->flags & EDGE_CROSSING) | |
| 759 && single_succ_p (e->dest) | |
| 760 && (single_succ_edge (e->dest)->flags | |
| 761 & EDGE_CAN_FALLTHRU) | |
| 762 && !(single_succ_edge (e->dest)->flags & EDGE_COMPLEX) | |
| 763 && single_succ (e->dest) == best_edge->dest | |
| 131 | 764 && (e->dest->count.apply_scale (2, 1) |
| 765 >= best_edge->count () || for_size)) | |
| 0 | 766 { |
| 767 best_edge = e; | |
| 768 if (dump_file) | |
| 769 fprintf (dump_file, "Selecting BB %d\n", | |
| 770 best_edge->dest->index); | |
| 771 break; | |
| 772 } | |
| 773 | |
| 774 bb->aux = best_edge->dest; | |
| 775 bbd[best_edge->dest->index].in_trace = (*n_traces) - 1; | |
| 776 bb = best_edge->dest; | |
| 777 } | |
| 778 } | |
| 779 } | |
| 780 while (best_edge); | |
| 781 trace->last = bb; | |
| 782 bbd[trace->first->index].start_of_trace = *n_traces - 1; | |
| 111 | 783 if (bbd[trace->last->index].end_of_trace != *n_traces - 1) |
| 784 { | |
| 785 bbd[trace->last->index].end_of_trace = *n_traces - 1; | |
| 786 /* Update the cached maximum frequency for interesting predecessor | |
| 787 edges for successors of the new trace end. */ | |
| 788 FOR_EACH_EDGE (e, ei, trace->last->succs) | |
| 789 if (EDGE_FREQUENCY (e) > bbd[e->dest->index].priority) | |
| 790 bbd[e->dest->index].priority = EDGE_FREQUENCY (e); | |
| 791 } | |
| 0 | 792 |
| 793 /* The trace is terminated so we have to recount the keys in heap | |
| 794 (some block can have a lower key because now one of its predecessors | |
| 795 is an end of the trace). */ | |
| 796 FOR_EACH_EDGE (e, ei, bb->succs) | |
| 797 { | |
| 111 | 798 if (e->dest == EXIT_BLOCK_PTR_FOR_FN (cfun) |
| 799 || bb_visited_trace (e->dest)) | |
| 0 | 800 continue; |
| 801 | |
| 802 if (bbd[e->dest->index].heap) | |
| 803 { | |
| 804 key = bb_to_key (e->dest); | |
| 111 | 805 if (key != bbd[e->dest->index].node->get_key ()) |
| 0 | 806 { |
| 807 if (dump_file) | |
| 808 { | |
| 809 fprintf (dump_file, | |
| 810 "Changing key for bb %d from %ld to %ld.\n", | |
| 811 e->dest->index, | |
| 111 | 812 (long) bbd[e->dest->index].node->get_key (), key); |
| 0 | 813 } |
| 111 | 814 bbd[e->dest->index].heap->replace_key |
| 815 (bbd[e->dest->index].node, key); | |
| 0 | 816 } |
| 817 } | |
| 818 } | |
| 819 } | |
| 820 | |
| 111 | 821 delete (*heap); |
| 0 | 822 |
| 823 /* "Return" the new heap. */ | |
| 824 *heap = new_heap; | |
| 825 } | |
| 826 | |
| 827 /* Create a duplicate of the basic block OLD_BB and redirect edge E to it, add | |
| 828 it to trace after BB, mark OLD_BB visited and update pass' data structures | |
| 829 (TRACE is a number of trace which OLD_BB is duplicated to). */ | |
| 830 | |
| 831 static basic_block | |
| 832 copy_bb (basic_block old_bb, edge e, basic_block bb, int trace) | |
| 833 { | |
| 834 basic_block new_bb; | |
| 835 | |
| 836 new_bb = duplicate_block (old_bb, e, bb); | |
| 837 BB_COPY_PARTITION (new_bb, old_bb); | |
| 838 | |
| 839 gcc_assert (e->dest == new_bb); | |
| 840 | |
| 841 if (dump_file) | |
| 842 fprintf (dump_file, | |
| 843 "Duplicated bb %d (created bb %d)\n", | |
| 844 old_bb->index, new_bb->index); | |
| 111 | 845 |
| 846 if (new_bb->index >= array_size | |
| 847 || last_basic_block_for_fn (cfun) > array_size) | |
| 0 | 848 { |
| 849 int i; | |
| 850 int new_size; | |
| 851 | |
| 111 | 852 new_size = MAX (last_basic_block_for_fn (cfun), new_bb->index + 1); |
| 0 | 853 new_size = GET_ARRAY_SIZE (new_size); |
| 854 bbd = XRESIZEVEC (bbro_basic_block_data, bbd, new_size); | |
| 855 for (i = array_size; i < new_size; i++) | |
| 856 { | |
| 857 bbd[i].start_of_trace = -1; | |
| 111 | 858 bbd[i].end_of_trace = -1; |
| 0 | 859 bbd[i].in_trace = -1; |
| 111 | 860 bbd[i].visited = 0; |
| 861 bbd[i].priority = -1; | |
| 0 | 862 bbd[i].heap = NULL; |
| 863 bbd[i].node = NULL; | |
| 864 } | |
| 865 array_size = new_size; | |
| 866 | |
| 867 if (dump_file) | |
| 868 { | |
| 869 fprintf (dump_file, | |
| 870 "Growing the dynamic array to %d elements.\n", | |
| 871 array_size); | |
| 872 } | |
| 873 } | |
| 874 | |
| 111 | 875 gcc_assert (!bb_visited_trace (e->dest)); |
| 876 mark_bb_visited (new_bb, trace); | |
| 877 new_bb->aux = bb->aux; | |
| 878 bb->aux = new_bb; | |
| 879 | |
| 0 | 880 bbd[new_bb->index].in_trace = trace; |
| 881 | |
| 882 return new_bb; | |
| 883 } | |
| 884 | |
| 885 /* Compute and return the key (for the heap) of the basic block BB. */ | |
| 886 | |
| 111 | 887 static long |
| 0 | 888 bb_to_key (basic_block bb) |
| 889 { | |
| 890 edge e; | |
| 891 edge_iterator ei; | |
| 111 | 892 |
| 893 /* Use index as key to align with its original order. */ | |
| 894 if (optimize_function_for_size_p (cfun)) | |
| 895 return bb->index; | |
| 0 | 896 |
| 897 /* Do not start in probably never executed blocks. */ | |
| 898 | |
| 899 if (BB_PARTITION (bb) == BB_COLD_PARTITION | |
| 111 | 900 || probably_never_executed_bb_p (cfun, bb)) |
| 0 | 901 return BB_FREQ_MAX; |
| 902 | |
| 903 /* Prefer blocks whose predecessor is an end of some trace | |
| 904 or whose predecessor edge is EDGE_DFS_BACK. */ | |
| 111 | 905 int priority = bbd[bb->index].priority; |
| 906 if (priority == -1) | |
| 0 | 907 { |
| 111 | 908 priority = 0; |
| 909 FOR_EACH_EDGE (e, ei, bb->preds) | |
| 0 | 910 { |
| 111 | 911 if ((e->src != ENTRY_BLOCK_PTR_FOR_FN (cfun) |
| 912 && bbd[e->src->index].end_of_trace >= 0) | |
| 913 || (e->flags & EDGE_DFS_BACK)) | |
| 914 { | |
| 915 int edge_freq = EDGE_FREQUENCY (e); | |
| 916 | |
| 917 if (edge_freq > priority) | |
| 918 priority = edge_freq; | |
| 919 } | |
| 0 | 920 } |
| 111 | 921 bbd[bb->index].priority = priority; |
| 0 | 922 } |
| 923 | |
| 924 if (priority) | |
| 925 /* The block with priority should have significantly lower key. */ | |
| 131 | 926 return -(100 * BB_FREQ_MAX + 100 * priority + bb->count.to_frequency (cfun)); |
| 927 | |
| 928 return -bb->count.to_frequency (cfun); | |
| 0 | 929 } |
| 930 | |
| 931 /* Return true when the edge E from basic block BB is better than the temporary | |
| 932 best edge (details are in function). The probability of edge E is PROB. The | |
| 131 | 933 count of the successor is COUNT. The current best probability is |
| 934 BEST_PROB, the best count is BEST_COUNT. | |
| 0 | 935 The edge is considered to be equivalent when PROB does not differ much from |
| 131 | 936 BEST_PROB; similarly for count. */ |
| 0 | 937 |
| 938 static bool | |
| 111 | 939 better_edge_p (const_basic_block bb, const_edge e, profile_probability prob, |
| 131 | 940 profile_count count, profile_probability best_prob, |
| 941 profile_count best_count, const_edge cur_best_edge) | |
| 0 | 942 { |
| 943 bool is_better_edge; | |
| 944 | |
| 945 /* The BEST_* values do not have to be best, but can be a bit smaller than | |
| 946 maximum values. */ | |
| 111 | 947 profile_probability diff_prob = best_prob.apply_scale (1, 10); |
| 0 | 948 |
| 111 | 949 /* The smaller one is better to keep the original order. */ |
| 950 if (optimize_function_for_size_p (cfun)) | |
| 951 return !cur_best_edge | |
| 952 || cur_best_edge->dest->index > e->dest->index; | |
| 953 | |
| 954 /* Those edges are so expensive that continuing a trace is not useful | |
| 955 performance wise. */ | |
| 956 if (e->flags & (EDGE_ABNORMAL | EDGE_EH)) | |
| 957 return false; | |
| 958 | |
| 959 if (prob > best_prob + diff_prob | |
| 960 || (!best_prob.initialized_p () | |
| 961 && prob > profile_probability::guessed_never ())) | |
| 0 | 962 /* The edge has higher probability than the temporary best edge. */ |
| 963 is_better_edge = true; | |
| 964 else if (prob < best_prob - diff_prob) | |
| 965 /* The edge has lower probability than the temporary best edge. */ | |
| 966 is_better_edge = false; | |
| 967 else | |
| 131 | 968 { |
| 969 profile_count diff_count = best_count.apply_scale (1, 10); | |
| 970 if (count < best_count - diff_count | |
| 971 || (!best_count.initialized_p () | |
| 972 && count.nonzero_p ())) | |
| 973 /* The edge and the temporary best edge have almost equivalent | |
| 974 probabilities. The higher countuency of a successor now means | |
| 975 that there is another edge going into that successor. | |
| 976 This successor has lower countuency so it is better. */ | |
| 977 is_better_edge = true; | |
| 978 else if (count > best_count + diff_count) | |
| 979 /* This successor has higher countuency so it is worse. */ | |
| 980 is_better_edge = false; | |
| 981 else if (e->dest->prev_bb == bb) | |
| 982 /* The edges have equivalent probabilities and the successors | |
| 983 have equivalent frequencies. Select the previous successor. */ | |
| 984 is_better_edge = true; | |
| 985 else | |
| 986 is_better_edge = false; | |
| 987 } | |
| 0 | 988 |
| 989 return is_better_edge; | |
| 990 } | |
| 991 | |
| 111 | 992 /* Return true when the edge E is better than the temporary best edge |
| 993 CUR_BEST_EDGE. If SRC_INDEX_P is true, the function compares the src bb of | |
| 994 E and CUR_BEST_EDGE; otherwise it will compare the dest bb. | |
| 995 BEST_LEN is the trace length of src (or dest) bb in CUR_BEST_EDGE. | |
| 996 TRACES record the information about traces. | |
| 997 When optimizing for size, the edge with smaller index is better. | |
| 998 When optimizing for speed, the edge with bigger probability or longer trace | |
| 999 is better. */ | |
| 1000 | |
| 1001 static bool | |
| 1002 connect_better_edge_p (const_edge e, bool src_index_p, int best_len, | |
| 1003 const_edge cur_best_edge, struct trace *traces) | |
| 1004 { | |
| 1005 int e_index; | |
| 1006 int b_index; | |
| 1007 bool is_better_edge; | |
| 1008 | |
| 1009 if (!cur_best_edge) | |
| 1010 return true; | |
| 1011 | |
| 1012 if (optimize_function_for_size_p (cfun)) | |
| 1013 { | |
| 1014 e_index = src_index_p ? e->src->index : e->dest->index; | |
| 1015 b_index = src_index_p ? cur_best_edge->src->index | |
| 1016 : cur_best_edge->dest->index; | |
| 1017 /* The smaller one is better to keep the original order. */ | |
| 1018 return b_index > e_index; | |
| 1019 } | |
| 1020 | |
| 1021 if (src_index_p) | |
| 1022 { | |
| 1023 e_index = e->src->index; | |
| 1024 | |
| 131 | 1025 /* We are looking for predecessor, so probabilities are not that |
| 1026 informative. We do not want to connect A to B becuse A has | |
| 1027 only one sucessor (probablity is 100%) while there is edge | |
| 1028 A' to B where probability is 90% but which is much more frequent. */ | |
| 1029 if (e->count () > cur_best_edge->count ()) | |
| 1030 /* The edge has higher probability than the temporary best edge. */ | |
| 1031 is_better_edge = true; | |
| 1032 else if (e->count () < cur_best_edge->count ()) | |
| 1033 /* The edge has lower probability than the temporary best edge. */ | |
| 1034 is_better_edge = false; | |
| 145 | 1035 else if (e->probability > cur_best_edge->probability) |
| 111 | 1036 /* The edge has higher probability than the temporary best edge. */ |
| 1037 is_better_edge = true; | |
| 1038 else if (e->probability < cur_best_edge->probability) | |
| 1039 /* The edge has lower probability than the temporary best edge. */ | |
| 1040 is_better_edge = false; | |
| 1041 else if (traces[bbd[e_index].end_of_trace].length > best_len) | |
| 1042 /* The edge and the temporary best edge have equivalent probabilities. | |
| 1043 The edge with longer trace is better. */ | |
| 1044 is_better_edge = true; | |
| 1045 else | |
| 1046 is_better_edge = false; | |
| 1047 } | |
| 1048 else | |
| 1049 { | |
| 1050 e_index = e->dest->index; | |
| 1051 | |
| 1052 if (e->probability > cur_best_edge->probability) | |
| 1053 /* The edge has higher probability than the temporary best edge. */ | |
| 1054 is_better_edge = true; | |
| 1055 else if (e->probability < cur_best_edge->probability) | |
| 1056 /* The edge has lower probability than the temporary best edge. */ | |
| 1057 is_better_edge = false; | |
| 1058 else if (traces[bbd[e_index].start_of_trace].length > best_len) | |
| 1059 /* The edge and the temporary best edge have equivalent probabilities. | |
| 1060 The edge with longer trace is better. */ | |
| 1061 is_better_edge = true; | |
| 1062 else | |
| 1063 is_better_edge = false; | |
| 1064 } | |
| 1065 | |
| 1066 return is_better_edge; | |
| 1067 } | |
| 1068 | |
| 0 | 1069 /* Connect traces in array TRACES, N_TRACES is the count of traces. */ |
| 1070 | |
| 1071 static void | |
| 1072 connect_traces (int n_traces, struct trace *traces) | |
| 1073 { | |
| 1074 int i; | |
| 1075 bool *connected; | |
| 1076 bool two_passes; | |
| 1077 int last_trace; | |
| 1078 int current_pass; | |
| 1079 int current_partition; | |
| 131 | 1080 profile_count count_threshold; |
| 111 | 1081 bool for_size = optimize_function_for_size_p (cfun); |
| 0 | 1082 |
| 131 | 1083 count_threshold = max_entry_count.apply_scale (DUPLICATION_THRESHOLD, 1000); |
| 0 | 1084 |
| 1085 connected = XCNEWVEC (bool, n_traces); | |
| 1086 last_trace = -1; | |
| 1087 current_pass = 1; | |
| 1088 current_partition = BB_PARTITION (traces[0].first); | |
| 1089 two_passes = false; | |
| 1090 | |
| 111 | 1091 if (crtl->has_bb_partition) |
| 0 | 1092 for (i = 0; i < n_traces && !two_passes; i++) |
| 1093 if (BB_PARTITION (traces[0].first) | |
| 1094 != BB_PARTITION (traces[i].first)) | |
| 1095 two_passes = true; | |
| 1096 | |
| 1097 for (i = 0; i < n_traces || (two_passes && current_pass == 1) ; i++) | |
| 1098 { | |
| 1099 int t = i; | |
| 1100 int t2; | |
| 1101 edge e, best; | |
| 1102 int best_len; | |
| 1103 | |
| 1104 if (i >= n_traces) | |
| 1105 { | |
| 1106 gcc_assert (two_passes && current_pass == 1); | |
| 1107 i = 0; | |
| 1108 t = i; | |
| 1109 current_pass = 2; | |
| 1110 if (current_partition == BB_HOT_PARTITION) | |
| 1111 current_partition = BB_COLD_PARTITION; | |
| 1112 else | |
| 1113 current_partition = BB_HOT_PARTITION; | |
| 1114 } | |
| 1115 | |
| 1116 if (connected[t]) | |
| 1117 continue; | |
| 1118 | |
| 1119 if (two_passes | |
| 1120 && BB_PARTITION (traces[t].first) != current_partition) | |
| 1121 continue; | |
| 1122 | |
| 1123 connected[t] = true; | |
| 1124 | |
| 1125 /* Find the predecessor traces. */ | |
| 1126 for (t2 = t; t2 > 0;) | |
| 1127 { | |
| 1128 edge_iterator ei; | |
| 1129 best = NULL; | |
| 1130 best_len = 0; | |
| 1131 FOR_EACH_EDGE (e, ei, traces[t2].first->preds) | |
| 1132 { | |
| 1133 int si = e->src->index; | |
| 1134 | |
| 111 | 1135 if (e->src != ENTRY_BLOCK_PTR_FOR_FN (cfun) |
| 0 | 1136 && (e->flags & EDGE_CAN_FALLTHRU) |
| 1137 && !(e->flags & EDGE_COMPLEX) | |
| 1138 && bbd[si].end_of_trace >= 0 | |
| 1139 && !connected[bbd[si].end_of_trace] | |
| 1140 && (BB_PARTITION (e->src) == current_partition) | |
| 111 | 1141 && connect_better_edge_p (e, true, best_len, best, traces)) |
| 0 | 1142 { |
| 1143 best = e; | |
| 1144 best_len = traces[bbd[si].end_of_trace].length; | |
| 1145 } | |
| 1146 } | |
| 1147 if (best) | |
| 1148 { | |
| 1149 best->src->aux = best->dest; | |
| 1150 t2 = bbd[best->src->index].end_of_trace; | |
| 1151 connected[t2] = true; | |
| 1152 | |
| 1153 if (dump_file) | |
| 1154 { | |
| 1155 fprintf (dump_file, "Connection: %d %d\n", | |
| 1156 best->src->index, best->dest->index); | |
| 1157 } | |
| 1158 } | |
| 1159 else | |
| 1160 break; | |
| 1161 } | |
| 1162 | |
| 1163 if (last_trace >= 0) | |
| 1164 traces[last_trace].last->aux = traces[t2].first; | |
| 1165 last_trace = t; | |
| 1166 | |
| 1167 /* Find the successor traces. */ | |
| 1168 while (1) | |
| 1169 { | |
| 1170 /* Find the continuation of the chain. */ | |
| 1171 edge_iterator ei; | |
| 1172 best = NULL; | |
| 1173 best_len = 0; | |
| 1174 FOR_EACH_EDGE (e, ei, traces[t].last->succs) | |
| 1175 { | |
| 1176 int di = e->dest->index; | |
| 1177 | |
| 111 | 1178 if (e->dest != EXIT_BLOCK_PTR_FOR_FN (cfun) |
| 0 | 1179 && (e->flags & EDGE_CAN_FALLTHRU) |
| 1180 && !(e->flags & EDGE_COMPLEX) | |
| 1181 && bbd[di].start_of_trace >= 0 | |
| 1182 && !connected[bbd[di].start_of_trace] | |
| 1183 && (BB_PARTITION (e->dest) == current_partition) | |
| 111 | 1184 && connect_better_edge_p (e, false, best_len, best, traces)) |
| 0 | 1185 { |
| 1186 best = e; | |
| 1187 best_len = traces[bbd[di].start_of_trace].length; | |
| 1188 } | |
| 1189 } | |
| 1190 | |
| 111 | 1191 if (for_size) |
| 1192 { | |
| 1193 if (!best) | |
| 1194 /* Stop finding the successor traces. */ | |
| 1195 break; | |
| 1196 | |
| 1197 /* It is OK to connect block n with block n + 1 or a block | |
| 1198 before n. For others, only connect to the loop header. */ | |
| 1199 if (best->dest->index > (traces[t].last->index + 1)) | |
| 1200 { | |
| 1201 int count = EDGE_COUNT (best->dest->preds); | |
| 1202 | |
| 1203 FOR_EACH_EDGE (e, ei, best->dest->preds) | |
| 1204 if (e->flags & EDGE_DFS_BACK) | |
| 1205 count--; | |
| 1206 | |
| 1207 /* If dest has multiple predecessors, skip it. We expect | |
| 1208 that one predecessor with smaller index connects with it | |
| 1209 later. */ | |
| 1210 if (count != 1) | |
| 1211 break; | |
| 1212 } | |
| 1213 | |
| 1214 /* Only connect Trace n with Trace n + 1. It is conservative | |
| 1215 to keep the order as close as possible to the original order. | |
| 1216 It also helps to reduce long jumps. */ | |
| 1217 if (last_trace != bbd[best->dest->index].start_of_trace - 1) | |
| 1218 break; | |
| 1219 | |
| 1220 if (dump_file) | |
| 1221 fprintf (dump_file, "Connection: %d %d\n", | |
| 1222 best->src->index, best->dest->index); | |
| 1223 | |
| 1224 t = bbd[best->dest->index].start_of_trace; | |
| 1225 traces[last_trace].last->aux = traces[t].first; | |
| 1226 connected[t] = true; | |
| 1227 last_trace = t; | |
| 1228 } | |
| 1229 else if (best) | |
| 0 | 1230 { |
| 1231 if (dump_file) | |
| 1232 { | |
| 1233 fprintf (dump_file, "Connection: %d %d\n", | |
| 1234 best->src->index, best->dest->index); | |
| 1235 } | |
| 1236 t = bbd[best->dest->index].start_of_trace; | |
| 1237 traces[last_trace].last->aux = traces[t].first; | |
| 1238 connected[t] = true; | |
| 1239 last_trace = t; | |
| 1240 } | |
| 1241 else | |
| 1242 { | |
| 1243 /* Try to connect the traces by duplication of 1 block. */ | |
| 1244 edge e2; | |
| 1245 basic_block next_bb = NULL; | |
| 1246 bool try_copy = false; | |
| 1247 | |
| 1248 FOR_EACH_EDGE (e, ei, traces[t].last->succs) | |
| 111 | 1249 if (e->dest != EXIT_BLOCK_PTR_FOR_FN (cfun) |
| 0 | 1250 && (e->flags & EDGE_CAN_FALLTHRU) |
| 1251 && !(e->flags & EDGE_COMPLEX) | |
| 1252 && (!best || e->probability > best->probability)) | |
| 1253 { | |
| 1254 edge_iterator ei; | |
| 1255 edge best2 = NULL; | |
| 1256 int best2_len = 0; | |
| 1257 | |
| 1258 /* If the destination is a start of a trace which is only | |
| 1259 one block long, then no need to search the successor | |
| 1260 blocks of the trace. Accept it. */ | |
| 1261 if (bbd[e->dest->index].start_of_trace >= 0 | |
| 1262 && traces[bbd[e->dest->index].start_of_trace].length | |
| 1263 == 1) | |
| 1264 { | |
| 1265 best = e; | |
| 1266 try_copy = true; | |
| 1267 continue; | |
| 1268 } | |
| 1269 | |
| 1270 FOR_EACH_EDGE (e2, ei, e->dest->succs) | |
| 1271 { | |
| 1272 int di = e2->dest->index; | |
| 1273 | |
| 111 | 1274 if (e2->dest == EXIT_BLOCK_PTR_FOR_FN (cfun) |
| 0 | 1275 || ((e2->flags & EDGE_CAN_FALLTHRU) |
| 1276 && !(e2->flags & EDGE_COMPLEX) | |
| 1277 && bbd[di].start_of_trace >= 0 | |
| 1278 && !connected[bbd[di].start_of_trace] | |
| 111 | 1279 && BB_PARTITION (e2->dest) == current_partition |
| 1280 && e2->count () >= count_threshold | |
| 0 | 1281 && (!best2 |
| 1282 || e2->probability > best2->probability | |
| 1283 || (e2->probability == best2->probability | |
| 1284 && traces[bbd[di].start_of_trace].length | |
| 1285 > best2_len)))) | |
| 1286 { | |
| 1287 best = e; | |
| 1288 best2 = e2; | |
| 111 | 1289 if (e2->dest != EXIT_BLOCK_PTR_FOR_FN (cfun)) |
| 0 | 1290 best2_len = traces[bbd[di].start_of_trace].length; |
| 1291 else | |
| 1292 best2_len = INT_MAX; | |
| 1293 next_bb = e2->dest; | |
| 1294 try_copy = true; | |
| 1295 } | |
| 1296 } | |
| 1297 } | |
| 1298 | |
| 1299 /* Copy tiny blocks always; copy larger blocks only when the | |
| 1300 edge is traversed frequently enough. */ | |
| 1301 if (try_copy | |
| 111 | 1302 && BB_PARTITION (best->src) == BB_PARTITION (best->dest) |
| 0 | 1303 && copy_bb_p (best->dest, |
| 1304 optimize_edge_for_speed_p (best) | |
| 111 | 1305 && (!best->count ().initialized_p () |
| 1306 || best->count () >= count_threshold))) | |
| 0 | 1307 { |
| 1308 basic_block new_bb; | |
| 1309 | |
| 1310 if (dump_file) | |
| 1311 { | |
| 1312 fprintf (dump_file, "Connection: %d %d ", | |
| 1313 traces[t].last->index, best->dest->index); | |
| 1314 if (!next_bb) | |
| 1315 fputc ('\n', dump_file); | |
| 111 | 1316 else if (next_bb == EXIT_BLOCK_PTR_FOR_FN (cfun)) |
| 0 | 1317 fprintf (dump_file, "exit\n"); |
| 1318 else | |
| 1319 fprintf (dump_file, "%d\n", next_bb->index); | |
| 1320 } | |
| 1321 | |
| 1322 new_bb = copy_bb (best->dest, best, traces[t].last, t); | |
| 1323 traces[t].last = new_bb; | |
| 111 | 1324 if (next_bb && next_bb != EXIT_BLOCK_PTR_FOR_FN (cfun)) |
| 0 | 1325 { |
| 1326 t = bbd[next_bb->index].start_of_trace; | |
| 1327 traces[last_trace].last->aux = traces[t].first; | |
| 1328 connected[t] = true; | |
| 1329 last_trace = t; | |
| 1330 } | |
| 1331 else | |
| 1332 break; /* Stop finding the successor traces. */ | |
| 1333 } | |
| 1334 else | |
| 1335 break; /* Stop finding the successor traces. */ | |
| 1336 } | |
| 1337 } | |
| 1338 } | |
| 1339 | |
| 1340 if (dump_file) | |
| 1341 { | |
| 1342 basic_block bb; | |
| 1343 | |
| 1344 fprintf (dump_file, "Final order:\n"); | |
| 1345 for (bb = traces[0].first; bb; bb = (basic_block) bb->aux) | |
| 1346 fprintf (dump_file, "%d ", bb->index); | |
| 1347 fprintf (dump_file, "\n"); | |
| 1348 fflush (dump_file); | |
| 1349 } | |
| 1350 | |
| 1351 FREE (connected); | |
| 1352 } | |
| 1353 | |
| 1354 /* Return true when BB can and should be copied. CODE_MAY_GROW is true | |
| 1355 when code size is allowed to grow by duplication. */ | |
| 1356 | |
| 1357 static bool | |
| 1358 copy_bb_p (const_basic_block bb, int code_may_grow) | |
| 1359 { | |
| 145 | 1360 unsigned int size = 0; |
| 1361 unsigned int max_size = uncond_jump_length; | |
| 111 | 1362 rtx_insn *insn; |
| 0 | 1363 |
| 1364 if (EDGE_COUNT (bb->preds) < 2) | |
| 1365 return false; | |
| 1366 if (!can_duplicate_block_p (bb)) | |
| 1367 return false; | |
| 1368 | |
| 1369 /* Avoid duplicating blocks which have many successors (PR/13430). */ | |
| 1370 if (EDGE_COUNT (bb->succs) > 8) | |
| 1371 return false; | |
| 1372 | |
| 1373 if (code_may_grow && optimize_bb_for_speed_p (bb)) | |
| 145 | 1374 max_size *= param_max_grow_copy_bb_insns; |
| 0 | 1375 |
| 1376 FOR_BB_INSNS (bb, insn) | |
| 1377 { | |
| 1378 if (INSN_P (insn)) | |
| 145 | 1379 { |
| 1380 size += get_attr_min_length (insn); | |
| 1381 if (size > max_size) | |
| 1382 break; | |
| 1383 } | |
| 0 | 1384 } |
| 1385 | |
| 1386 if (size <= max_size) | |
| 1387 return true; | |
| 1388 | |
| 1389 if (dump_file) | |
| 1390 { | |
| 1391 fprintf (dump_file, | |
| 145 | 1392 "Block %d can't be copied because its size = %u.\n", |
| 0 | 1393 bb->index, size); |
| 1394 } | |
| 1395 | |
| 1396 return false; | |
| 1397 } | |
| 1398 | |
| 1399 /* Return the length of unconditional jump instruction. */ | |
| 1400 | |
| 111 | 1401 int |
| 0 | 1402 get_uncond_jump_length (void) |
| 1403 { | |
| 145 | 1404 unsigned int length; |
| 0 | 1405 |
| 111 | 1406 start_sequence (); |
| 1407 rtx_code_label *label = emit_label (gen_label_rtx ()); | |
| 1408 rtx_insn *jump = emit_jump_insn (targetm.gen_jump (label)); | |
| 0 | 1409 length = get_attr_min_length (jump); |
| 111 | 1410 end_sequence (); |
| 1411 | |
| 145 | 1412 gcc_assert (length < INT_MAX); |
| 0 | 1413 return length; |
| 1414 } | |
| 1415 | |
| 131 | 1416 /* Create a forwarder block to OLD_BB starting with NEW_LABEL and in the |
| 1417 other partition wrt OLD_BB. */ | |
| 1418 | |
| 1419 static basic_block | |
| 1420 create_eh_forwarder_block (rtx_code_label *new_label, basic_block old_bb) | |
| 1421 { | |
| 1422 /* Split OLD_BB, so that EH pads have always only incoming EH edges, | |
| 1423 bb_has_eh_pred bbs are treated specially by DF infrastructure. */ | |
| 1424 old_bb = split_block_after_labels (old_bb)->dest; | |
| 1425 | |
| 1426 /* Put the new label and a jump in the new basic block. */ | |
| 1427 rtx_insn *label = emit_label (new_label); | |
| 1428 rtx_code_label *old_label = block_label (old_bb); | |
| 1429 rtx_insn *jump = emit_jump_insn (targetm.gen_jump (old_label)); | |
| 1430 JUMP_LABEL (jump) = old_label; | |
| 1431 | |
| 1432 /* Create the new basic block and put it in last position. */ | |
| 1433 basic_block last_bb = EXIT_BLOCK_PTR_FOR_FN (cfun)->prev_bb; | |
| 1434 basic_block new_bb = create_basic_block (label, jump, last_bb); | |
| 1435 new_bb->aux = last_bb->aux; | |
| 1436 new_bb->count = old_bb->count; | |
| 1437 last_bb->aux = new_bb; | |
| 1438 | |
| 1439 emit_barrier_after_bb (new_bb); | |
| 1440 | |
| 1441 make_single_succ_edge (new_bb, old_bb, 0); | |
| 1442 | |
| 1443 /* Make sure the new basic block is in the other partition. */ | |
| 1444 unsigned new_partition = BB_PARTITION (old_bb); | |
| 1445 new_partition ^= BB_HOT_PARTITION | BB_COLD_PARTITION; | |
| 1446 BB_SET_PARTITION (new_bb, new_partition); | |
| 1447 | |
| 1448 return new_bb; | |
| 1449 } | |
| 1450 | |
| 1451 /* The common landing pad in block OLD_BB has edges from both partitions. | |
| 1452 Add a new landing pad that will just jump to the old one and split the | |
| 1453 edges so that no EH edge crosses partitions. */ | |
| 111 | 1454 |
| 1455 static void | |
| 131 | 1456 sjlj_fix_up_crossing_landing_pad (basic_block old_bb) |
| 1457 { | |
| 1458 const unsigned lp_len = cfun->eh->lp_array->length (); | |
| 1459 edge_iterator ei; | |
| 1460 edge e; | |
| 1461 | |
| 1462 /* Generate the new common landing-pad label. */ | |
| 1463 rtx_code_label *new_label = gen_label_rtx (); | |
| 1464 LABEL_PRESERVE_P (new_label) = 1; | |
| 1465 | |
| 1466 /* Create the forwarder block. */ | |
| 1467 basic_block new_bb = create_eh_forwarder_block (new_label, old_bb); | |
| 1468 | |
| 1469 /* Create the map from old to new lp index and initialize it. */ | |
| 1470 unsigned *index_map = (unsigned *) alloca (lp_len * sizeof (unsigned)); | |
| 1471 memset (index_map, 0, lp_len * sizeof (unsigned)); | |
| 1472 | |
| 1473 /* Fix up the edges. */ | |
| 1474 for (ei = ei_start (old_bb->preds); (e = ei_safe_edge (ei)) != NULL; ) | |
| 1475 if (e->src != new_bb && BB_PARTITION (e->src) == BB_PARTITION (new_bb)) | |
| 1476 { | |
| 1477 rtx_insn *insn = BB_END (e->src); | |
| 1478 rtx note = find_reg_note (insn, REG_EH_REGION, NULL_RTX); | |
| 1479 | |
| 1480 gcc_assert (note != NULL); | |
| 1481 const unsigned old_index = INTVAL (XEXP (note, 0)); | |
| 1482 | |
| 1483 /* Generate the new landing-pad structure. */ | |
| 1484 if (index_map[old_index] == 0) | |
| 1485 { | |
| 1486 eh_landing_pad old_lp = (*cfun->eh->lp_array)[old_index]; | |
| 1487 eh_landing_pad new_lp = gen_eh_landing_pad (old_lp->region); | |
| 1488 new_lp->post_landing_pad = old_lp->post_landing_pad; | |
| 1489 new_lp->landing_pad = new_label; | |
| 1490 index_map[old_index] = new_lp->index; | |
| 1491 } | |
| 1492 XEXP (note, 0) = GEN_INT (index_map[old_index]); | |
| 1493 | |
| 1494 /* Adjust the edge to the new destination. */ | |
| 1495 redirect_edge_succ (e, new_bb); | |
| 1496 } | |
| 1497 else | |
| 1498 ei_next (&ei); | |
| 1499 } | |
| 1500 | |
| 1501 /* The landing pad OLD_LP, in block OLD_BB, has edges from both partitions. | |
| 1502 Add a new landing pad that will just jump to the old one and split the | |
| 1503 edges so that no EH edge crosses partitions. */ | |
| 1504 | |
| 1505 static void | |
| 1506 dw2_fix_up_crossing_landing_pad (eh_landing_pad old_lp, basic_block old_bb) | |
| 111 | 1507 { |
| 1508 eh_landing_pad new_lp; | |
| 1509 edge_iterator ei; | |
| 1510 edge e; | |
| 1511 | |
| 1512 /* Generate the new landing-pad structure. */ | |
| 1513 new_lp = gen_eh_landing_pad (old_lp->region); | |
| 1514 new_lp->post_landing_pad = old_lp->post_landing_pad; | |
| 1515 new_lp->landing_pad = gen_label_rtx (); | |
| 1516 LABEL_PRESERVE_P (new_lp->landing_pad) = 1; | |
| 1517 | |
| 131 | 1518 /* Create the forwarder block. */ |
| 1519 basic_block new_bb = create_eh_forwarder_block (new_lp->landing_pad, old_bb); | |
| 111 | 1520 |
| 1521 /* Fix up the edges. */ | |
| 1522 for (ei = ei_start (old_bb->preds); (e = ei_safe_edge (ei)) != NULL; ) | |
| 131 | 1523 if (e->src != new_bb && BB_PARTITION (e->src) == BB_PARTITION (new_bb)) |
| 111 | 1524 { |
| 1525 rtx_insn *insn = BB_END (e->src); | |
| 1526 rtx note = find_reg_note (insn, REG_EH_REGION, NULL_RTX); | |
| 1527 | |
| 1528 gcc_assert (note != NULL); | |
| 1529 gcc_checking_assert (INTVAL (XEXP (note, 0)) == old_lp->index); | |
| 1530 XEXP (note, 0) = GEN_INT (new_lp->index); | |
| 1531 | |
| 1532 /* Adjust the edge to the new destination. */ | |
| 1533 redirect_edge_succ (e, new_bb); | |
| 1534 } | |
| 1535 else | |
| 1536 ei_next (&ei); | |
| 1537 } | |
| 1538 | |
| 1539 | |
| 1540 /* Ensure that all hot bbs are included in a hot path through the | |
| 1541 procedure. This is done by calling this function twice, once | |
| 1542 with WALK_UP true (to look for paths from the entry to hot bbs) and | |
| 1543 once with WALK_UP false (to look for paths from hot bbs to the exit). | |
| 1544 Returns the updated value of COLD_BB_COUNT and adds newly-hot bbs | |
| 1545 to BBS_IN_HOT_PARTITION. */ | |
| 1546 | |
| 1547 static unsigned int | |
| 1548 sanitize_hot_paths (bool walk_up, unsigned int cold_bb_count, | |
| 1549 vec<basic_block> *bbs_in_hot_partition) | |
| 1550 { | |
| 1551 /* Callers check this. */ | |
| 1552 gcc_checking_assert (cold_bb_count); | |
| 1553 | |
| 1554 /* Keep examining hot bbs while we still have some left to check | |
| 1555 and there are remaining cold bbs. */ | |
| 1556 vec<basic_block> hot_bbs_to_check = bbs_in_hot_partition->copy (); | |
| 1557 while (! hot_bbs_to_check.is_empty () | |
| 1558 && cold_bb_count) | |
| 1559 { | |
| 1560 basic_block bb = hot_bbs_to_check.pop (); | |
| 1561 vec<edge, va_gc> *edges = walk_up ? bb->preds : bb->succs; | |
| 1562 edge e; | |
| 1563 edge_iterator ei; | |
| 1564 profile_probability highest_probability | |
| 1565 = profile_probability::uninitialized (); | |
| 1566 profile_count highest_count = profile_count::uninitialized (); | |
| 1567 bool found = false; | |
| 1568 | |
| 1569 /* Walk the preds/succs and check if there is at least one already | |
| 1570 marked hot. Keep track of the most frequent pred/succ so that we | |
| 1571 can mark it hot if we don't find one. */ | |
| 1572 FOR_EACH_EDGE (e, ei, edges) | |
| 1573 { | |
| 1574 basic_block reach_bb = walk_up ? e->src : e->dest; | |
| 1575 | |
| 1576 if (e->flags & EDGE_DFS_BACK) | |
| 1577 continue; | |
| 1578 | |
| 1579 /* Do not expect profile insanities when profile was not adjusted. */ | |
| 1580 if (e->probability == profile_probability::never () | |
| 1581 || e->count () == profile_count::zero ()) | |
| 1582 continue; | |
| 1583 | |
| 1584 if (BB_PARTITION (reach_bb) != BB_COLD_PARTITION) | |
| 1585 { | |
| 1586 found = true; | |
| 1587 break; | |
| 1588 } | |
| 1589 /* The following loop will look for the hottest edge via | |
| 131 | 1590 the edge count, if it is non-zero, then fallback to |
| 1591 the edge probability. */ | |
| 1592 if (!(e->count () > highest_count)) | |
| 111 | 1593 highest_count = e->count (); |
| 1594 if (!highest_probability.initialized_p () | |
| 1595 || e->probability > highest_probability) | |
| 1596 highest_probability = e->probability; | |
| 1597 } | |
| 1598 | |
| 1599 /* If bb is reached by (or reaches, in the case of !WALK_UP) another hot | |
| 1600 block (or unpartitioned, e.g. the entry block) then it is ok. If not, | |
| 1601 then the most frequent pred (or succ) needs to be adjusted. In the | |
| 1602 case where multiple preds/succs have the same frequency (e.g. a | |
| 1603 50-50 branch), then both will be adjusted. */ | |
| 1604 if (found) | |
| 1605 continue; | |
| 1606 | |
| 1607 FOR_EACH_EDGE (e, ei, edges) | |
| 1608 { | |
| 1609 if (e->flags & EDGE_DFS_BACK) | |
| 1610 continue; | |
| 1611 /* Do not expect profile insanities when profile was not adjusted. */ | |
| 1612 if (e->probability == profile_probability::never () | |
| 1613 || e->count () == profile_count::zero ()) | |
| 1614 continue; | |
| 1615 /* Select the hottest edge using the edge count, if it is non-zero, | |
| 131 | 1616 then fallback to the edge probability. */ |
| 1617 if (highest_count.initialized_p ()) | |
| 111 | 1618 { |
| 131 | 1619 if (!(e->count () >= highest_count)) |
| 111 | 1620 continue; |
| 1621 } | |
| 131 | 1622 else if (!(e->probability >= highest_probability)) |
| 111 | 1623 continue; |
| 1624 | |
| 1625 basic_block reach_bb = walk_up ? e->src : e->dest; | |
| 1626 | |
| 1627 /* We have a hot bb with an immediate dominator that is cold. | |
| 1628 The dominator needs to be re-marked hot. */ | |
| 1629 BB_SET_PARTITION (reach_bb, BB_HOT_PARTITION); | |
| 1630 if (dump_file) | |
| 1631 fprintf (dump_file, "Promoting bb %i to hot partition to sanitize " | |
| 1632 "profile of bb %i in %s walk\n", reach_bb->index, | |
| 1633 bb->index, walk_up ? "backward" : "forward"); | |
| 1634 cold_bb_count--; | |
| 1635 | |
| 1636 /* Now we need to examine newly-hot reach_bb to see if it is also | |
| 1637 dominated by a cold bb. */ | |
| 1638 bbs_in_hot_partition->safe_push (reach_bb); | |
| 1639 hot_bbs_to_check.safe_push (reach_bb); | |
| 1640 } | |
| 1641 } | |
| 131 | 1642 hot_bbs_to_check.release (); |
| 111 | 1643 |
| 1644 return cold_bb_count; | |
| 1645 } | |
| 1646 | |
| 1647 | |
| 0 | 1648 /* Find the basic blocks that are rarely executed and need to be moved to |
| 1649 a separate section of the .o file (to cut down on paging and improve | |
| 111 | 1650 cache locality). Return a vector of all edges that cross. */ |
| 1651 | |
| 1652 static vec<edge> | |
| 1653 find_rarely_executed_basic_blocks_and_crossing_edges (void) | |
| 0 | 1654 { |
| 111 | 1655 vec<edge> crossing_edges = vNULL; |
| 0 | 1656 basic_block bb; |
| 1657 edge e; | |
| 1658 edge_iterator ei; | |
| 111 | 1659 unsigned int cold_bb_count = 0; |
| 1660 auto_vec<basic_block> bbs_in_hot_partition; | |
| 1661 | |
| 1662 propagate_unlikely_bbs_forward (); | |
| 0 | 1663 |
| 1664 /* Mark which partition (hot/cold) each basic block belongs in. */ | |
| 111 | 1665 FOR_EACH_BB_FN (bb, cfun) |
| 1666 { | |
| 1667 bool cold_bb = false; | |
| 1668 | |
| 1669 if (probably_never_executed_bb_p (cfun, bb)) | |
| 1670 { | |
| 145 | 1671 cold_bb = true; |
| 1672 | |
| 111 | 1673 /* Handle profile insanities created by upstream optimizations |
| 1674 by also checking the incoming edge weights. If there is a non-cold | |
| 1675 incoming edge, conservatively prevent this block from being split | |
| 1676 into the cold section. */ | |
| 145 | 1677 if (!bb->count.precise_p ()) |
| 1678 FOR_EACH_EDGE (e, ei, bb->preds) | |
| 1679 if (!probably_never_executed_edge_p (cfun, e)) | |
| 1680 { | |
| 1681 cold_bb = false; | |
| 1682 break; | |
| 1683 } | |
| 111 | 1684 } |
| 1685 if (cold_bb) | |
| 1686 { | |
| 1687 BB_SET_PARTITION (bb, BB_COLD_PARTITION); | |
| 1688 cold_bb_count++; | |
| 1689 } | |
| 1690 else | |
| 1691 { | |
| 1692 BB_SET_PARTITION (bb, BB_HOT_PARTITION); | |
| 1693 bbs_in_hot_partition.safe_push (bb); | |
| 1694 } | |
| 1695 } | |
| 1696 | |
| 1697 /* Ensure that hot bbs are included along a hot path from the entry to exit. | |
| 1698 Several different possibilities may include cold bbs along all paths | |
| 1699 to/from a hot bb. One is that there are edge weight insanities | |
| 1700 due to optimization phases that do not properly update basic block profile | |
| 1701 counts. The second is that the entry of the function may not be hot, because | |
| 1702 it is entered fewer times than the number of profile training runs, but there | |
| 1703 is a loop inside the function that causes blocks within the function to be | |
| 1704 above the threshold for hotness. This is fixed by walking up from hot bbs | |
| 1705 to the entry block, and then down from hot bbs to the exit, performing | |
| 1706 partitioning fixups as necessary. */ | |
| 1707 if (cold_bb_count) | |
| 0 | 1708 { |
| 111 | 1709 mark_dfs_back_edges (); |
| 1710 cold_bb_count = sanitize_hot_paths (true, cold_bb_count, | |
| 1711 &bbs_in_hot_partition); | |
| 1712 if (cold_bb_count) | |
| 1713 sanitize_hot_paths (false, cold_bb_count, &bbs_in_hot_partition); | |
| 1714 | |
| 1715 hash_set <basic_block> set; | |
| 1716 find_bbs_reachable_by_hot_paths (&set); | |
| 1717 FOR_EACH_BB_FN (bb, cfun) | |
| 1718 if (!set.contains (bb)) | |
| 1719 BB_SET_PARTITION (bb, BB_COLD_PARTITION); | |
| 1720 } | |
| 1721 | |
| 1722 /* The format of .gcc_except_table does not allow landing pads to | |
| 1723 be in a different partition as the throw. Fix this by either | |
| 131 | 1724 moving the landing pads or inserting forwarder landing pads. */ |
| 111 | 1725 if (cfun->eh->lp_array) |
| 1726 { | |
| 131 | 1727 const bool sjlj |
| 1728 = (targetm_common.except_unwind_info (&global_options) == UI_SJLJ); | |
| 111 | 1729 unsigned i; |
| 1730 eh_landing_pad lp; | |
| 1731 | |
| 1732 FOR_EACH_VEC_ELT (*cfun->eh->lp_array, i, lp) | |
| 1733 { | |
| 1734 bool all_same, all_diff; | |
| 1735 | |
| 1736 if (lp == NULL | |
| 1737 || lp->landing_pad == NULL_RTX | |
| 1738 || !LABEL_P (lp->landing_pad)) | |
| 1739 continue; | |
| 1740 | |
| 1741 all_same = all_diff = true; | |
| 1742 bb = BLOCK_FOR_INSN (lp->landing_pad); | |
| 1743 FOR_EACH_EDGE (e, ei, bb->preds) | |
| 1744 { | |
| 1745 gcc_assert (e->flags & EDGE_EH); | |
| 1746 if (BB_PARTITION (bb) == BB_PARTITION (e->src)) | |
| 1747 all_diff = false; | |
| 1748 else | |
| 1749 all_same = false; | |
| 1750 } | |
| 1751 | |
| 1752 if (all_same) | |
| 1753 ; | |
| 1754 else if (all_diff) | |
| 1755 { | |
| 1756 int which = BB_PARTITION (bb); | |
| 1757 which ^= BB_HOT_PARTITION | BB_COLD_PARTITION; | |
| 1758 BB_SET_PARTITION (bb, which); | |
| 1759 } | |
| 131 | 1760 else if (sjlj) |
| 1761 sjlj_fix_up_crossing_landing_pad (bb); | |
| 111 | 1762 else |
| 131 | 1763 dw2_fix_up_crossing_landing_pad (lp, bb); |
| 1764 | |
| 1765 /* There is a single, common landing pad in SJLJ mode. */ | |
| 1766 if (sjlj) | |
| 1767 break; | |
| 111 | 1768 } |
| 0 | 1769 } |
| 1770 | |
| 1771 /* Mark every edge that crosses between sections. */ | |
| 111 | 1772 FOR_EACH_BB_FN (bb, cfun) |
| 0 | 1773 FOR_EACH_EDGE (e, ei, bb->succs) |
| 111 | 1774 { |
| 1775 unsigned int flags = e->flags; | |
| 1776 | |
| 1777 /* We should never have EDGE_CROSSING set yet. */ | |
| 1778 gcc_checking_assert ((flags & EDGE_CROSSING) == 0); | |
| 1779 | |
| 1780 if (e->src != ENTRY_BLOCK_PTR_FOR_FN (cfun) | |
| 1781 && e->dest != EXIT_BLOCK_PTR_FOR_FN (cfun) | |
| 1782 && BB_PARTITION (e->src) != BB_PARTITION (e->dest)) | |
| 1783 { | |
| 1784 crossing_edges.safe_push (e); | |
| 1785 flags |= EDGE_CROSSING; | |
| 1786 } | |
| 1787 | |
| 1788 /* Now that we've split eh edges as appropriate, allow landing pads | |
| 1789 to be merged with the post-landing pads. */ | |
| 1790 flags &= ~EDGE_PRESERVE; | |
| 1791 | |
| 1792 e->flags = flags; | |
| 1793 } | |
| 1794 | |
| 1795 return crossing_edges; | |
| 1796 } | |
| 1797 | |
| 1798 /* Set the flag EDGE_CAN_FALLTHRU for edges that can be fallthru. */ | |
| 1799 | |
| 1800 static void | |
| 1801 set_edge_can_fallthru_flag (void) | |
| 1802 { | |
| 1803 basic_block bb; | |
| 1804 | |
| 1805 FOR_EACH_BB_FN (bb, cfun) | |
| 0 | 1806 { |
| 111 | 1807 edge e; |
| 1808 edge_iterator ei; | |
| 1809 | |
| 1810 FOR_EACH_EDGE (e, ei, bb->succs) | |
| 0 | 1811 { |
| 111 | 1812 e->flags &= ~EDGE_CAN_FALLTHRU; |
| 1813 | |
| 1814 /* The FALLTHRU edge is also CAN_FALLTHRU edge. */ | |
| 1815 if (e->flags & EDGE_FALLTHRU) | |
| 1816 e->flags |= EDGE_CAN_FALLTHRU; | |
| 0 | 1817 } |
| 111 | 1818 |
| 1819 /* If the BB ends with an invertible condjump all (2) edges are | |
| 1820 CAN_FALLTHRU edges. */ | |
| 1821 if (EDGE_COUNT (bb->succs) != 2) | |
| 1822 continue; | |
| 1823 if (!any_condjump_p (BB_END (bb))) | |
| 1824 continue; | |
| 1825 | |
| 1826 rtx_jump_insn *bb_end_jump = as_a <rtx_jump_insn *> (BB_END (bb)); | |
| 1827 if (!invert_jump (bb_end_jump, JUMP_LABEL (bb_end_jump), 0)) | |
| 1828 continue; | |
| 1829 invert_jump (bb_end_jump, JUMP_LABEL (bb_end_jump), 0); | |
| 1830 EDGE_SUCC (bb, 0)->flags |= EDGE_CAN_FALLTHRU; | |
| 1831 EDGE_SUCC (bb, 1)->flags |= EDGE_CAN_FALLTHRU; | |
| 0 | 1832 } |
| 1833 } | |
| 1834 | |
| 1835 /* If any destination of a crossing edge does not have a label, add label; | |
| 111 | 1836 Convert any easy fall-through crossing edges to unconditional jumps. */ |
| 0 | 1837 |
| 1838 static void | |
| 111 | 1839 add_labels_and_missing_jumps (vec<edge> crossing_edges) |
| 0 | 1840 { |
| 111 | 1841 size_t i; |
| 1842 edge e; | |
| 1843 | |
| 1844 FOR_EACH_VEC_ELT (crossing_edges, i, e) | |
| 0 | 1845 { |
| 111 | 1846 basic_block src = e->src; |
| 1847 basic_block dest = e->dest; | |
| 1848 rtx_jump_insn *new_jump; | |
| 1849 | |
| 1850 if (dest == EXIT_BLOCK_PTR_FOR_FN (cfun)) | |
| 1851 continue; | |
| 1852 | |
| 1853 /* Make sure dest has a label. */ | |
| 1854 rtx_code_label *label = block_label (dest); | |
| 1855 | |
| 1856 /* Nothing to do for non-fallthru edges. */ | |
| 1857 if (src == ENTRY_BLOCK_PTR_FOR_FN (cfun)) | |
| 1858 continue; | |
| 1859 if ((e->flags & EDGE_FALLTHRU) == 0) | |
| 1860 continue; | |
| 1861 | |
| 1862 /* If the block does not end with a control flow insn, then we | |
| 1863 can trivially add a jump to the end to fixup the crossing. | |
| 1864 Otherwise the jump will have to go in a new bb, which will | |
| 1865 be handled by fix_up_fall_thru_edges function. */ | |
| 1866 if (control_flow_insn_p (BB_END (src))) | |
| 1867 continue; | |
| 1868 | |
| 1869 /* Make sure there's only one successor. */ | |
| 1870 gcc_assert (single_succ_p (src)); | |
| 1871 | |
| 1872 new_jump = emit_jump_insn_after (targetm.gen_jump (label), BB_END (src)); | |
| 1873 BB_END (src) = new_jump; | |
| 1874 JUMP_LABEL (new_jump) = label; | |
| 1875 LABEL_NUSES (label) += 1; | |
| 1876 | |
| 1877 emit_barrier_after_bb (src); | |
| 1878 | |
| 1879 /* Mark edge as non-fallthru. */ | |
| 1880 e->flags &= ~EDGE_FALLTHRU; | |
| 1881 } | |
| 0 | 1882 } |
| 1883 | |
| 1884 /* Find any bb's where the fall-through edge is a crossing edge (note that | |
| 1885 these bb's must also contain a conditional jump or end with a call | |
| 1886 instruction; we've already dealt with fall-through edges for blocks | |
| 1887 that didn't have a conditional jump or didn't end with call instruction | |
| 1888 in the call to add_labels_and_missing_jumps). Convert the fall-through | |
| 1889 edge to non-crossing edge by inserting a new bb to fall-through into. | |
| 1890 The new bb will contain an unconditional jump (crossing edge) to the | |
| 1891 original fall through destination. */ | |
| 1892 | |
| 1893 static void | |
| 1894 fix_up_fall_thru_edges (void) | |
| 1895 { | |
| 1896 basic_block cur_bb; | |
| 111 | 1897 |
| 1898 FOR_EACH_BB_FN (cur_bb, cfun) | |
| 0 | 1899 { |
| 111 | 1900 edge succ1; |
| 1901 edge succ2; | |
| 1902 edge fall_thru = NULL; | |
| 1903 edge cond_jump = NULL; | |
| 1904 | |
| 0 | 1905 fall_thru = NULL; |
| 1906 if (EDGE_COUNT (cur_bb->succs) > 0) | |
| 1907 succ1 = EDGE_SUCC (cur_bb, 0); | |
| 1908 else | |
| 1909 succ1 = NULL; | |
| 1910 | |
| 1911 if (EDGE_COUNT (cur_bb->succs) > 1) | |
| 1912 succ2 = EDGE_SUCC (cur_bb, 1); | |
| 1913 else | |
| 1914 succ2 = NULL; | |
| 1915 | |
| 1916 /* Find the fall-through edge. */ | |
| 1917 | |
| 1918 if (succ1 | |
| 1919 && (succ1->flags & EDGE_FALLTHRU)) | |
| 1920 { | |
| 1921 fall_thru = succ1; | |
| 1922 cond_jump = succ2; | |
| 1923 } | |
| 1924 else if (succ2 | |
| 1925 && (succ2->flags & EDGE_FALLTHRU)) | |
| 1926 { | |
| 1927 fall_thru = succ2; | |
| 1928 cond_jump = succ1; | |
| 1929 } | |
| 111 | 1930 else if (succ2 && EDGE_COUNT (cur_bb->succs) > 2) |
| 1931 fall_thru = find_fallthru_edge (cur_bb->succs); | |
| 1932 | |
| 1933 if (fall_thru && (fall_thru->dest != EXIT_BLOCK_PTR_FOR_FN (cfun))) | |
| 0 | 1934 { |
| 1935 /* Check to see if the fall-thru edge is a crossing edge. */ | |
| 1936 | |
| 1937 if (fall_thru->flags & EDGE_CROSSING) | |
| 1938 { | |
| 1939 /* The fall_thru edge crosses; now check the cond jump edge, if | |
| 1940 it exists. */ | |
| 1941 | |
| 111 | 1942 bool cond_jump_crosses = true; |
| 1943 int invert_worked = 0; | |
| 1944 rtx_insn *old_jump = BB_END (cur_bb); | |
| 0 | 1945 |
| 1946 /* Find the jump instruction, if there is one. */ | |
| 1947 | |
| 1948 if (cond_jump) | |
| 1949 { | |
| 1950 if (!(cond_jump->flags & EDGE_CROSSING)) | |
| 1951 cond_jump_crosses = false; | |
| 1952 | |
| 1953 /* We know the fall-thru edge crosses; if the cond | |
| 1954 jump edge does NOT cross, and its destination is the | |
| 1955 next block in the bb order, invert the jump | |
| 111 | 1956 (i.e. fix it so the fall through does not cross and |
| 0 | 1957 the cond jump does). */ |
| 1958 | |
| 111 | 1959 if (!cond_jump_crosses) |
| 0 | 1960 { |
| 1961 /* Find label in fall_thru block. We've already added | |
| 1962 any missing labels, so there must be one. */ | |
| 1963 | |
| 111 | 1964 rtx_code_label *fall_thru_label |
| 1965 = block_label (fall_thru->dest); | |
| 1966 | |
| 1967 if (old_jump && fall_thru_label) | |
| 1968 { | |
| 1969 rtx_jump_insn *old_jump_insn | |
| 1970 = dyn_cast <rtx_jump_insn *> (old_jump); | |
| 1971 if (old_jump_insn) | |
| 1972 invert_worked = invert_jump (old_jump_insn, | |
| 1973 fall_thru_label, 0); | |
| 1974 } | |
| 1975 | |
| 0 | 1976 if (invert_worked) |
| 1977 { | |
| 1978 fall_thru->flags &= ~EDGE_FALLTHRU; | |
| 1979 cond_jump->flags |= EDGE_FALLTHRU; | |
| 1980 update_br_prob_note (cur_bb); | |
| 111 | 1981 std::swap (fall_thru, cond_jump); |
| 0 | 1982 cond_jump->flags |= EDGE_CROSSING; |
| 1983 fall_thru->flags &= ~EDGE_CROSSING; | |
| 1984 } | |
| 1985 } | |
| 1986 } | |
| 1987 | |
| 1988 if (cond_jump_crosses || !invert_worked) | |
| 1989 { | |
| 1990 /* This is the case where both edges out of the basic | |
| 1991 block are crossing edges. Here we will fix up the | |
| 1992 fall through edge. The jump edge will be taken care | |
|
55
77e2b8dfacca
update it from 4.4.3 to 4.5.0
ryoma <e075725@ie.u-ryukyu.ac.jp>
parents:
0
diff
changeset
|
1993 of later. The EDGE_CROSSING flag of fall_thru edge |
| 111 | 1994 is unset before the call to force_nonfallthru |
| 1995 function because if a new basic-block is created | |
| 1996 this edge remains in the current section boundary | |
| 1997 while the edge between new_bb and the fall_thru->dest | |
| 1998 becomes EDGE_CROSSING. */ | |
| 1999 | |
| 2000 fall_thru->flags &= ~EDGE_CROSSING; | |
| 2001 basic_block new_bb = force_nonfallthru (fall_thru); | |
| 0 | 2002 |
| 2003 if (new_bb) | |
| 2004 { | |
| 2005 new_bb->aux = cur_bb->aux; | |
| 2006 cur_bb->aux = new_bb; | |
| 2007 | |
| 111 | 2008 /* This is done by force_nonfallthru_and_redirect. */ |
| 2009 gcc_assert (BB_PARTITION (new_bb) | |
| 2010 == BB_PARTITION (cur_bb)); | |
| 2011 | |
| 0 | 2012 single_succ_edge (new_bb)->flags |= EDGE_CROSSING; |
| 2013 } | |
| 2014 else | |
| 2015 { | |
| 111 | 2016 /* If a new basic-block was not created; restore |
| 2017 the EDGE_CROSSING flag. */ | |
| 2018 fall_thru->flags |= EDGE_CROSSING; | |
| 0 | 2019 } |
| 111 | 2020 |
| 2021 /* Add barrier after new jump */ | |
| 2022 emit_barrier_after_bb (new_bb ? new_bb : cur_bb); | |
| 0 | 2023 } |
| 2024 } | |
| 2025 } | |
| 2026 } | |
| 2027 } | |
| 2028 | |
| 2029 /* This function checks the destination block of a "crossing jump" to | |
| 2030 see if it has any crossing predecessors that begin with a code label | |
| 2031 and end with an unconditional jump. If so, it returns that predecessor | |
| 2032 block. (This is to avoid creating lots of new basic blocks that all | |
| 2033 contain unconditional jumps to the same destination). */ | |
| 2034 | |
| 2035 static basic_block | |
| 2036 find_jump_block (basic_block jump_dest) | |
| 2037 { | |
| 2038 basic_block source_bb = NULL; | |
| 2039 edge e; | |
| 111 | 2040 rtx_insn *insn; |
| 0 | 2041 edge_iterator ei; |
| 2042 | |
| 2043 FOR_EACH_EDGE (e, ei, jump_dest->preds) | |
| 2044 if (e->flags & EDGE_CROSSING) | |
| 2045 { | |
| 2046 basic_block src = e->src; | |
| 2047 | |
| 2048 /* Check each predecessor to see if it has a label, and contains | |
| 2049 only one executable instruction, which is an unconditional jump. | |
| 2050 If so, we can use it. */ | |
| 2051 | |
| 2052 if (LABEL_P (BB_HEAD (src))) | |
| 2053 for (insn = BB_HEAD (src); | |
| 2054 !INSN_P (insn) && insn != NEXT_INSN (BB_END (src)); | |
| 2055 insn = NEXT_INSN (insn)) | |
| 2056 { | |
| 2057 if (INSN_P (insn) | |
| 2058 && insn == BB_END (src) | |
| 2059 && JUMP_P (insn) | |
| 2060 && !any_condjump_p (insn)) | |
| 2061 { | |
| 2062 source_bb = src; | |
| 2063 break; | |
| 2064 } | |
| 2065 } | |
| 2066 | |
| 2067 if (source_bb) | |
| 2068 break; | |
| 2069 } | |
| 2070 | |
| 2071 return source_bb; | |
| 2072 } | |
| 2073 | |
| 2074 /* Find all BB's with conditional jumps that are crossing edges; | |
| 2075 insert a new bb and make the conditional jump branch to the new | |
| 2076 bb instead (make the new bb same color so conditional branch won't | |
| 2077 be a 'crossing' edge). Insert an unconditional jump from the | |
| 2078 new bb to the original destination of the conditional jump. */ | |
| 2079 | |
| 2080 static void | |
| 2081 fix_crossing_conditional_branches (void) | |
| 2082 { | |
| 2083 basic_block cur_bb; | |
| 2084 basic_block new_bb; | |
| 2085 basic_block dest; | |
| 2086 edge succ1; | |
| 2087 edge succ2; | |
| 2088 edge crossing_edge; | |
| 2089 edge new_edge; | |
| 2090 rtx set_src; | |
| 2091 rtx old_label = NULL_RTX; | |
| 111 | 2092 rtx_code_label *new_label; |
| 2093 | |
| 2094 FOR_EACH_BB_FN (cur_bb, cfun) | |
| 0 | 2095 { |
| 2096 crossing_edge = NULL; | |
| 2097 if (EDGE_COUNT (cur_bb->succs) > 0) | |
| 2098 succ1 = EDGE_SUCC (cur_bb, 0); | |
| 2099 else | |
| 2100 succ1 = NULL; | |
| 2101 | |
| 2102 if (EDGE_COUNT (cur_bb->succs) > 1) | |
| 2103 succ2 = EDGE_SUCC (cur_bb, 1); | |
| 2104 else | |
| 2105 succ2 = NULL; | |
| 2106 | |
| 2107 /* We already took care of fall-through edges, so only one successor | |
| 2108 can be a crossing edge. */ | |
| 2109 | |
| 2110 if (succ1 && (succ1->flags & EDGE_CROSSING)) | |
| 2111 crossing_edge = succ1; | |
| 2112 else if (succ2 && (succ2->flags & EDGE_CROSSING)) | |
| 2113 crossing_edge = succ2; | |
| 2114 | |
| 2115 if (crossing_edge) | |
| 2116 { | |
| 111 | 2117 rtx_insn *old_jump = BB_END (cur_bb); |
| 0 | 2118 |
| 2119 /* Check to make sure the jump instruction is a | |
| 2120 conditional jump. */ | |
| 2121 | |
| 2122 set_src = NULL_RTX; | |
| 2123 | |
| 2124 if (any_condjump_p (old_jump)) | |
| 2125 { | |
| 2126 if (GET_CODE (PATTERN (old_jump)) == SET) | |
| 2127 set_src = SET_SRC (PATTERN (old_jump)); | |
| 2128 else if (GET_CODE (PATTERN (old_jump)) == PARALLEL) | |
| 2129 { | |
| 2130 set_src = XVECEXP (PATTERN (old_jump), 0,0); | |
| 2131 if (GET_CODE (set_src) == SET) | |
| 2132 set_src = SET_SRC (set_src); | |
| 2133 else | |
| 2134 set_src = NULL_RTX; | |
| 2135 } | |
| 2136 } | |
| 2137 | |
| 2138 if (set_src && (GET_CODE (set_src) == IF_THEN_ELSE)) | |
| 2139 { | |
| 111 | 2140 rtx_jump_insn *old_jump_insn = |
| 2141 as_a <rtx_jump_insn *> (old_jump); | |
| 2142 | |
| 0 | 2143 if (GET_CODE (XEXP (set_src, 1)) == PC) |
| 2144 old_label = XEXP (set_src, 2); | |
| 2145 else if (GET_CODE (XEXP (set_src, 2)) == PC) | |
| 2146 old_label = XEXP (set_src, 1); | |
| 2147 | |
| 2148 /* Check to see if new bb for jumping to that dest has | |
| 2149 already been created; if so, use it; if not, create | |
| 2150 a new one. */ | |
| 2151 | |
| 2152 new_bb = find_jump_block (crossing_edge->dest); | |
| 2153 | |
| 2154 if (new_bb) | |
| 2155 new_label = block_label (new_bb); | |
| 2156 else | |
| 2157 { | |
| 111 | 2158 basic_block last_bb; |
| 2159 rtx_code_label *old_jump_target; | |
| 2160 rtx_jump_insn *new_jump; | |
| 2161 | |
| 0 | 2162 /* Create new basic block to be dest for |
| 2163 conditional jump. */ | |
| 2164 | |
| 2165 /* Put appropriate instructions in new bb. */ | |
| 2166 | |
| 2167 new_label = gen_label_rtx (); | |
| 111 | 2168 emit_label (new_label); |
| 2169 | |
| 2170 gcc_assert (GET_CODE (old_label) == LABEL_REF); | |
| 2171 old_jump_target = old_jump_insn->jump_target (); | |
| 2172 new_jump = as_a <rtx_jump_insn *> | |
| 2173 (emit_jump_insn (targetm.gen_jump (old_jump_target))); | |
| 2174 new_jump->set_jump_target (old_jump_target); | |
| 2175 | |
| 2176 last_bb = EXIT_BLOCK_PTR_FOR_FN (cfun)->prev_bb; | |
| 2177 new_bb = create_basic_block (new_label, new_jump, last_bb); | |
| 2178 new_bb->aux = last_bb->aux; | |
| 2179 last_bb->aux = new_bb; | |
| 2180 | |
| 2181 emit_barrier_after_bb (new_bb); | |
| 0 | 2182 |
| 2183 /* Make sure new bb is in same partition as source | |
| 2184 of conditional branch. */ | |
| 2185 BB_COPY_PARTITION (new_bb, cur_bb); | |
| 2186 } | |
| 2187 | |
| 2188 /* Make old jump branch to new bb. */ | |
| 2189 | |
| 111 | 2190 redirect_jump (old_jump_insn, new_label, 0); |
| 0 | 2191 |
| 2192 /* Remove crossing_edge as predecessor of 'dest'. */ | |
| 2193 | |
| 2194 dest = crossing_edge->dest; | |
| 2195 | |
| 2196 redirect_edge_succ (crossing_edge, new_bb); | |
| 2197 | |
| 2198 /* Make a new edge from new_bb to old dest; new edge | |
| 2199 will be a successor for new_bb and a predecessor | |
| 2200 for 'dest'. */ | |
| 2201 | |
| 2202 if (EDGE_COUNT (new_bb->succs) == 0) | |
| 111 | 2203 new_edge = make_single_succ_edge (new_bb, dest, 0); |
| 0 | 2204 else |
| 2205 new_edge = EDGE_SUCC (new_bb, 0); | |
| 2206 | |
| 2207 crossing_edge->flags &= ~EDGE_CROSSING; | |
| 2208 new_edge->flags |= EDGE_CROSSING; | |
| 2209 } | |
| 2210 } | |
| 2211 } | |
| 2212 } | |
| 2213 | |
| 2214 /* Find any unconditional branches that cross between hot and cold | |
| 2215 sections. Convert them into indirect jumps instead. */ | |
| 2216 | |
| 2217 static void | |
| 2218 fix_crossing_unconditional_branches (void) | |
| 2219 { | |
| 2220 basic_block cur_bb; | |
| 111 | 2221 rtx_insn *last_insn; |
| 0 | 2222 rtx label; |
| 2223 rtx label_addr; | |
| 111 | 2224 rtx_insn *indirect_jump_sequence; |
| 2225 rtx_insn *jump_insn = NULL; | |
| 0 | 2226 rtx new_reg; |
| 111 | 2227 rtx_insn *cur_insn; |
| 0 | 2228 edge succ; |
| 2229 | |
| 111 | 2230 FOR_EACH_BB_FN (cur_bb, cfun) |
| 0 | 2231 { |
| 2232 last_insn = BB_END (cur_bb); | |
| 2233 | |
| 2234 if (EDGE_COUNT (cur_bb->succs) < 1) | |
| 2235 continue; | |
| 2236 | |
| 2237 succ = EDGE_SUCC (cur_bb, 0); | |
| 2238 | |
| 2239 /* Check to see if bb ends in a crossing (unconditional) jump. At | |
| 2240 this point, no crossing jumps should be conditional. */ | |
| 2241 | |
| 2242 if (JUMP_P (last_insn) | |
| 2243 && (succ->flags & EDGE_CROSSING)) | |
| 2244 { | |
| 2245 gcc_assert (!any_condjump_p (last_insn)); | |
| 2246 | |
| 2247 /* Make sure the jump is not already an indirect or table jump. */ | |
| 2248 | |
| 2249 if (!computed_jump_p (last_insn) | |
| 111 | 2250 && !tablejump_p (last_insn, NULL, NULL)) |
| 0 | 2251 { |
| 2252 /* We have found a "crossing" unconditional branch. Now | |
| 2253 we must convert it to an indirect jump. First create | |
| 2254 reference of label, as target for jump. */ | |
| 2255 | |
| 2256 label = JUMP_LABEL (last_insn); | |
| 2257 label_addr = gen_rtx_LABEL_REF (Pmode, label); | |
| 2258 LABEL_NUSES (label) += 1; | |
| 2259 | |
| 2260 /* Get a register to use for the indirect jump. */ | |
| 2261 | |
| 2262 new_reg = gen_reg_rtx (Pmode); | |
| 2263 | |
| 2264 /* Generate indirect the jump sequence. */ | |
| 2265 | |
| 2266 start_sequence (); | |
| 2267 emit_move_insn (new_reg, label_addr); | |
| 2268 emit_indirect_jump (new_reg); | |
| 2269 indirect_jump_sequence = get_insns (); | |
| 2270 end_sequence (); | |
| 2271 | |
| 2272 /* Make sure every instruction in the new jump sequence has | |
| 2273 its basic block set to be cur_bb. */ | |
| 2274 | |
| 2275 for (cur_insn = indirect_jump_sequence; cur_insn; | |
| 2276 cur_insn = NEXT_INSN (cur_insn)) | |
| 2277 { | |
| 2278 if (!BARRIER_P (cur_insn)) | |
| 2279 BLOCK_FOR_INSN (cur_insn) = cur_bb; | |
| 2280 if (JUMP_P (cur_insn)) | |
| 2281 jump_insn = cur_insn; | |
| 2282 } | |
| 2283 | |
| 2284 /* Insert the new (indirect) jump sequence immediately before | |
| 2285 the unconditional jump, then delete the unconditional jump. */ | |
| 2286 | |
| 2287 emit_insn_before (indirect_jump_sequence, last_insn); | |
| 2288 delete_insn (last_insn); | |
| 2289 | |
| 111 | 2290 JUMP_LABEL (jump_insn) = label; |
| 2291 LABEL_NUSES (label)++; | |
| 2292 | |
| 0 | 2293 /* Make BB_END for cur_bb be the jump instruction (NOT the |
| 2294 barrier instruction at the end of the sequence...). */ | |
| 2295 | |
| 2296 BB_END (cur_bb) = jump_insn; | |
| 2297 } | |
| 2298 } | |
| 2299 } | |
| 2300 } | |
| 2301 | |
| 111 | 2302 /* Update CROSSING_JUMP_P flags on all jump insns. */ |
| 0 | 2303 |
| 2304 static void | |
| 111 | 2305 update_crossing_jump_flags (void) |
| 0 | 2306 { |
| 2307 basic_block bb; | |
| 2308 edge e; | |
| 2309 edge_iterator ei; | |
| 2310 | |
| 111 | 2311 FOR_EACH_BB_FN (bb, cfun) |
| 0 | 2312 FOR_EACH_EDGE (e, ei, bb->succs) |
| 111 | 2313 if (e->flags & EDGE_CROSSING) |
| 2314 { | |
| 131 | 2315 if (JUMP_P (BB_END (bb))) |
| 111 | 2316 CROSSING_JUMP_P (BB_END (bb)) = 1; |
| 2317 break; | |
| 2318 } | |
| 0 | 2319 } |
| 2320 | |
| 111 | 2321 /* Reorder basic blocks using the software trace cache (STC) algorithm. */ |
| 0 | 2322 |
| 2323 static void | |
| 111 | 2324 reorder_basic_blocks_software_trace_cache (void) |
| 0 | 2325 { |
| 111 | 2326 if (dump_file) |
| 2327 fprintf (dump_file, "\nReordering with the STC algorithm.\n\n"); | |
| 2328 | |
| 0 | 2329 int n_traces; |
| 2330 int i; | |
| 2331 struct trace *traces; | |
| 2332 | |
| 2333 /* We are estimating the length of uncond jump insn only once since the code | |
| 2334 for getting the insn length always returns the minimal length now. */ | |
| 2335 if (uncond_jump_length == 0) | |
| 2336 uncond_jump_length = get_uncond_jump_length (); | |
| 2337 | |
| 2338 /* We need to know some information for each basic block. */ | |
| 111 | 2339 array_size = GET_ARRAY_SIZE (last_basic_block_for_fn (cfun)); |
| 0 | 2340 bbd = XNEWVEC (bbro_basic_block_data, array_size); |
| 2341 for (i = 0; i < array_size; i++) | |
| 2342 { | |
| 2343 bbd[i].start_of_trace = -1; | |
| 111 | 2344 bbd[i].end_of_trace = -1; |
| 0 | 2345 bbd[i].in_trace = -1; |
| 111 | 2346 bbd[i].visited = 0; |
| 2347 bbd[i].priority = -1; | |
| 0 | 2348 bbd[i].heap = NULL; |
| 2349 bbd[i].node = NULL; | |
| 2350 } | |
| 2351 | |
| 111 | 2352 traces = XNEWVEC (struct trace, n_basic_blocks_for_fn (cfun)); |
| 0 | 2353 n_traces = 0; |
| 2354 find_traces (&n_traces, traces); | |
| 2355 connect_traces (n_traces, traces); | |
| 2356 FREE (traces); | |
| 2357 FREE (bbd); | |
| 111 | 2358 } |
| 2359 | |
| 131 | 2360 /* Order edges by execution frequency, higher first. */ |
| 2361 | |
| 2362 static int | |
| 2363 edge_order (const void *ve1, const void *ve2) | |
| 111 | 2364 { |
| 131 | 2365 edge e1 = *(const edge *) ve1; |
| 2366 edge e2 = *(const edge *) ve2; | |
| 2367 profile_count c1 = e1->count (); | |
| 2368 profile_count c2 = e2->count (); | |
| 2369 /* Since profile_count::operator< does not establish a strict weak order | |
| 2370 in presence of uninitialized counts, use 'max': this makes them appear | |
| 2371 as if having execution frequency less than any initialized count. */ | |
| 2372 profile_count m = c1.max (c2); | |
| 2373 return (m == c2) - (m == c1); | |
| 111 | 2374 } |
| 2375 | |
| 2376 /* Reorder basic blocks using the "simple" algorithm. This tries to | |
| 2377 maximize the dynamic number of branches that are fallthrough, without | |
| 2378 copying instructions. The algorithm is greedy, looking at the most | |
| 2379 frequently executed branch first. */ | |
| 2380 | |
| 2381 static void | |
| 2382 reorder_basic_blocks_simple (void) | |
| 2383 { | |
| 2384 if (dump_file) | |
| 2385 fprintf (dump_file, "\nReordering with the \"simple\" algorithm.\n\n"); | |
| 2386 | |
| 2387 edge *edges = new edge[2 * n_basic_blocks_for_fn (cfun)]; | |
| 2388 | |
| 2389 /* First, collect all edges that can be optimized by reordering blocks: | |
| 2390 simple jumps and conditional jumps, as well as the function entry edge. */ | |
| 2391 | |
| 2392 int n = 0; | |
| 2393 edges[n++] = EDGE_SUCC (ENTRY_BLOCK_PTR_FOR_FN (cfun), 0); | |
| 2394 | |
| 2395 basic_block bb; | |
| 2396 FOR_EACH_BB_FN (bb, cfun) | |
| 2397 { | |
| 2398 rtx_insn *end = BB_END (bb); | |
| 2399 | |
| 2400 if (computed_jump_p (end) || tablejump_p (end, NULL, NULL)) | |
| 2401 continue; | |
| 2402 | |
| 2403 /* We cannot optimize asm goto. */ | |
| 2404 if (JUMP_P (end) && extract_asm_operands (end)) | |
| 2405 continue; | |
| 2406 | |
| 2407 if (single_succ_p (bb)) | |
| 2408 edges[n++] = EDGE_SUCC (bb, 0); | |
| 2409 else if (any_condjump_p (end)) | |
| 2410 { | |
| 2411 edge e0 = EDGE_SUCC (bb, 0); | |
| 2412 edge e1 = EDGE_SUCC (bb, 1); | |
| 2413 /* When optimizing for size it is best to keep the original | |
| 2414 fallthrough edges. */ | |
| 2415 if (e1->flags & EDGE_FALLTHRU) | |
| 2416 std::swap (e0, e1); | |
| 2417 edges[n++] = e0; | |
| 2418 edges[n++] = e1; | |
| 2419 } | |
| 2420 } | |
| 2421 | |
| 2422 /* Sort the edges, the most desirable first. When optimizing for size | |
| 2423 all edges are equally desirable. */ | |
| 2424 | |
| 2425 if (optimize_function_for_speed_p (cfun)) | |
| 131 | 2426 gcc_stablesort (edges, n, sizeof *edges, edge_order); |
| 111 | 2427 |
| 2428 /* Now decide which of those edges to make fallthrough edges. We set | |
| 2429 BB_VISITED if a block already has a fallthrough successor assigned | |
| 2430 to it. We make ->AUX of an endpoint point to the opposite endpoint | |
| 2431 of a sequence of blocks that fall through, and ->AUX will be NULL | |
| 2432 for a block that is in such a sequence but not an endpoint anymore. | |
| 2433 | |
| 2434 To start with, everything points to itself, nothing is assigned yet. */ | |
| 2435 | |
| 2436 FOR_ALL_BB_FN (bb, cfun) | |
| 2437 { | |
| 2438 bb->aux = bb; | |
| 2439 bb->flags &= ~BB_VISITED; | |
| 2440 } | |
| 2441 | |
| 2442 EXIT_BLOCK_PTR_FOR_FN (cfun)->aux = 0; | |
| 2443 | |
| 2444 /* Now for all edges, the most desirable first, see if that edge can | |
| 2445 connect two sequences. If it can, update AUX and BB_VISITED; if it | |
| 2446 cannot, zero out the edge in the table. */ | |
| 2447 | |
| 2448 for (int j = 0; j < n; j++) | |
| 2449 { | |
| 2450 edge e = edges[j]; | |
| 2451 | |
| 2452 basic_block tail_a = e->src; | |
| 2453 basic_block head_b = e->dest; | |
| 2454 basic_block head_a = (basic_block) tail_a->aux; | |
| 2455 basic_block tail_b = (basic_block) head_b->aux; | |
| 2456 | |
| 2457 /* An edge cannot connect two sequences if: | |
| 2458 - it crosses partitions; | |
| 2459 - its src is not a current endpoint; | |
| 2460 - its dest is not a current endpoint; | |
| 2461 - or, it would create a loop. */ | |
| 2462 | |
| 2463 if (e->flags & EDGE_CROSSING | |
| 2464 || tail_a->flags & BB_VISITED | |
| 2465 || !tail_b | |
| 2466 || (!(head_b->flags & BB_VISITED) && head_b != tail_b) | |
| 2467 || tail_a == tail_b) | |
| 2468 { | |
| 2469 edges[j] = 0; | |
| 2470 continue; | |
| 2471 } | |
| 2472 | |
| 2473 tail_a->aux = 0; | |
| 2474 head_b->aux = 0; | |
| 2475 head_a->aux = tail_b; | |
| 2476 tail_b->aux = head_a; | |
| 2477 tail_a->flags |= BB_VISITED; | |
| 2478 } | |
| 2479 | |
| 2480 /* Put the pieces together, in the same order that the start blocks of | |
| 2481 the sequences already had. The hot/cold partitioning gives a little | |
| 2482 complication: as a first pass only do this for blocks in the same | |
| 2483 partition as the start block, and (if there is anything left to do) | |
| 2484 in a second pass handle the other partition. */ | |
| 2485 | |
| 2486 basic_block last_tail = (basic_block) ENTRY_BLOCK_PTR_FOR_FN (cfun)->aux; | |
| 2487 | |
| 131 | 2488 int current_partition |
| 2489 = BB_PARTITION (last_tail == ENTRY_BLOCK_PTR_FOR_FN (cfun) | |
| 2490 ? EDGE_SUCC (ENTRY_BLOCK_PTR_FOR_FN (cfun), 0)->dest | |
| 2491 : last_tail); | |
| 111 | 2492 bool need_another_pass = true; |
| 2493 | |
| 2494 for (int pass = 0; pass < 2 && need_another_pass; pass++) | |
| 2495 { | |
| 2496 need_another_pass = false; | |
| 2497 | |
| 2498 FOR_EACH_BB_FN (bb, cfun) | |
| 2499 if ((bb->flags & BB_VISITED && bb->aux) || bb->aux == bb) | |
| 2500 { | |
| 2501 if (BB_PARTITION (bb) != current_partition) | |
| 2502 { | |
| 2503 need_another_pass = true; | |
| 2504 continue; | |
| 2505 } | |
| 2506 | |
| 2507 last_tail->aux = bb; | |
| 2508 last_tail = (basic_block) bb->aux; | |
| 2509 } | |
| 2510 | |
| 2511 current_partition ^= BB_HOT_PARTITION | BB_COLD_PARTITION; | |
| 2512 } | |
| 2513 | |
| 2514 last_tail->aux = 0; | |
| 2515 | |
| 2516 /* Finally, link all the chosen fallthrough edges. */ | |
| 2517 | |
| 2518 for (int j = 0; j < n; j++) | |
| 2519 if (edges[j]) | |
| 2520 edges[j]->src->aux = edges[j]->dest; | |
| 2521 | |
| 2522 delete[] edges; | |
| 2523 | |
| 2524 /* If the entry edge no longer falls through we have to make a new | |
| 2525 block so it can do so again. */ | |
| 2526 | |
| 2527 edge e = EDGE_SUCC (ENTRY_BLOCK_PTR_FOR_FN (cfun), 0); | |
| 2528 if (e->dest != ENTRY_BLOCK_PTR_FOR_FN (cfun)->aux) | |
| 2529 { | |
| 2530 force_nonfallthru (e); | |
| 2531 e->src->aux = ENTRY_BLOCK_PTR_FOR_FN (cfun)->aux; | |
| 2532 } | |
| 2533 } | |
| 2534 | |
| 2535 /* Reorder basic blocks. The main entry point to this file. */ | |
| 2536 | |
| 2537 static void | |
| 2538 reorder_basic_blocks (void) | |
| 2539 { | |
| 2540 gcc_assert (current_ir_type () == IR_RTL_CFGLAYOUT); | |
| 2541 | |
| 2542 if (n_basic_blocks_for_fn (cfun) <= NUM_FIXED_BLOCKS + 1) | |
| 2543 return; | |
| 2544 | |
| 2545 set_edge_can_fallthru_flag (); | |
| 2546 mark_dfs_back_edges (); | |
| 2547 | |
| 2548 switch (flag_reorder_blocks_algorithm) | |
| 2549 { | |
| 2550 case REORDER_BLOCKS_ALGORITHM_SIMPLE: | |
| 2551 reorder_basic_blocks_simple (); | |
| 2552 break; | |
| 2553 | |
| 2554 case REORDER_BLOCKS_ALGORITHM_STC: | |
| 2555 reorder_basic_blocks_software_trace_cache (); | |
| 2556 break; | |
| 2557 | |
| 2558 default: | |
| 2559 gcc_unreachable (); | |
| 2560 } | |
| 0 | 2561 |
| 2562 relink_block_chain (/*stay_in_cfglayout_mode=*/true); | |
| 2563 | |
| 2564 if (dump_file) | |
| 111 | 2565 { |
| 2566 if (dump_flags & TDF_DETAILS) | |
| 2567 dump_reg_info (dump_file); | |
| 2568 dump_flow_info (dump_file, dump_flags); | |
| 2569 } | |
| 2570 | |
| 2571 /* Signal that rtl_verify_flow_info_1 can now verify that there | |
| 2572 is at most one switch between hot/cold sections. */ | |
| 2573 crtl->bb_reorder_complete = true; | |
| 0 | 2574 } |
| 2575 | |
| 2576 /* Determine which partition the first basic block in the function | |
| 2577 belongs to, then find the first basic block in the current function | |
| 2578 that belongs to a different section, and insert a | |
| 2579 NOTE_INSN_SWITCH_TEXT_SECTIONS note immediately before it in the | |
| 2580 instruction stream. When writing out the assembly code, | |
| 2581 encountering this note will make the compiler switch between the | |
| 2582 hot and cold text sections. */ | |
| 2583 | |
| 111 | 2584 void |
| 0 | 2585 insert_section_boundary_note (void) |
| 2586 { | |
| 2587 basic_block bb; | |
| 111 | 2588 bool switched_sections = false; |
| 2589 int current_partition = 0; | |
| 2590 | |
| 2591 if (!crtl->has_bb_partition) | |
| 2592 return; | |
| 2593 | |
| 2594 FOR_EACH_BB_FN (bb, cfun) | |
| 0 | 2595 { |
| 111 | 2596 if (!current_partition) |
| 2597 current_partition = BB_PARTITION (bb); | |
| 2598 if (BB_PARTITION (bb) != current_partition) | |
| 0 | 2599 { |
| 111 | 2600 gcc_assert (!switched_sections); |
| 2601 switched_sections = true; | |
| 2602 emit_note_before (NOTE_INSN_SWITCH_TEXT_SECTIONS, BB_HEAD (bb)); | |
| 2603 current_partition = BB_PARTITION (bb); | |
| 0 | 2604 } |
| 2605 } | |
| 131 | 2606 |
| 2607 /* Make sure crtl->has_bb_partition matches reality even if bbpart finds | |
| 2608 some hot and some cold basic blocks, but later one of those kinds is | |
| 2609 optimized away. */ | |
| 2610 crtl->has_bb_partition = switched_sections; | |
| 0 | 2611 } |
| 2612 | |
| 111 | 2613 namespace { |
| 2614 | |
| 2615 const pass_data pass_data_reorder_blocks = | |
| 2616 { | |
| 2617 RTL_PASS, /* type */ | |
| 2618 "bbro", /* name */ | |
| 2619 OPTGROUP_NONE, /* optinfo_flags */ | |
| 2620 TV_REORDER_BLOCKS, /* tv_id */ | |
| 2621 0, /* properties_required */ | |
| 2622 0, /* properties_provided */ | |
| 2623 0, /* properties_destroyed */ | |
| 2624 0, /* todo_flags_start */ | |
| 2625 0, /* todo_flags_finish */ | |
| 2626 }; | |
| 2627 | |
| 2628 class pass_reorder_blocks : public rtl_opt_pass | |
| 2629 { | |
| 2630 public: | |
| 2631 pass_reorder_blocks (gcc::context *ctxt) | |
| 2632 : rtl_opt_pass (pass_data_reorder_blocks, ctxt) | |
| 2633 {} | |
| 2634 | |
| 2635 /* opt_pass methods: */ | |
| 2636 virtual bool gate (function *) | |
| 2637 { | |
| 2638 if (targetm.cannot_modify_jumps_p ()) | |
| 2639 return false; | |
| 2640 return (optimize > 0 | |
| 2641 && (flag_reorder_blocks || flag_reorder_blocks_and_partition)); | |
| 2642 } | |
| 2643 | |
| 2644 virtual unsigned int execute (function *); | |
| 2645 | |
| 2646 }; // class pass_reorder_blocks | |
| 2647 | |
| 2648 unsigned int | |
| 2649 pass_reorder_blocks::execute (function *fun) | |
| 2650 { | |
| 2651 basic_block bb; | |
| 2652 | |
| 2653 /* Last attempt to optimize CFG, as scheduling, peepholing and insn | |
| 2654 splitting possibly introduced more crossjumping opportunities. */ | |
| 2655 cfg_layout_initialize (CLEANUP_EXPENSIVE); | |
| 2656 | |
| 2657 reorder_basic_blocks (); | |
| 131 | 2658 cleanup_cfg (CLEANUP_EXPENSIVE | CLEANUP_NO_PARTITIONING); |
| 111 | 2659 |
| 2660 FOR_EACH_BB_FN (bb, fun) | |
| 2661 if (bb->next_bb != EXIT_BLOCK_PTR_FOR_FN (fun)) | |
| 2662 bb->aux = bb->next_bb; | |
| 2663 cfg_layout_finalize (); | |
| 2664 | |
| 145 | 2665 FOR_EACH_BB_FN (bb, fun) |
| 2666 df_recompute_luids (bb); | |
| 111 | 2667 return 0; |
| 2668 } | |
| 2669 | |
| 2670 } // anon namespace | |
| 2671 | |
| 2672 rtl_opt_pass * | |
| 2673 make_pass_reorder_blocks (gcc::context *ctxt) | |
| 2674 { | |
| 2675 return new pass_reorder_blocks (ctxt); | |
| 2676 } | |
| 2677 | |
| 2678 /* Duplicate a block (that we already know ends in a computed jump) into its | |
| 2679 predecessors, where possible. Return whether anything is changed. */ | |
| 2680 static bool | |
| 2681 maybe_duplicate_computed_goto (basic_block bb, int max_size) | |
| 2682 { | |
| 2683 if (single_pred_p (bb)) | |
| 2684 return false; | |
| 2685 | |
| 2686 /* Make sure that the block is small enough. */ | |
| 2687 rtx_insn *insn; | |
| 2688 FOR_BB_INSNS (bb, insn) | |
| 2689 if (INSN_P (insn)) | |
| 2690 { | |
| 2691 max_size -= get_attr_min_length (insn); | |
| 2692 if (max_size < 0) | |
| 2693 return false; | |
| 2694 } | |
| 2695 | |
| 2696 bool changed = false; | |
| 2697 edge e; | |
| 2698 edge_iterator ei; | |
| 2699 for (ei = ei_start (bb->preds); (e = ei_safe_edge (ei)); ) | |
| 2700 { | |
| 2701 basic_block pred = e->src; | |
| 2702 | |
| 2703 /* Do not duplicate BB into PRED if that is the last predecessor, or if | |
| 2704 we cannot merge a copy of BB with PRED. */ | |
| 2705 if (single_pred_p (bb) | |
| 2706 || !single_succ_p (pred) | |
| 2707 || e->flags & EDGE_COMPLEX | |
| 2708 || pred->index < NUM_FIXED_BLOCKS | |
| 2709 || (JUMP_P (BB_END (pred)) && !simplejump_p (BB_END (pred))) | |
| 2710 || (JUMP_P (BB_END (pred)) && CROSSING_JUMP_P (BB_END (pred)))) | |
| 2711 { | |
| 2712 ei_next (&ei); | |
| 2713 continue; | |
| 2714 } | |
| 2715 | |
| 2716 if (dump_file) | |
| 2717 fprintf (dump_file, "Duplicating computed goto bb %d into bb %d\n", | |
| 2718 bb->index, e->src->index); | |
| 2719 | |
| 2720 /* Remember if PRED can be duplicated; if so, the copy of BB merged | |
| 2721 with PRED can be duplicated as well. */ | |
| 2722 bool can_dup_more = can_duplicate_block_p (pred); | |
| 2723 | |
| 2724 /* Make a copy of BB, merge it into PRED. */ | |
| 2725 basic_block copy = duplicate_block (bb, e, NULL); | |
| 2726 emit_barrier_after_bb (copy); | |
| 2727 reorder_insns_nobb (BB_HEAD (copy), BB_END (copy), BB_END (pred)); | |
| 2728 merge_blocks (pred, copy); | |
| 2729 | |
| 2730 changed = true; | |
| 2731 | |
| 2732 /* Try to merge the resulting merged PRED into further predecessors. */ | |
| 2733 if (can_dup_more) | |
| 2734 maybe_duplicate_computed_goto (pred, max_size); | |
| 2735 } | |
| 2736 | |
| 2737 return changed; | |
| 2738 } | |
| 2739 | |
| 0 | 2740 /* Duplicate the blocks containing computed gotos. This basically unfactors |
| 2741 computed gotos that were factored early on in the compilation process to | |
| 111 | 2742 speed up edge based data flow. We used to not unfactor them again, which |
| 2743 can seriously pessimize code with many computed jumps in the source code, | |
| 2744 such as interpreters. See e.g. PR15242. */ | |
| 2745 static void | |
| 2746 duplicate_computed_gotos (function *fun) | |
| 0 | 2747 { |
| 2748 /* We are estimating the length of uncond jump insn only once | |
| 2749 since the code for getting the insn length always returns | |
| 2750 the minimal length now. */ | |
| 2751 if (uncond_jump_length == 0) | |
| 2752 uncond_jump_length = get_uncond_jump_length (); | |
| 2753 | |
| 111 | 2754 /* Never copy a block larger than this. */ |
| 2755 int max_size | |
| 145 | 2756 = uncond_jump_length * param_max_goto_duplication_insns; |
| 111 | 2757 |
| 2758 bool changed = false; | |
| 2759 | |
| 2760 /* Try to duplicate all blocks that end in a computed jump and that | |
| 2761 can be duplicated at all. */ | |
| 2762 basic_block bb; | |
| 2763 FOR_EACH_BB_FN (bb, fun) | |
| 2764 if (computed_jump_p (BB_END (bb)) && can_duplicate_block_p (bb)) | |
| 2765 changed |= maybe_duplicate_computed_goto (bb, max_size); | |
| 2766 | |
| 2767 /* Duplicating blocks will redirect edges and may cause hot blocks | |
| 2768 previously reached by both hot and cold blocks to become dominated | |
| 2769 only by cold blocks. */ | |
| 2770 if (changed) | |
| 2771 fixup_partitions (); | |
| 2772 } | |
| 2773 | |
| 2774 namespace { | |
| 2775 | |
| 2776 const pass_data pass_data_duplicate_computed_gotos = | |
| 2777 { | |
| 2778 RTL_PASS, /* type */ | |
| 2779 "compgotos", /* name */ | |
| 2780 OPTGROUP_NONE, /* optinfo_flags */ | |
| 2781 TV_REORDER_BLOCKS, /* tv_id */ | |
| 2782 0, /* properties_required */ | |
| 2783 0, /* properties_provided */ | |
| 2784 0, /* properties_destroyed */ | |
| 2785 0, /* todo_flags_start */ | |
| 2786 0, /* todo_flags_finish */ | |
| 2787 }; | |
| 2788 | |
| 2789 class pass_duplicate_computed_gotos : public rtl_opt_pass | |
| 2790 { | |
| 2791 public: | |
| 2792 pass_duplicate_computed_gotos (gcc::context *ctxt) | |
| 2793 : rtl_opt_pass (pass_data_duplicate_computed_gotos, ctxt) | |
| 2794 {} | |
| 2795 | |
| 2796 /* opt_pass methods: */ | |
| 2797 virtual bool gate (function *); | |
| 2798 virtual unsigned int execute (function *); | |
| 2799 | |
| 2800 }; // class pass_duplicate_computed_gotos | |
| 2801 | |
| 2802 bool | |
| 2803 pass_duplicate_computed_gotos::gate (function *fun) | |
| 2804 { | |
| 2805 if (targetm.cannot_modify_jumps_p ()) | |
| 2806 return false; | |
| 2807 return (optimize > 0 | |
| 2808 && flag_expensive_optimizations | |
| 2809 && ! optimize_function_for_size_p (fun)); | |
| 2810 } | |
| 2811 | |
| 2812 unsigned int | |
| 2813 pass_duplicate_computed_gotos::execute (function *fun) | |
| 2814 { | |
| 2815 duplicate_computed_gotos (fun); | |
| 2816 | |
| 0 | 2817 return 0; |
| 2818 } | |
| 2819 | |
| 111 | 2820 } // anon namespace |
| 2821 | |
| 2822 rtl_opt_pass * | |
| 2823 make_pass_duplicate_computed_gotos (gcc::context *ctxt) | |
| 0 | 2824 { |
| 111 | 2825 return new pass_duplicate_computed_gotos (ctxt); |
| 2826 } | |
| 0 | 2827 |
| 2828 /* This function is the main 'entrance' for the optimization that | |
| 2829 partitions hot and cold basic blocks into separate sections of the | |
| 2830 .o file (to improve performance and cache locality). Ideally it | |
| 2831 would be called after all optimizations that rearrange the CFG have | |
| 2832 been called. However part of this optimization may introduce new | |
| 2833 register usage, so it must be called before register allocation has | |
| 2834 occurred. This means that this optimization is actually called | |
| 2835 well before the optimization that reorders basic blocks (see | |
| 2836 function above). | |
| 2837 | |
| 2838 This optimization checks the feedback information to determine | |
| 2839 which basic blocks are hot/cold, updates flags on the basic blocks | |
| 2840 to indicate which section they belong in. This information is | |
| 2841 later used for writing out sections in the .o file. Because hot | |
| 2842 and cold sections can be arbitrarily large (within the bounds of | |
| 2843 memory), far beyond the size of a single function, it is necessary | |
| 2844 to fix up all edges that cross section boundaries, to make sure the | |
| 2845 instructions used can actually span the required distance. The | |
| 2846 fixes are described below. | |
| 2847 | |
| 2848 Fall-through edges must be changed into jumps; it is not safe or | |
| 2849 legal to fall through across a section boundary. Whenever a | |
| 2850 fall-through edge crossing a section boundary is encountered, a new | |
| 2851 basic block is inserted (in the same section as the fall-through | |
| 2852 source), and the fall through edge is redirected to the new basic | |
| 2853 block. The new basic block contains an unconditional jump to the | |
| 2854 original fall-through target. (If the unconditional jump is | |
| 2855 insufficient to cross section boundaries, that is dealt with a | |
| 2856 little later, see below). | |
| 2857 | |
| 2858 In order to deal with architectures that have short conditional | |
| 2859 branches (which cannot span all of memory) we take any conditional | |
| 2860 jump that attempts to cross a section boundary and add a level of | |
| 2861 indirection: it becomes a conditional jump to a new basic block, in | |
| 2862 the same section. The new basic block contains an unconditional | |
| 2863 jump to the original target, in the other section. | |
| 2864 | |
| 2865 For those architectures whose unconditional branch is also | |
| 2866 incapable of reaching all of memory, those unconditional jumps are | |
| 2867 converted into indirect jumps, through a register. | |
| 2868 | |
| 2869 IMPORTANT NOTE: This optimization causes some messy interactions | |
| 2870 with the cfg cleanup optimizations; those optimizations want to | |
| 2871 merge blocks wherever possible, and to collapse indirect jump | |
| 2872 sequences (change "A jumps to B jumps to C" directly into "A jumps | |
| 2873 to C"). Those optimizations can undo the jump fixes that | |
| 2874 partitioning is required to make (see above), in order to ensure | |
| 2875 that jumps attempting to cross section boundaries are really able | |
| 2876 to cover whatever distance the jump requires (on many architectures | |
| 2877 conditional or unconditional jumps are not able to reach all of | |
| 2878 memory). Therefore tests have to be inserted into each such | |
| 2879 optimization to make sure that it does not undo stuff necessary to | |
| 2880 cross partition boundaries. This would be much less of a problem | |
| 2881 if we could perform this optimization later in the compilation, but | |
| 2882 unfortunately the fact that we may need to create indirect jumps | |
| 2883 (through registers) requires that this optimization be performed | |
| 111 | 2884 before register allocation. |
| 2885 | |
| 2886 Hot and cold basic blocks are partitioned and put in separate | |
| 2887 sections of the .o file, to reduce paging and improve cache | |
| 2888 performance (hopefully). This can result in bits of code from the | |
| 2889 same function being widely separated in the .o file. However this | |
| 2890 is not obvious to the current bb structure. Therefore we must take | |
| 2891 care to ensure that: 1). There are no fall_thru edges that cross | |
| 2892 between sections; 2). For those architectures which have "short" | |
| 2893 conditional branches, all conditional branches that attempt to | |
| 2894 cross between sections are converted to unconditional branches; | |
| 2895 and, 3). For those architectures which have "short" unconditional | |
| 2896 branches, all unconditional branches that attempt to cross between | |
| 2897 sections are converted to indirect jumps. | |
| 2898 | |
| 2899 The code for fixing up fall_thru edges that cross between hot and | |
| 2900 cold basic blocks does so by creating new basic blocks containing | |
| 2901 unconditional branches to the appropriate label in the "other" | |
| 2902 section. The new basic block is then put in the same (hot or cold) | |
| 2903 section as the original conditional branch, and the fall_thru edge | |
| 2904 is modified to fall into the new basic block instead. By adding | |
| 2905 this level of indirection we end up with only unconditional branches | |
| 2906 crossing between hot and cold sections. | |
| 2907 | |
| 2908 Conditional branches are dealt with by adding a level of indirection. | |
| 2909 A new basic block is added in the same (hot/cold) section as the | |
| 2910 conditional branch, and the conditional branch is retargeted to the | |
| 2911 new basic block. The new basic block contains an unconditional branch | |
| 2912 to the original target of the conditional branch (in the other section). | |
| 2913 | |
| 2914 Unconditional branches are dealt with by converting them into | |
| 2915 indirect jumps. */ | |
| 2916 | |
| 2917 namespace { | |
| 2918 | |
| 2919 const pass_data pass_data_partition_blocks = | |
| 0 | 2920 { |
| 111 | 2921 RTL_PASS, /* type */ |
| 2922 "bbpart", /* name */ | |
| 2923 OPTGROUP_NONE, /* optinfo_flags */ | |
| 2924 TV_REORDER_BLOCKS, /* tv_id */ | |
| 2925 PROP_cfglayout, /* properties_required */ | |
| 2926 0, /* properties_provided */ | |
| 2927 0, /* properties_destroyed */ | |
| 2928 0, /* todo_flags_start */ | |
| 2929 0, /* todo_flags_finish */ | |
| 2930 }; | |
| 2931 | |
| 2932 class pass_partition_blocks : public rtl_opt_pass | |
| 0 | 2933 { |
| 111 | 2934 public: |
| 2935 pass_partition_blocks (gcc::context *ctxt) | |
| 2936 : rtl_opt_pass (pass_data_partition_blocks, ctxt) | |
| 2937 {} | |
| 2938 | |
| 2939 /* opt_pass methods: */ | |
| 2940 virtual bool gate (function *); | |
| 2941 virtual unsigned int execute (function *); | |
| 2942 | |
| 2943 }; // class pass_partition_blocks | |
| 2944 | |
| 2945 bool | |
| 2946 pass_partition_blocks::gate (function *fun) | |
| 0 | 2947 { |
| 2948 /* The optimization to partition hot/cold basic blocks into separate | |
| 2949 sections of the .o file does not work well with linkonce or with | |
| 131 | 2950 user defined section attributes or with naked attribute. Don't call |
| 2951 it if either case arises. */ | |
| 0 | 2952 return (flag_reorder_blocks_and_partition |
| 111 | 2953 && optimize |
| 2954 /* See pass_reorder_blocks::gate. We should not partition if | |
| 2955 we are going to omit the reordering. */ | |
| 2956 && optimize_function_for_speed_p (fun) | |
| 2957 && !DECL_COMDAT_GROUP (current_function_decl) | |
| 131 | 2958 && !lookup_attribute ("section", DECL_ATTRIBUTES (fun->decl)) |
| 2959 && !lookup_attribute ("naked", DECL_ATTRIBUTES (fun->decl)) | |
| 2960 /* Workaround a bug in GDB where read_partial_die doesn't cope | |
| 2961 with DIEs with DW_AT_ranges, see PR81115. */ | |
| 2962 && !(in_lto_p && MAIN_NAME_P (DECL_NAME (fun->decl)))); | |
| 0 | 2963 } |
| 2964 | |
| 111 | 2965 unsigned |
| 2966 pass_partition_blocks::execute (function *fun) | |
| 0 | 2967 { |
| 111 | 2968 vec<edge> crossing_edges; |
| 2969 | |
| 2970 if (n_basic_blocks_for_fn (fun) <= NUM_FIXED_BLOCKS + 1) | |
| 2971 return 0; | |
| 2972 | |
| 2973 df_set_flags (DF_DEFER_INSN_RESCAN); | |
| 2974 | |
| 2975 crossing_edges = find_rarely_executed_basic_blocks_and_crossing_edges (); | |
| 2976 if (!crossing_edges.exists ()) | |
| 2977 /* Make sure to process deferred rescans and clear changeable df flags. */ | |
| 2978 return TODO_df_finish; | |
| 2979 | |
| 2980 crtl->has_bb_partition = true; | |
| 2981 | |
| 2982 /* Make sure the source of any crossing edge ends in a jump and the | |
| 2983 destination of any crossing edge has a label. */ | |
| 2984 add_labels_and_missing_jumps (crossing_edges); | |
| 2985 | |
| 2986 /* Convert all crossing fall_thru edges to non-crossing fall | |
| 2987 thrus to unconditional jumps (that jump to the original fall | |
| 2988 through dest). */ | |
| 2989 fix_up_fall_thru_edges (); | |
| 2990 | |
| 2991 /* If the architecture does not have conditional branches that can | |
| 2992 span all of memory, convert crossing conditional branches into | |
| 2993 crossing unconditional branches. */ | |
| 2994 if (!HAS_LONG_COND_BRANCH) | |
| 2995 fix_crossing_conditional_branches (); | |
| 2996 | |
| 2997 /* If the architecture does not have unconditional branches that | |
| 2998 can span all of memory, convert crossing unconditional branches | |
| 2999 into indirect jumps. Since adding an indirect jump also adds | |
| 3000 a new register usage, update the register usage information as | |
| 3001 well. */ | |
| 3002 if (!HAS_LONG_UNCOND_BRANCH) | |
| 3003 fix_crossing_unconditional_branches (); | |
| 3004 | |
| 3005 update_crossing_jump_flags (); | |
| 3006 | |
| 3007 /* Clear bb->aux fields that the above routines were using. */ | |
| 3008 clear_aux_for_blocks (); | |
| 3009 | |
| 3010 crossing_edges.release (); | |
| 3011 | |
| 3012 /* ??? FIXME: DF generates the bb info for a block immediately. | |
| 3013 And by immediately, I mean *during* creation of the block. | |
| 3014 | |
| 3015 #0 df_bb_refs_collect | |
| 3016 #1 in df_bb_refs_record | |
| 3017 #2 in create_basic_block_structure | |
| 3018 | |
| 3019 Which means that the bb_has_eh_pred test in df_bb_refs_collect | |
| 3020 will *always* fail, because no edges can have been added to the | |
| 3021 block yet. Which of course means we don't add the right | |
| 3022 artificial refs, which means we fail df_verify (much) later. | |
| 3023 | |
| 3024 Cleanest solution would seem to make DF_DEFER_INSN_RESCAN imply | |
| 3025 that we also shouldn't grab data from the new blocks those new | |
| 3026 insns are in either. In this way one can create the block, link | |
| 3027 it up properly, and have everything Just Work later, when deferred | |
| 3028 insns are processed. | |
| 3029 | |
| 3030 In the meantime, we have no other option but to throw away all | |
| 3031 of the DF data and recompute it all. */ | |
| 3032 if (fun->eh->lp_array) | |
| 3033 { | |
| 3034 df_finish_pass (true); | |
| 3035 df_scan_alloc (NULL); | |
| 3036 df_scan_blocks (); | |
| 3037 /* Not all post-landing pads use all of the EH_RETURN_DATA_REGNO | |
| 3038 data. We blindly generated all of them when creating the new | |
| 3039 landing pad. Delete those assignments we don't use. */ | |
| 3040 df_set_flags (DF_LR_RUN_DCE); | |
| 3041 df_analyze (); | |
| 3042 } | |
| 3043 | |
| 3044 /* Make sure to process deferred rescans and clear changeable df flags. */ | |
| 3045 return TODO_df_finish; | |
| 3046 } | |
| 3047 | |
| 3048 } // anon namespace | |
| 3049 | |
| 3050 rtl_opt_pass * | |
| 3051 make_pass_partition_blocks (gcc::context *ctxt) | |
| 3052 { | |
| 3053 return new pass_partition_blocks (ctxt); | |
| 3054 } |