gcc/stringpool.cc

   1 /* String pool for GCC.
   2    Copyright (C) 2000-2024 Free Software Foundation, Inc.
   3
   4 This file is part of GCC.
   5
   6 GCC is free software; you can redistribute it and/or modify it under
   7 the terms of the GNU General Public License as published by the Free
   8 Software Foundation; either version 3, or (at your option) any later
   9 version.
  10
  11 GCC is distributed in the hope that it will be useful, but WITHOUT ANY
  12 WARRANTY; without even the implied warranty of MERCHANTABILITY or
  13 FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
  14 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
  20 /* String text, identifier text and identifier node allocator.
  21    Identifiers are uniquely stored in a hash table.
  22
  23    We use cpplib's hash table implementation.  libiberty's
  24    hashtab.c is not used because it requires 100% average space
  25    overhead per string, which is unacceptable.  Also, this algorithm
  26    is faster.  */
  27
  28 #include "config.h"
  29 #include "system.h"
  30 #include "coretypes.h"
  31 #include "tree.h"
  32 #include "cpplib.h"
  33
  34 struct ht *ident_hash;
  35 struct ht *ident_hash_extra;
  36
  37 static hashnode alloc_node (cpp_hash_table *);
  38 static int mark_ident (struct cpp_reader *, hashnode, const void *);
  39
  40 static void *
  41 stringpool_ggc_alloc (size_t x)
  42 {
  43   return ggc_alloc_atomic (x);
  44 }
  45
  46 /* Initialize the string pool.  */
  47 void
  48 init_stringpool (void)
  49 {
  50   /* Clean up if we're called more than once.
  51      (We can't make this idempotent since identifiers contain state) */
  52   if (ident_hash)
  53     ht_destroy (ident_hash);
  54   if (ident_hash_extra)
  55     ht_destroy (ident_hash_extra);
  56
  57   /* Create with 16K (2^14) entries.  */
  58   ident_hash = ht_create (14);
  59   ident_hash->alloc_node = alloc_node;
  60   ident_hash->alloc_subobject = stringpool_ggc_alloc;
  61
  62   /* Create with 64 (2^6) entries.  */
  63   ident_hash_extra = ht_create (6);
  64   ident_hash_extra->alloc_node = [] (cpp_hash_table *)
  65   {
  66     return HT_NODE (ggc_cleared_alloc<cpp_hashnode_extra> ());
  67   };
  68   ident_hash_extra->alloc_subobject = stringpool_ggc_alloc;
  69 }
  70
  71 /* Allocate a hash node.  */
  72 static hashnode
  73 alloc_node (cpp_hash_table *table ATTRIBUTE_UNUSED)
  74 {
  75   return GCC_IDENT_TO_HT_IDENT (make_node (IDENTIFIER_NODE));
  76 }
  77
  78 /* Allocate and return a string constant of length LENGTH, containing
  79    CONTENTS.  If LENGTH is -1, CONTENTS is assumed to be a
  80    nul-terminated string, and the length is calculated using strlen.  */
  81
  82 const char *
  83 ggc_alloc_string (const char *contents, int length MEM_STAT_DECL)
  84 {
  85   if (length == -1)
  86     length = strlen (contents);
  87
  88   if (!length)
  89     return "";
  90
  91   char *result = (char *) ggc_alloc_atomic (length + 1);
  92   memcpy (result, contents, length);
  93   result[length] = '\0';
  94
  95   return (const char *) result;
  96 }
  97
  98 /* Return an IDENTIFIER_NODE whose name is TEXT (a null-terminated string).
  99    If an identifier with that name has previously been referred to,
 100    the same node is returned this time.  */
 101
 102 #undef get_identifier
 103
 104 tree
 105 get_identifier (const char *text)
 106 {
 107   hashnode ht_node = ht_lookup (ident_hash,
 108                                 (const unsigned char *) text,
 109                                 strlen (text), HT_ALLOC);
 110
 111   /* ht_node can't be NULL here.  */
 112   return HT_IDENT_TO_GCC_IDENT (ht_node);
 113 }
 114
 115 /* Identical to get_identifier, except that the length is assumed
 116    known.  */
 117
 118 tree
 119 get_identifier_with_length (const char *text, size_t length)
 120 {
 121   hashnode ht_node = ht_lookup (ident_hash,
 122                                 (const unsigned char *) text,
 123                                 length, HT_ALLOC);
 124
 125   /* ht_node can't be NULL here.  */
 126   return HT_IDENT_TO_GCC_IDENT (ht_node);
 127 }
 128
 129 /* If an identifier with the name TEXT (a null-terminated string) has
 130    previously been referred to, return that node; otherwise return
 131    NULL_TREE.  */
 132
 133 tree
 134 maybe_get_identifier (const char *text)
 135 {
 136   hashnode ht_node;
 137
 138   ht_node = ht_lookup (ident_hash, (const unsigned char *) text,
 139                        strlen (text), HT_NO_INSERT);
 140   if (ht_node)
 141     return HT_IDENT_TO_GCC_IDENT (ht_node);
 142
 143   return NULL_TREE;
 144 }
 145
 146 /* Report some basic statistics about the string pool.  */
 147
 148 void
 149 stringpool_statistics (void)
 150 {
 151   ht_dump_statistics (ident_hash);
 152 }
 153 \f
 154 /* Mark an identifier for GC.  */
 155
 156 static int
 157 mark_ident (struct cpp_reader *pfile ATTRIBUTE_UNUSED, hashnode h,
 158             const void *v ATTRIBUTE_UNUSED)
 159 {
 160   gt_ggc_m_9tree_node (HT_IDENT_TO_GCC_IDENT (h));
 161   return 1;
 162 }
 163
 164 /* Return true if an identifier should be removed from the table.  */
 165
 166 static int
 167 maybe_delete_ident (struct cpp_reader *pfile ATTRIBUTE_UNUSED, hashnode h,
 168                     const void *v ATTRIBUTE_UNUSED)
 169 {
 170   return !ggc_marked_p (HT_IDENT_TO_GCC_IDENT (h));
 171 }
 172
 173 /* Mark the trees hanging off the identifier node for GGC.  These are
 174    handled specially (not using gengtype) because identifiers are only
 175    roots during one part of compilation.  */
 176
 177 void
 178 ggc_mark_stringpool (void)
 179 {
 180   ht_forall (ident_hash, mark_ident, NULL);
 181   ht_forall (ident_hash_extra,
 182              [] (cpp_reader *, hashnode h, const void *)
 183              {
 184                gt_ggc_m_18cpp_hashnode_extra (h);
 185                return 1;
 186              }, nullptr);
 187 }
 188
 189 /* Purge the identifier hash of identifiers which are no longer
 190    referenced.  */
 191
 192 void
 193 ggc_purge_stringpool (void)
 194 {
 195   ht_purge (ident_hash, maybe_delete_ident, NULL);
 196   ht_purge (ident_hash_extra,
 197             [] (cpp_reader *, hashnode h, const void *) -> int
 198             {
 199               return !ggc_marked_p (h);
 200             }, nullptr);
 201 }
 202
 203 /* Pointer-walking routine for strings (not very interesting, since
 204    strings don't contain pointers).  */
 205
 206 void
 207 gt_pch_p_S (void *obj ATTRIBUTE_UNUSED, void *x ATTRIBUTE_UNUSED,
 208             gt_pointer_operator op ATTRIBUTE_UNUSED,
 209             void *cookie ATTRIBUTE_UNUSED)
 210 {
 211   gcc_unreachable ();
 212 }
 213
 214 /* PCH pointer-walking routine for strings.  */
 215
 216 void
 217 gt_pch_n_S (const void *x)
 218 {
 219   gt_pch_note_object (CONST_CAST (void *, x), CONST_CAST (void *, x),
 220                       &gt_pch_p_S);
 221 }
 222
 223 void
 224 gt_pch_n_S2 (const void *x, size_t string_len)
 225 {
 226   gt_pch_note_object (CONST_CAST (void *, x), CONST_CAST (void *, x),
 227                       &gt_pch_p_S, string_len);
 228 }
 229
 230
 231 /* User-callable entry point for marking string X.  */
 232
 233 void
 234 gt_pch_nx (const char *& x)
 235 {
 236   gt_pch_n_S (x);
 237 }
 238
 239 void
 240 gt_pch_nx (char *& x)
 241 {
 242   gt_pch_n_S (x);
 243 }
 244
 245 void
 246 gt_pch_nx (unsigned char *& x)
 247 {
 248   gt_pch_n_S (x);
 249 }
 250
 251 void
 252 gt_pch_nx (unsigned char& x ATTRIBUTE_UNUSED)
 253 {
 254 }
 255
 256 void
 257 gt_pch_nx (unsigned char *x, gt_pointer_operator op, void *cookie)
 258 {
 259   op (x, NULL, cookie);
 260 }
 261 \f
 262 /* Handle saving and restoring the string pool for PCH.  */
 263
 264 /* SPD is saved in the PCH file and holds the information needed
 265    to restore the string pool.  */
 266
 267 struct GTY(()) string_pool_data {
 268   ht_identifier_ptr *
 269     GTY((length ("%h.nslots"),
 270          nested_ptr (union tree_node, "%h ? GCC_IDENT_TO_HT_IDENT (%h) : NULL",
 271                      "%h ? HT_IDENT_TO_GCC_IDENT (%h) : NULL")))
 272     entries;
 273   unsigned int nslots;
 274   unsigned int nelements;
 275 };
 276
 277 struct GTY (()) string_pool_data_extra
 278 {
 279   ht_identifier_ptr *
 280     GTY((length ("%h.nslots"),
 281          nested_ptr (cpp_hashnode_extra, "%h ? HT_NODE (%h) : nullptr",
 282                      "(cpp_hashnode_extra *)%h")))
 283     entries;
 284   unsigned int nslots;
 285   unsigned int nelements;
 286 };
 287
 288 static GTY(()) struct string_pool_data * spd;
 289 static GTY(()) struct string_pool_data_extra *spd2;
 290
 291 /* Save the stringpool data in SPD.  */
 292
 293 void
 294 gt_pch_save_stringpool (void)
 295 {
 296   spd = ggc_alloc<string_pool_data> ();
 297   spd->nslots = ident_hash->nslots;
 298   spd->nelements = ident_hash->nelements;
 299   spd->entries = ggc_vec_alloc<ht_identifier_ptr> (spd->nslots);
 300   memcpy (spd->entries, ident_hash->entries,
 301           spd->nslots * sizeof (spd->entries[0]));
 302
 303   spd2 = ggc_alloc<string_pool_data_extra> ();
 304   spd2->nslots = ident_hash_extra->nslots;
 305   spd2->nelements = ident_hash_extra->nelements;
 306   spd2->entries = ggc_vec_alloc<ht_identifier_ptr> (spd2->nslots);
 307   memcpy (spd2->entries, ident_hash_extra->entries,
 308           spd2->nslots * sizeof (spd2->entries[0]));
 309 }
 310
 311 /* Return the stringpool to its state before gt_pch_save_stringpool
 312    was called.  */
 313
 314 void
 315 gt_pch_fixup_stringpool (void)
 316 {
 317 }
 318
 319 /* A PCH file has been restored, which loaded SPD; fill the real hash table
 320    from SPD.  */
 321
 322 void
 323 gt_pch_restore_stringpool (void)
 324 {
 325   ht_load (ident_hash, spd->entries, spd->nslots, spd->nelements, false);
 326   ht_load (ident_hash_extra, spd2->entries, spd2->nslots, spd2->nelements,
 327            false);
 328   spd = NULL;
 329   spd2 = NULL;
 330 }
 331
 332 #include "gt-stringpool.h"