boehm-gc/backgraph.c

   1 /*
   2  * Copyright (c) 2001 by Hewlett-Packard Company. All rights reserved.
   3  *
   4  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
   5  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
   6  *
   7  * Permission is hereby granted to use or copy this program
   8  * for any purpose,  provided the above notices are retained on all copies.
   9  * Permission to modify the code and to distribute modified code is granted,
  10  * provided the above notices are retained, and a notice that the code was
  11  * modified is included with the above copyright notice.
  12  *
  13  */
  14
  15 /*
  16  * This implements a full, though not well-tuned, representation of the
  17  * backwards points-to graph.  This is used to test for non-GC-robust
  18  * data structures; the code is not used during normal garbage collection.
  19  *
  20  * One restriction is that we drop all back-edges from nodes with very
  21  * high in-degree, and simply add them add them to a list of such
  22  * nodes.  They are then treated as permanent roots.  Id this by itself
  23  * doesn't introduce a space leak, then such nodes can't contribute to
  24  * a growing space leak.
  25  */
  26
  27 #include "gc.h" /* For configuration information. */
  28
  29 #ifdef MAKE_BACK_GRAPH
  30
  31 #define MAX_IN  10      /* Maximum in-degree we handle directly */
  32
  33 #include "private/dbg_mlc.h"
  34 #include <unistd.h>
  35
  36 #if !defined(DBG_HDRS_ALL) || (ALIGNMENT != CPP_WORDSZ/8) || !defined(UNIX_LIKE)
  37 # error Configuration doesnt support MAKE_BACK_GRAPH
  38 #endif
  39
  40 /* We store single back pointers directly in the object's oh_bg_ptr field.   */
  41 /* If there is more than one ptr to an object, we store q | FLAG_MANY,       */
  42 /* where q is a pointer to a back_edges object.                              */
  43 /* Every once in a while we use a back_edges object even for a single        */
  44 /* pointer, since we need the other fields in the back_edges structure to    */
  45 /* be present in some fraction of the objects.  Otherwise we get serious     */
  46 /* performance issues.                                                       */
  47 #define FLAG_MANY 2
  48
  49 typedef struct back_edges_struct {
  50   word n_edges; /* Number of edges, including those in continuation     */
  51                 /* structures.                                          */
  52   unsigned short flags;
  53 #       define RETAIN 1 /* Directly points to a reachable object;       */
  54                         /* retain for next GC.                          */
  55   unsigned short height_gc_no;
  56                 /* If height > 0, then the GC_gc_no value when it       */
  57                 /* was computed.  If it was computed this cycle, then   */
  58                 /* it is current.  If it was computed during the        */
  59                 /* last cycle, then it represents the old height,       */
  60                 /* which is only saved for live objects referenced by   */
  61                 /* dead ones.  This may grow due to refs from newly     */
  62                 /* dead objects.                                        */
  63   signed_word height;
  64                 /* Longest path through unreachable nodes to this node  */
  65                 /* that we found using depth first search.              */
  66
  67 #   define HEIGHT_UNKNOWN ((signed_word)(-2))
  68 #   define HEIGHT_IN_PROGRESS ((signed_word)(-1))
  69   ptr_t edges[MAX_IN];
  70   struct back_edges_struct *cont;
  71                 /* Pointer to continuation structure; we use only the   */
  72                 /* edges field in the continuation.                     */
  73                 /* also used as free list link.                         */
  74 } back_edges;
  75
  76 /* Allocate a new back edge structure.  Should be more sophisticated    */
  77 /* if this were production code.                                        */
  78 #define MAX_BACK_EDGE_STRUCTS 100000
  79 static back_edges *back_edge_space = 0;
  80 int GC_n_back_edge_structs = 0; /* Serves as pointer to never used      */
  81                                 /* back_edges space.                    */
  82 static back_edges *avail_back_edges = 0;
  83                                 /* Pointer to free list of deallocated  */
  84                                 /* back_edges structures.               */
  85
  86 static back_edges * new_back_edges(void)
  87 {
  88   if (0 == back_edge_space) {
  89     back_edge_space = (back_edges *)
  90                         GET_MEM(MAX_BACK_EDGE_STRUCTS*sizeof(back_edges));
  91   }
  92   if (0 != avail_back_edges) {
  93     back_edges * result = avail_back_edges;
  94     avail_back_edges = result -> cont;
  95     result -> cont = 0;
  96     return result;
  97   }
  98   if (GC_n_back_edge_structs >= MAX_BACK_EDGE_STRUCTS - 1) {
  99     ABORT("needed too much space for back edges: adjust "
 100           "MAX_BACK_EDGE_STRUCTS");
 101   }
 102   return back_edge_space + (GC_n_back_edge_structs++);
 103 }
 104
 105 /* Deallocate p and its associated continuation structures.     */
 106 static void deallocate_back_edges(back_edges *p)
 107 {
 108    back_edges *last = p;
 109
 110    while (0 != last -> cont) last = last -> cont;
 111    last -> cont = avail_back_edges;
 112    avail_back_edges = p;
 113 }
 114
 115 /* Table of objects that are currently on the depth-first search        */
 116 /* stack.  Only objects with in-degree one are in this table.           */
 117 /* Other objects are identified using HEIGHT_IN_PROGRESS.               */
 118 /* FIXME: This data structure NEEDS IMPROVEMENT.                        */
 119 #define INITIAL_IN_PROGRESS 10000
 120 static ptr_t * in_progress_space = 0;
 121 static size_t in_progress_size = 0;
 122 static size_t n_in_progress = 0;
 123
 124 static void push_in_progress(ptr_t p)
 125 {
 126   if (n_in_progress >= in_progress_size)
 127     if (in_progress_size == 0) {
 128       in_progress_size = INITIAL_IN_PROGRESS;
 129       in_progress_space = (ptr_t *)GET_MEM(in_progress_size * sizeof(ptr_t));
 130     } else {
 131       ptr_t * new_in_progress_space;
 132       in_progress_size *= 2;
 133       new_in_progress_space = (ptr_t *)
 134                                 GET_MEM(in_progress_size * sizeof(ptr_t));
 135       BCOPY(in_progress_space, new_in_progress_space,
 136             n_in_progress * sizeof(ptr_t));
 137       in_progress_space = new_in_progress_space;
 138       /* FIXME: This just drops the old space.  */
 139     }
 140   if (in_progress_space == 0)
 141       ABORT("MAKE_BACK_GRAPH: Out of in-progress space: "
 142             "Huge linear data structure?");
 143   in_progress_space[n_in_progress++] = p;
 144 }
 145
 146 static GC_bool is_in_progress(ptr_t p)
 147 {
 148   int i;
 149   for (i = 0; i < n_in_progress; ++i) {
 150     if (in_progress_space[i] == p) return TRUE;
 151   }
 152   return FALSE;
 153 }
 154
 155 static void pop_in_progress(ptr_t p)
 156 {
 157   --n_in_progress;
 158   GC_ASSERT(in_progress_space[n_in_progress] == p);
 159 }
 160
 161 #define GET_OH_BG_PTR(p) \
 162                 (ptr_t)REVEAL_POINTER(((oh *)(p)) -> oh_bg_ptr)
 163 #define SET_OH_BG_PTR(p,q) (((oh *)(p)) -> oh_bg_ptr) = HIDE_POINTER(q)
 164
 165 /* Execute s once for each predecessor q of p in the points-to graph.   */
 166 /* s should be a bracketed statement.  We declare q.                    */
 167 #define FOR_EACH_PRED(q, p, s) \
 168   { \
 169     ptr_t q = GET_OH_BG_PTR(p); \
 170     if (!((word)q & FLAG_MANY)) { \
 171       if (q && !((word)q & 1)) s \
 172               /* !((word)q & 1) checks for a misnterpreted freelist link */ \
 173     } else { \
 174       back_edges *orig_be_ = (back_edges *)((word)q & ~FLAG_MANY); \
 175       back_edges *be_ = orig_be_; \
 176       int total_, local_; \
 177       int n_edges_ = be_ -> n_edges; \
 178       for (total_ = 0, local_ = 0; total_ < n_edges_; ++local_, ++total_) { \
 179           if (local_ == MAX_IN) { \
 180               be_ = be_ -> cont; \
 181               local_ = 0; \
 182           } \
 183           q = be_ -> edges[local_]; s \
 184       } \
 185     } \
 186   }
 187
 188 /* Ensure that p has a back_edges structure associated with it. */
 189 static void ensure_struct(ptr_t p)
 190 {
 191   ptr_t old_back_ptr = GET_OH_BG_PTR(p);
 192
 193   if (!((word)old_back_ptr & FLAG_MANY)) {
 194     back_edges *be = new_back_edges();
 195     be -> flags = 0;
 196     if (0 == old_back_ptr) {
 197       be -> n_edges = 0;
 198     } else {
 199       be -> n_edges = 1;
 200       be -> edges[0] = old_back_ptr;
 201     }
 202     be -> height = HEIGHT_UNKNOWN;
 203     be -> height_gc_no = GC_gc_no - 1;
 204     GC_ASSERT(be >= back_edge_space);
 205     SET_OH_BG_PTR(p, (word)be | FLAG_MANY);
 206   }
 207 }
 208
 209 /* Add the (forward) edge from p to q to the backward graph.  Both p    */
 210 /* q are pointers to the object base, i.e. pointers to an oh.           */
 211 static void add_edge(ptr_t p,  ptr_t q)
 212 {
 213     ptr_t old_back_ptr = GET_OH_BG_PTR(q);
 214     back_edges * be, *be_cont;
 215     word i;
 216     static unsigned random_number = 13;
 217 #   define GOT_LUCKY_NUMBER (((++random_number) & 0x7f) == 0)
 218       /* A not very random number we use to occasionally allocate a     */
 219       /* back_edges structure even for a single backward edge.  This    */
 220       /* prevents us from repeatedly tracing back through very long     */
 221       /* chains, since we will have some place to store height and      */
 222       /* in_progress flags along the way.                               */
 223
 224     GC_ASSERT(p == GC_base(p) && q == GC_base(q));
 225     if (!GC_HAS_DEBUG_INFO(q) || !GC_HAS_DEBUG_INFO(p)) {
 226       /* This is really a misinterpreted free list link, since we saw */
 227       /* a pointer to a free list.  Dont overwrite it!                */
 228       return;
 229     }
 230     if (0 == old_back_ptr) {
 231         SET_OH_BG_PTR(q, p);
 232         if (GOT_LUCKY_NUMBER) ensure_struct(q);
 233         return;
 234     }
 235     /* Check whether it was already in the list of predecessors. */
 236       FOR_EACH_PRED(pred, q, { if (p == pred) return; });
 237     ensure_struct(q);
 238     old_back_ptr = GET_OH_BG_PTR(q);
 239     be = (back_edges *)((word)old_back_ptr & ~FLAG_MANY);
 240     for (i = be -> n_edges, be_cont = be; i > MAX_IN;
 241         be_cont = be_cont -> cont, i -= MAX_IN) {}
 242     if (i == MAX_IN) {
 243         be_cont -> cont = new_back_edges();
 244         be_cont = be_cont -> cont;
 245         i = 0;
 246     }
 247     be_cont -> edges[i] = p;
 248     be -> n_edges++;
 249     if (be -> n_edges == 100) {
 250 #       if 0
 251           if (GC_print_stats) {
 252             GC_err_printf0("The following object has in-degree >= 100:\n");
 253             GC_print_heap_obj(q);
 254           }
 255 #       endif
 256     }
 257 }
 258
 259 typedef void (*per_object_func)(ptr_t p, word n_words, word gc_descr);
 260
 261 static void per_object_helper(struct hblk *h, word fn)
 262 {
 263   hdr * hhdr = HDR(h);
 264   word sz = hhdr -> hb_sz;
 265   word descr = hhdr -> hb_descr;
 266   per_object_func f = (per_object_func)fn;
 267   int i = 0;
 268
 269   do {
 270     f((ptr_t)(h -> hb_body + i), sz, descr);
 271     i += sz;
 272   } while (i + sz <= BYTES_TO_WORDS(HBLKSIZE));
 273 }
 274
 275 void GC_apply_to_each_object(per_object_func f)
 276 {
 277   GC_apply_to_all_blocks(per_object_helper, (word)f);
 278 }
 279
 280 static void reset_back_edge(ptr_t p, word n_words, word gc_descr)
 281 {
 282   /* Skip any free list links, or dropped blocks */
 283   if (GC_HAS_DEBUG_INFO(p)) {
 284     ptr_t old_back_ptr = GET_OH_BG_PTR(p);
 285     if ((word)old_back_ptr & FLAG_MANY) {
 286       back_edges *be = (back_edges *)((word)old_back_ptr & ~FLAG_MANY);
 287       if (!(be -> flags & RETAIN)) {
 288         deallocate_back_edges(be);
 289         SET_OH_BG_PTR(p, 0);
 290       } else {
 291         word *currentp;
 292
 293         GC_ASSERT(GC_is_marked(p));
 294
 295         /* Back edges may point to objects that will not be retained.   */
 296         /* Delete them for now, but remember the height.                */
 297         /* Some will be added back at next GC.                          */
 298           be -> n_edges = 0;
 299           if (0 != be -> cont) {
 300             deallocate_back_edges(be -> cont);
 301             be -> cont = 0;
 302           }
 303
 304         GC_ASSERT(GC_is_marked(p));
 305
 306         /* We only retain things for one GC cycle at a time.            */
 307           be -> flags &= ~RETAIN;
 308       }
 309     } else /* Simple back pointer */ {
 310       /* Clear to avoid dangling pointer. */
 311       SET_OH_BG_PTR(p, 0);
 312     }
 313   }
 314 }
 315
 316 static void add_back_edges(ptr_t p, word n_words, word gc_descr)
 317 {
 318   word *currentp = (word *)(p + sizeof(oh));
 319
 320   /* For now, fix up non-length descriptors conservatively.     */
 321     if((gc_descr & GC_DS_TAGS) != GC_DS_LENGTH) {
 322       gc_descr = WORDS_TO_BYTES(n_words);
 323     }
 324   while (currentp < (word *)(p + gc_descr)) {
 325     word current = *currentp++;
 326     FIXUP_POINTER(current);
 327     if (current >= (word)GC_least_plausible_heap_addr &&
 328         current <= (word)GC_greatest_plausible_heap_addr) {
 329        ptr_t target = GC_base((GC_PTR)current);
 330        if (0 != target) {
 331          add_edge(p, target);
 332        }
 333     }
 334   }
 335 }
 336
 337 /* Rebuild the representation of the backward reachability graph.       */
 338 /* Does not examine mark bits.  Can be called before GC.                */
 339 void GC_build_back_graph(void)
 340 {
 341   GC_apply_to_each_object(add_back_edges);
 342 }
 343
 344 /* Return an approximation to the length of the longest simple path     */
 345 /* through unreachable objects to p.  We refer to this as the height    */
 346 /* of p.                                                                */
 347 static word backwards_height(ptr_t p)
 348 {
 349   word result;
 350   ptr_t back_ptr = GET_OH_BG_PTR(p);
 351   back_edges *be;
 352
 353   if (0 == back_ptr) return 1;
 354   if (!((word)back_ptr & FLAG_MANY)) {
 355     if (is_in_progress(p)) return 0;  /* DFS back edge, i.e. we followed  */
 356                                       /* an edge to an object already     */
 357                                       /* on our stack: ignore             */
 358     push_in_progress(p);
 359     result = backwards_height(back_ptr)+1;
 360     pop_in_progress(p);
 361     return result;
 362   }
 363   be = (back_edges *)((word)back_ptr & ~FLAG_MANY);
 364   if (be -> height >= 0 && be -> height_gc_no == GC_gc_no)
 365       return be -> height;
 366   /* Ignore back edges in DFS */
 367     if (be -> height == HEIGHT_IN_PROGRESS) return 0;
 368   result = (be -> height > 0? be -> height : 1);
 369   be -> height = HEIGHT_IN_PROGRESS;
 370   FOR_EACH_PRED(q, p, {
 371     word this_height;
 372     if (GC_is_marked(q) && !(FLAG_MANY & (word)GET_OH_BG_PTR(p))) {
 373       if (GC_print_stats)
 374           GC_printf2("Found bogus pointer from 0x%lx to 0x%lx\n", q, p);
 375         /* Reachable object "points to" unreachable one.                */
 376         /* Could be caused by our lax treatment of GC descriptors.      */
 377       this_height = 1;
 378     } else {
 379         this_height = backwards_height(q);
 380     }
 381     if (this_height >= result) result = this_height + 1;
 382   });
 383   be -> height = result;
 384   be -> height_gc_no = GC_gc_no;
 385   return result;
 386 }
 387
 388 word GC_max_height;
 389 ptr_t GC_deepest_obj;
 390
 391 /* Compute the maximum height of every unreachable predecessor p of  a  */
 392 /* reachable object.  Arrange to save the heights of all such objects p */
 393 /* so that they can be used in calculating the height of objects in the */
 394 /* next GC.                                                             */
 395 /* Set GC_max_height to be the maximum height we encounter, and         */
 396 /* GC_deepest_obj to be the corresponding object.                       */
 397 static void update_max_height(ptr_t p, word n_words, word gc_descr)
 398 {
 399   if (GC_is_marked(p) && GC_HAS_DEBUG_INFO(p)) {
 400     int i;
 401     word p_height = 0;
 402     ptr_t p_deepest_obj = 0;
 403     ptr_t back_ptr;
 404     back_edges *be = 0;
 405
 406     /* If we remembered a height last time, use it as a minimum.        */
 407     /* It may have increased due to newly unreachable chains pointing   */
 408     /* to p, but it can't have decreased.                               */
 409     back_ptr = GET_OH_BG_PTR(p);
 410     if (0 != back_ptr && ((word)back_ptr & FLAG_MANY)) {
 411       be = (back_edges *)((word)back_ptr & ~FLAG_MANY);
 412       if (be -> height != HEIGHT_UNKNOWN) p_height = be -> height;
 413     }
 414     FOR_EACH_PRED(q, p, {
 415       if (!GC_is_marked(q) && GC_HAS_DEBUG_INFO(q)) {
 416         word q_height;
 417
 418         q_height = backwards_height(q);
 419         if (q_height > p_height) {
 420           p_height = q_height;
 421           p_deepest_obj = q;
 422         }
 423       }
 424     });
 425     if (p_height > 0) {
 426       /* Remember the height for next time. */
 427         if (be == 0) {
 428           ensure_struct(p);
 429           back_ptr = GET_OH_BG_PTR(p);
 430           be = (back_edges *)((word)back_ptr & ~FLAG_MANY);
 431         }
 432         be -> flags |= RETAIN;
 433         be -> height = p_height;
 434         be -> height_gc_no = GC_gc_no;
 435     }
 436     if (p_height > GC_max_height) {
 437         GC_max_height = p_height;
 438         GC_deepest_obj = p_deepest_obj;
 439     }
 440   }
 441 }
 442
 443 word GC_max_max_height = 0;
 444
 445 void GC_traverse_back_graph(void)
 446 {
 447   GC_max_height = 0;
 448   GC_apply_to_each_object(update_max_height);
 449 }
 450
 451 void GC_print_back_graph_stats(void)
 452 {
 453   GC_printf2("Maximum backwards height of reachable objects at GC %lu is %ld\n",
 454              (unsigned long) GC_gc_no, GC_max_height);
 455   if (GC_max_height > GC_max_max_height) {
 456     GC_max_max_height = GC_max_height;
 457     GC_printf0("The following unreachable object is last in a longest chain "
 458                "of unreachable objects:\n");
 459     GC_print_heap_obj(GC_deepest_obj);
 460   }
 461   if (GC_print_stats) {
 462     GC_printf1("Needed max total of %ld back-edge structs\n",
 463                GC_n_back_edge_structs);
 464   }
 465   GC_apply_to_each_object(reset_back_edge);
 466   GC_deepest_obj = 0;
 467 }
 468
 469 #endif /* MAKE_BACK_GRAPH */