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                         sbrk(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 /* This data structure NEEDS IMPROVEMENT.                               */
 119 #define MAX_IN_PROGRESS 10000
 120 static ptr_t * in_progress_space = 0;
 121 static int n_in_progress = 0;
 122
 123 static void push_in_progress(ptr_t p)
 124 {
 125   if (in_progress_space == 0)
 126       in_progress_space = sbrk(MAX_IN_PROGRESS * sizeof(ptr_t));
 127   if (n_in_progress == MAX_IN_PROGRESS)
 128       ABORT("Exceeded MAX_IN_PROGRESS");
 129   in_progress_space[n_in_progress++] = p;
 130 }
 131
 132 static GC_bool is_in_progress(ptr_t p)
 133 {
 134   int i;
 135   for (i = 0; i < n_in_progress; ++i) {
 136     if (in_progress_space[i] == p) return TRUE;
 137   }
 138   return FALSE;
 139 }
 140
 141 static void pop_in_progress(ptr_t p)
 142 {
 143   --n_in_progress;
 144   GC_ASSERT(in_progress_space[n_in_progress] == p);
 145 }
 146
 147 #define GET_OH_BG_PTR(p) \
 148                 (ptr_t)REVEAL_POINTER(((oh *)(p)) -> oh_bg_ptr)
 149 #define SET_OH_BG_PTR(p,q) (((oh *)(p)) -> oh_bg_ptr) = HIDE_POINTER(q)
 150
 151 /* Execute s once for each predecessor q of p in the points-to graph.   */
 152 /* s should be a bracketed statement.  We declare q.                    */
 153 #define FOR_EACH_PRED(q, p, s) \
 154   { \
 155     ptr_t q = GET_OH_BG_PTR(p); \
 156     if (!((word)q & FLAG_MANY)) { \
 157       if (q && !((word)q & 1)) s \
 158               /* !((word)q & 1) checks for a misnterpreted freelist link */ \
 159     } else { \
 160       back_edges *orig_be_ = (back_edges *)((word)q & ~FLAG_MANY); \
 161       back_edges *be_ = orig_be_; \
 162       int total_, local_; \
 163       int n_edges_ = be_ -> n_edges; \
 164       for (total_ = 0, local_ = 0; total_ < n_edges_; ++local_, ++total_) { \
 165           if (local_ == MAX_IN) { \
 166               be_ = be_ -> cont; \
 167               local_ = 0; \
 168           } \
 169           q = be_ -> edges[local_]; s \
 170       } \
 171     } \
 172   }
 173
 174 /* Ensure that p has a back_edges structure associated with it. */
 175 static void ensure_struct(ptr_t p)
 176 {
 177   ptr_t old_back_ptr = GET_OH_BG_PTR(p);
 178
 179   if (!((word)old_back_ptr & FLAG_MANY)) {
 180     back_edges *be = new_back_edges();
 181     be -> flags = 0;
 182     if (0 == old_back_ptr) {
 183       be -> n_edges = 0;
 184     } else {
 185       be -> n_edges = 1;
 186       be -> edges[0] = old_back_ptr;
 187     }
 188     be -> height = HEIGHT_UNKNOWN;
 189     be -> height_gc_no = GC_gc_no - 1;
 190     GC_ASSERT(be >= back_edge_space);
 191     SET_OH_BG_PTR(p, (word)be | FLAG_MANY);
 192   }
 193 }
 194
 195 /* Add the (forward) edge from p to q to the backward graph.  Both p    */
 196 /* q are pointers to the object base, i.e. pointers to an oh.           */
 197 static void add_edge(ptr_t p,  ptr_t q)
 198 {
 199     ptr_t old_back_ptr = GET_OH_BG_PTR(q);
 200     back_edges * be, *be_cont;
 201     word i;
 202     static unsigned random_number = 13;
 203 #   define GOT_LUCKY_NUMBER (((++random_number) & 0x7f) == 0)
 204       /* A not very random number we use to occasionally allocate a     */
 205       /* back_edges structure even for a single backward edge.  This    */
 206       /* prevents us from repeatedly tracing back through very long     */
 207       /* chains, since we will have some place to store height and      */
 208       /* in_progress flags along the way.                               */
 209
 210     GC_ASSERT(p == GC_base(p) && q == GC_base(q));
 211     if (!GC_HAS_DEBUG_INFO(q) || !GC_HAS_DEBUG_INFO(p)) {
 212       /* This is really a misinterpreted free list link, since we saw */
 213       /* a pointer to a free list.  Dont overwrite it!                */
 214       return;
 215     }
 216     if (0 == old_back_ptr) {
 217         SET_OH_BG_PTR(q, p);
 218         if (GOT_LUCKY_NUMBER) ensure_struct(q);
 219         return;
 220     }
 221     /* Check whether it was already in the list of predecessors. */
 222       FOR_EACH_PRED(pred, q, { if (p == pred) return; });
 223     ensure_struct(q);
 224     old_back_ptr = GET_OH_BG_PTR(q);
 225     be = (back_edges *)((word)old_back_ptr & ~FLAG_MANY);
 226     for (i = be -> n_edges, be_cont = be; i > MAX_IN;
 227         be_cont = be_cont -> cont, i -= MAX_IN) {}
 228     if (i == MAX_IN) {
 229         be_cont -> cont = new_back_edges();
 230         be_cont = be_cont -> cont;
 231         i = 0;
 232     }
 233     be_cont -> edges[i] = p;
 234     be -> n_edges++;
 235     if (be -> n_edges == 100) {
 236 #       if 0
 237           if (GC_print_stats) {
 238             GC_err_printf0("The following object has in-degree >= 100:\n");
 239             GC_print_heap_obj(q);
 240           }
 241 #       endif
 242     }
 243 }
 244
 245 typedef void (*per_object_func)(ptr_t p, word n_words, word gc_descr);
 246
 247 static void per_object_helper(struct hblk *h, word fn)
 248 {
 249   hdr * hhdr = HDR(h);
 250   word sz = hhdr -> hb_sz;
 251   word descr = hhdr -> hb_descr;
 252   per_object_func f = (per_object_func)fn;
 253   int i = 0;
 254
 255   do {
 256     f((ptr_t)(h -> hb_body + i), sz, descr);
 257     i += sz;
 258   } while (i + sz <= BYTES_TO_WORDS(HBLKSIZE));
 259 }
 260
 261 void GC_apply_to_each_object(per_object_func f)
 262 {
 263   GC_apply_to_all_blocks(per_object_helper, (word)f);
 264 }
 265
 266 static void reset_back_edge(ptr_t p, word n_words, word gc_descr)
 267 {
 268   /* Skip any free list links, or dropped blocks */
 269   if (GC_HAS_DEBUG_INFO(p)) {
 270     ptr_t old_back_ptr = GET_OH_BG_PTR(p);
 271     if ((word)old_back_ptr & FLAG_MANY) {
 272       back_edges *be = (back_edges *)((word)old_back_ptr & ~FLAG_MANY);
 273       if (!(be -> flags & RETAIN)) {
 274         deallocate_back_edges(be);
 275         SET_OH_BG_PTR(p, 0);
 276       } else {
 277         word *currentp;
 278
 279         GC_ASSERT(GC_is_marked(p));
 280
 281         /* Back edges may point to objects that will not be retained.   */
 282         /* Delete them for now, but remember the height.                */
 283         /* Some will be added back at next GC.                          */
 284           be -> n_edges = 0;
 285           if (0 != be -> cont) {
 286             deallocate_back_edges(be -> cont);
 287             be -> cont = 0;
 288           }
 289
 290         GC_ASSERT(GC_is_marked(p));
 291
 292         /* We only retain things for one GC cycle at a time.            */
 293           be -> flags &= ~RETAIN;
 294       }
 295     } else /* Simple back pointer */ {
 296       /* Clear to avoid dangling pointer. */
 297       SET_OH_BG_PTR(p, 0);
 298     }
 299   }
 300 }
 301
 302 static void add_back_edges(ptr_t p, word n_words, word gc_descr)
 303 {
 304   word *currentp = (word *)(p + sizeof(oh));
 305
 306   /* For now, fix up non-length descriptors conservatively.     */
 307     if((gc_descr & GC_DS_TAGS) != GC_DS_LENGTH) {
 308       gc_descr = WORDS_TO_BYTES(n_words);
 309     }
 310   while (currentp < (word *)(p + gc_descr)) {
 311     word current = *currentp++;
 312     FIXUP_POINTER(current);
 313     if (current >= (word)GC_least_plausible_heap_addr &&
 314         current <= (word)GC_greatest_plausible_heap_addr) {
 315        ptr_t target = GC_base((GC_PTR)current);
 316        if (0 != target) {
 317          add_edge(p, target);
 318        }
 319     }
 320   }
 321 }
 322
 323 /* Rebuild the reprentation of the backward reachability graph. */
 324 /* Does not examine mark bits.  Can be called before GC.        */
 325 void GC_build_back_graph(void)
 326 {
 327   GC_apply_to_each_object(add_back_edges);
 328 }
 329
 330 /* Return an approximation to the length of the longest simple path     */
 331 /* through unreachable objects to p.  We refer to this as the height    */
 332 /* of p.                                                                */
 333 static word backwards_height(ptr_t p)
 334 {
 335   word result;
 336   ptr_t back_ptr = GET_OH_BG_PTR(p);
 337   back_edges *be;
 338
 339   if (0 == back_ptr) return 1;
 340   if (!((word)back_ptr & FLAG_MANY)) {
 341     if (is_in_progress(p)) return 0;  /* DFS back edge, i.e. we followed  */
 342                                       /* an edge to an object already     */
 343                                       /* on our stack: ignore             */
 344     push_in_progress(p);
 345     result = backwards_height(back_ptr)+1;
 346     pop_in_progress(p);
 347     return result;
 348   }
 349   be = (back_edges *)((word)back_ptr & ~FLAG_MANY);
 350   if (be -> height >= 0 && be -> height_gc_no == GC_gc_no)
 351       return be -> height;
 352   /* Ignore back edges in DFS */
 353     if (be -> height == HEIGHT_IN_PROGRESS) return 0;
 354   result = (be -> height > 0? be -> height : 1);
 355   be -> height = HEIGHT_IN_PROGRESS;
 356   FOR_EACH_PRED(q, p, {
 357     word this_height;
 358     if (GC_is_marked(q) && !(FLAG_MANY & (word)GET_OH_BG_PTR(p))) {
 359       if (GC_print_stats)
 360           GC_printf2("Found bogus pointer from 0x%lx to 0x%lx\n", q, p);
 361         /* Reachable object "points to" unreachable one.                */
 362         /* Could be caused by our lax treatment of GC descriptors.      */
 363       this_height = 1;
 364     } else {
 365         this_height = backwards_height(q);
 366     }
 367     if (this_height >= result) result = this_height + 1;
 368   });
 369   be -> height = result;
 370   be -> height_gc_no = GC_gc_no;
 371   return result;
 372 }
 373
 374 word GC_max_height;
 375 ptr_t GC_deepest_obj;
 376
 377 /* Compute the maximum height of every unreachable predecessor p of  a  */
 378 /* reachable object.  Arrange to save the heights of all such objects p */
 379 /* so that they can be used in calculating the height of objects in the */
 380 /* next GC.                                                             */
 381 /* Set GC_max_height to be the maximum height we encounter, and         */
 382 /* GC_deepest_obj to be the corresponding object.                       */
 383 static void update_max_height(ptr_t p, word n_words, word gc_descr)
 384 {
 385   if (GC_is_marked(p) && GC_HAS_DEBUG_INFO(p)) {
 386     int i;
 387     word p_height = 0;
 388     ptr_t p_deepest_obj = 0;
 389     ptr_t back_ptr;
 390     back_edges *be = 0;
 391
 392     /* If we remembered a height last time, use it as a minimum.        */
 393     /* It may have increased due to newly unreachable chains pointing   */
 394     /* to p, but it can't have decreased.                               */
 395     back_ptr = GET_OH_BG_PTR(p);
 396     if (0 != back_ptr && ((word)back_ptr & FLAG_MANY)) {
 397       be = (back_edges *)((word)back_ptr & ~FLAG_MANY);
 398       if (be -> height != HEIGHT_UNKNOWN) p_height = be -> height;
 399     }
 400     FOR_EACH_PRED(q, p, {
 401       if (!GC_is_marked(q) && GC_HAS_DEBUG_INFO(q)) {
 402         word q_height;
 403
 404         q_height = backwards_height(q);
 405         if (q_height > p_height) {
 406           p_height = q_height;
 407           p_deepest_obj = q;
 408         }
 409       }
 410     });
 411     if (p_height > 0) {
 412       /* Remember the height for next time. */
 413         if (be == 0) {
 414           ensure_struct(p);
 415           back_ptr = GET_OH_BG_PTR(p);
 416           be = (back_edges *)((word)back_ptr & ~FLAG_MANY);
 417         }
 418         be -> flags |= RETAIN;
 419         be -> height = p_height;
 420         be -> height_gc_no = GC_gc_no;
 421     }
 422     if (p_height > GC_max_height) {
 423         GC_max_height = p_height;
 424         GC_deepest_obj = p_deepest_obj;
 425     }
 426   }
 427 }
 428
 429 void GC_traverse_back_graph(void)
 430 {
 431   static word max_max_height = 0;
 432   GC_max_height = 0;
 433   GC_apply_to_each_object(update_max_height);
 434   GC_printf2("Maximum backwards height of reachable objects at GC %lu is %ld\n",
 435              (unsigned long) GC_gc_no, GC_max_height);
 436   if (GC_max_height > max_max_height) {
 437     max_max_height = GC_max_height;
 438     GC_printf0("The following unreachable object is last in a longest chain "
 439                "of unreachable objects:\n");
 440     GC_print_heap_obj(GC_deepest_obj);
 441   }
 442   if (GC_print_stats) {
 443     GC_printf1("Needed max total of %ld back-edge structs\n",
 444                GC_n_back_edge_structs);
 445   }
 446   GC_apply_to_each_object(reset_back_edge);
 447   GC_deepest_obj = 0;
 448 }
 449
 450 #endif /* MAKE_BACK_GRAPH */