gcc/stringpool.c

   1 /* String pool for GCC.
   2    Copyright (C) 2000-2015 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 "ggc.h"
  32 #include "ggc-internal.h"
  33 #include "hash-set.h"
  34 #include "vec.h"
  35 #include "input.h"
  36 #include "alias.h"
  37 #include "symtab.h"
  38 #include "options.h"
  39 #include "inchash.h"
  40 #include "tree.h"
  41 #include "cpplib.h"
  42
  43 /* The "" allocated string.  */
  44 const char empty_string[] = "";
  45
  46 /* Character strings, each containing a single decimal digit.
  47    Written this way to save space.  */
  48 static const char digit_vector[] = {
  49   '0', 0, '1', 0, '2', 0, '3', 0, '4', 0,
  50   '5', 0, '6', 0, '7', 0, '8', 0, '9', 0
  51 };
  52
  53 #define digit_string(d) (digit_vector + ((d) * 2))
  54
  55 struct ht *ident_hash;
  56
  57 static hashnode alloc_node (cpp_hash_table *);
  58 static int mark_ident (struct cpp_reader *, hashnode, const void *);
  59
  60 static void *
  61 stringpool_ggc_alloc (size_t x)
  62 {
  63   return ggc_alloc_atomic (x);
  64 }
  65
  66 /* Initialize the string pool.  */
  67 void
  68 init_stringpool (void)
  69 {
  70   /* Clean up if we're called more than once.
  71      (We can't make this idempotent since identifiers contain state) */
  72   if (ident_hash)
  73     ht_destroy (ident_hash);
  74
  75   /* Create with 16K (2^14) entries.  */
  76   ident_hash = ht_create (14);
  77   ident_hash->alloc_node = alloc_node;
  78   ident_hash->alloc_subobject = stringpool_ggc_alloc;
  79 }
  80
  81 /* Allocate a hash node.  */
  82 static hashnode
  83 alloc_node (cpp_hash_table *table ATTRIBUTE_UNUSED)
  84 {
  85   return GCC_IDENT_TO_HT_IDENT (make_node (IDENTIFIER_NODE));
  86 }
  87
  88 /* Allocate and return a string constant of length LENGTH, containing
  89    CONTENTS.  If LENGTH is -1, CONTENTS is assumed to be a
  90    nul-terminated string, and the length is calculated using strlen.  */
  91
  92 const char *
  93 ggc_alloc_string (const char *contents, int length MEM_STAT_DECL)
  94 {
  95   char *result;
  96
  97   if (length == -1)
  98     length = strlen (contents);
  99
 100   if (length == 0)
 101     return empty_string;
 102   if (length == 1 && ISDIGIT (contents[0]))
 103     return digit_string (contents[0] - '0');
 104
 105   result = (char *) ggc_internal_cleared_alloc (length + 1 PASS_MEM_STAT);
 106   memcpy (result, contents, length);
 107   result[length] = '\0';
 108   return (const char *) result;
 109 }
 110
 111 /* Return an IDENTIFIER_NODE whose name is TEXT (a null-terminated string).
 112    If an identifier with that name has previously been referred to,
 113    the same node is returned this time.  */
 114
 115 #undef get_identifier
 116
 117 tree
 118 get_identifier (const char *text)
 119 {
 120   hashnode ht_node = ht_lookup (ident_hash,
 121                                 (const unsigned char *) text,
 122                                 strlen (text), HT_ALLOC);
 123
 124   /* ht_node can't be NULL here.  */
 125   return HT_IDENT_TO_GCC_IDENT (ht_node);
 126 }
 127
 128 /* Identical to get_identifier, except that the length is assumed
 129    known.  */
 130
 131 tree
 132 get_identifier_with_length (const char *text, size_t length)
 133 {
 134   hashnode ht_node = ht_lookup (ident_hash,
 135                                 (const unsigned char *) text,
 136                                 length, HT_ALLOC);
 137
 138   /* ht_node can't be NULL here.  */
 139   return HT_IDENT_TO_GCC_IDENT (ht_node);
 140 }
 141
 142 /* If an identifier with the name TEXT (a null-terminated string) has
 143    previously been referred to, return that node; otherwise return
 144    NULL_TREE.  */
 145
 146 tree
 147 maybe_get_identifier (const char *text)
 148 {
 149   hashnode ht_node;
 150
 151   ht_node = ht_lookup (ident_hash, (const unsigned char *) text,
 152                        strlen (text), HT_NO_INSERT);
 153   if (ht_node)
 154     return HT_IDENT_TO_GCC_IDENT (ht_node);
 155
 156   return NULL_TREE;
 157 }
 158
 159 /* Report some basic statistics about the string pool.  */
 160
 161 void
 162 stringpool_statistics (void)
 163 {
 164   ht_dump_statistics (ident_hash);
 165 }
 166 \f
 167 /* Mark an identifier for GC.  */
 168
 169 static int
 170 mark_ident (struct cpp_reader *pfile ATTRIBUTE_UNUSED, hashnode h,
 171             const void *v ATTRIBUTE_UNUSED)
 172 {
 173   gt_ggc_m_9tree_node (HT_IDENT_TO_GCC_IDENT (h));
 174   return 1;
 175 }
 176
 177 /* Return true if an identifier should be removed from the table.  */
 178
 179 static int
 180 maybe_delete_ident (struct cpp_reader *pfile ATTRIBUTE_UNUSED, hashnode h,
 181                     const void *v ATTRIBUTE_UNUSED)
 182 {
 183   return !ggc_marked_p (HT_IDENT_TO_GCC_IDENT (h));
 184 }
 185
 186 /* Mark the trees hanging off the identifier node for GGC.  These are
 187    handled specially (not using gengtype) because identifiers are only
 188    roots during one part of compilation.  */
 189
 190 void
 191 ggc_mark_stringpool (void)
 192 {
 193   ht_forall (ident_hash, mark_ident, NULL);
 194 }
 195
 196 /* Purge the identifier hash of identifiers which are no longer
 197    referenced.  */
 198
 199 void
 200 ggc_purge_stringpool (void)
 201 {
 202   ht_purge (ident_hash, maybe_delete_ident, NULL);
 203 }
 204
 205 /* Pointer-walking routine for strings (not very interesting, since
 206    strings don't contain pointers).  */
 207
 208 void
 209 gt_pch_p_S (void *obj ATTRIBUTE_UNUSED, void *x ATTRIBUTE_UNUSED,
 210             gt_pointer_operator op ATTRIBUTE_UNUSED,
 211             void *cookie ATTRIBUTE_UNUSED)
 212 {
 213 }
 214
 215 /* PCH pointer-walking routine for strings.  */
 216
 217 void
 218 gt_pch_n_S (const void *x)
 219 {
 220   gt_pch_note_object (CONST_CAST (void *, x), CONST_CAST (void *, x),
 221                       &gt_pch_p_S);
 222 }
 223
 224
 225 /* User-callable entry point for marking string X.  */
 226
 227 void
 228 gt_pch_nx (const char *& x)
 229 {
 230   gt_pch_n_S (x);
 231 }
 232
 233 void
 234 gt_pch_nx (unsigned char *& x)
 235 {
 236   gt_pch_n_S (x);
 237 }
 238
 239 void
 240 gt_pch_nx (unsigned char& x ATTRIBUTE_UNUSED)
 241 {
 242 }
 243
 244 void
 245 gt_pch_nx (unsigned char *x, gt_pointer_operator op, void *cookie)
 246 {
 247   op (x, cookie);
 248 }
 249 \f
 250 /* Handle saving and restoring the string pool for PCH.  */
 251
 252 /* SPD is saved in the PCH file and holds the information needed
 253    to restore the string pool.  */
 254
 255 struct GTY(()) string_pool_data {
 256   ht_identifier_ptr *
 257     GTY((length ("%h.nslots"),
 258          nested_ptr (union tree_node, "%h ? GCC_IDENT_TO_HT_IDENT (%h) : NULL",
 259                      "%h ? HT_IDENT_TO_GCC_IDENT (%h) : NULL")))
 260     entries;
 261   unsigned int nslots;
 262   unsigned int nelements;
 263 };
 264
 265 static GTY(()) struct string_pool_data * spd;
 266
 267 /* Save the stringpool data in SPD.  */
 268
 269 void
 270 gt_pch_save_stringpool (void)
 271 {
 272   spd = ggc_alloc<string_pool_data> ();
 273   spd->nslots = ident_hash->nslots;
 274   spd->nelements = ident_hash->nelements;
 275   spd->entries = ggc_vec_alloc<ht_identifier_ptr> (spd->nslots);
 276   memcpy (spd->entries, ident_hash->entries,
 277           spd->nslots * sizeof (spd->entries[0]));
 278 }
 279
 280 /* Return the stringpool to its state before gt_pch_save_stringpool
 281    was called.  */
 282
 283 void
 284 gt_pch_fixup_stringpool (void)
 285 {
 286 }
 287
 288 /* A PCH file has been restored, which loaded SPD; fill the real hash table
 289    from SPD.  */
 290
 291 void
 292 gt_pch_restore_stringpool (void)
 293 {
 294   ht_load (ident_hash, spd->entries, spd->nslots, spd->nelements, false);
 295   spd = NULL;
 296 }
 297
 298 #include "gt-stringpool.h"