pack-revindex.c

   1 #include "cache.h"
   2 #include "pack-revindex.h"
   3
   4 /*
   5  * Pack index for existing packs give us easy access to the offsets into
   6  * corresponding pack file where each object's data starts, but the entries
   7  * do not store the size of the compressed representation (uncompressed
   8  * size is easily available by examining the pack entry header).  It is
   9  * also rather expensive to find the sha1 for an object given its offset.
  10  *
  11  * We build a hashtable of existing packs (pack_revindex), and keep reverse
  12  * index here -- pack index file is sorted by object name mapping to offset;
  13  * this pack_revindex[].revindex array is a list of offset/index_nr pairs
  14  * ordered by offset, so if you know the offset of an object, next offset
  15  * is where its packed representation ends and the index_nr can be used to
  16  * get the object sha1 from the main index.
  17  */
  18
  19 struct pack_revindex {
  20         struct packed_git *p;
  21         struct revindex_entry *revindex;
  22 };
  23
  24 static struct pack_revindex *pack_revindex;
  25 static int pack_revindex_hashsz;
  26
  27 static int pack_revindex_ix(struct packed_git *p)
  28 {
  29         unsigned long ui = (unsigned long)p;
  30         int i;
  31
  32         ui = ui ^ (ui >> 16); /* defeat structure alignment */
  33         i = (int)(ui % pack_revindex_hashsz);
  34         while (pack_revindex[i].p) {
  35                 if (pack_revindex[i].p == p)
  36                         return i;
  37                 if (++i == pack_revindex_hashsz)
  38                         i = 0;
  39         }
  40         return -1 - i;
  41 }
  42
  43 static void init_pack_revindex(void)
  44 {
  45         int num;
  46         struct packed_git *p;
  47
  48         for (num = 0, p = packed_git; p; p = p->next)
  49                 num++;
  50         if (!num)
  51                 return;
  52         pack_revindex_hashsz = num * 11;
  53         pack_revindex = xcalloc(sizeof(*pack_revindex), pack_revindex_hashsz);
  54         for (p = packed_git; p; p = p->next) {
  55                 num = pack_revindex_ix(p);
  56                 num = - 1 - num;
  57                 pack_revindex[num].p = p;
  58         }
  59         /* revindex elements are lazily initialized */
  60 }
  61
  62 /*
  63  * This is a least-significant-digit radix sort.
  64  *
  65  * It sorts each of the "n" items in "entries" by its offset field. The "max"
  66  * parameter must be at least as large as the largest offset in the array,
  67  * and lets us quit the sort early.
  68  */
  69 static void sort_revindex(struct revindex_entry *entries, unsigned n, off_t max)
  70 {
  71         /*
  72          * We use a "digit" size of 16 bits. That keeps our memory
  73          * usage reasonable, and we can generally (for a 4G or smaller
  74          * packfile) quit after two rounds of radix-sorting.
  75          */
  76 #define DIGIT_SIZE (16)
  77 #define BUCKETS (1 << DIGIT_SIZE)
  78         /*
  79          * We want to know the bucket that a[i] will go into when we are using
  80          * the digit that is N bits from the (least significant) end.
  81          */
  82 #define BUCKET_FOR(a, i, bits) (((a)[(i)].offset >> (bits)) & (BUCKETS-1))
  83
  84         /*
  85          * We need O(n) temporary storage. Rather than do an extra copy of the
  86          * partial results into "entries", we sort back and forth between the
  87          * real array and temporary storage. In each iteration of the loop, we
  88          * keep track of them with alias pointers, always sorting from "from"
  89          * to "to".
  90          */
  91         struct revindex_entry *tmp = xmalloc(n * sizeof(*tmp));
  92         struct revindex_entry *from = entries, *to = tmp;
  93         int bits;
  94         unsigned *pos = xmalloc(BUCKETS * sizeof(*pos));
  95
  96         /*
  97          * If (max >> bits) is zero, then we know that the radix digit we are
  98          * on (and any higher) will be zero for all entries, and our loop will
  99          * be a no-op, as everybody lands in the same zero-th bucket.
 100          */
 101         for (bits = 0; max >> bits; bits += DIGIT_SIZE) {
 102                 struct revindex_entry *swap;
 103                 unsigned i;
 104
 105                 memset(pos, 0, BUCKETS * sizeof(*pos));
 106
 107                 /*
 108                  * We want pos[i] to store the index of the last element that
 109                  * will go in bucket "i" (actually one past the last element).
 110                  * To do this, we first count the items that will go in each
 111                  * bucket, which gives us a relative offset from the last
 112                  * bucket. We can then cumulatively add the index from the
 113                  * previous bucket to get the true index.
 114                  */
 115                 for (i = 0; i < n; i++)
 116                         pos[BUCKET_FOR(from, i, bits)]++;
 117                 for (i = 1; i < BUCKETS; i++)
 118                         pos[i] += pos[i-1];
 119
 120                 /*
 121                  * Now we can drop the elements into their correct buckets (in
 122                  * our temporary array).  We iterate the pos counter backwards
 123                  * to avoid using an extra index to count up. And since we are
 124                  * going backwards there, we must also go backwards through the
 125                  * array itself, to keep the sort stable.
 126                  *
 127                  * Note that we use an unsigned iterator to make sure we can
 128                  * handle 2^32-1 objects, even on a 32-bit system. But this
 129                  * means we cannot use the more obvious "i >= 0" loop condition
 130                  * for counting backwards, and must instead check for
 131                  * wrap-around with UINT_MAX.
 132                  */
 133                 for (i = n - 1; i != UINT_MAX; i--)
 134                         to[--pos[BUCKET_FOR(from, i, bits)]] = from[i];
 135
 136                 /*
 137                  * Now "to" contains the most sorted list, so we swap "from" and
 138                  * "to" for the next iteration.
 139                  */
 140                 swap = from;
 141                 from = to;
 142                 to = swap;
 143         }
 144
 145         /*
 146          * If we ended with our data in the original array, great. If not,
 147          * we have to move it back from the temporary storage.
 148          */
 149         if (from != entries)
 150                 memcpy(entries, tmp, n * sizeof(*entries));
 151         free(tmp);
 152         free(pos);
 153
 154 #undef BUCKET_FOR
 155 #undef BUCKETS
 156 #undef DIGIT_SIZE
 157 }
 158
 159 /*
 160  * Ordered list of offsets of objects in the pack.
 161  */
 162 static void create_pack_revindex(struct pack_revindex *rix)
 163 {
 164         struct packed_git *p = rix->p;
 165         unsigned num_ent = p->num_objects;
 166         unsigned i;
 167         const char *index = p->index_data;
 168
 169         rix->revindex = xmalloc(sizeof(*rix->revindex) * (num_ent + 1));
 170         index += 4 * 256;
 171
 172         if (p->index_version > 1) {
 173                 const uint32_t *off_32 =
 174                         (uint32_t *)(index + 8 + p->num_objects * (20 + 4));
 175                 const uint32_t *off_64 = off_32 + p->num_objects;
 176                 for (i = 0; i < num_ent; i++) {
 177                         uint32_t off = ntohl(*off_32++);
 178                         if (!(off & 0x80000000)) {
 179                                 rix->revindex[i].offset = off;
 180                         } else {
 181                                 rix->revindex[i].offset =
 182                                         ((uint64_t)ntohl(*off_64++)) << 32;
 183                                 rix->revindex[i].offset |=
 184                                         ntohl(*off_64++);
 185                         }
 186                         rix->revindex[i].nr = i;
 187                 }
 188         } else {
 189                 for (i = 0; i < num_ent; i++) {
 190                         uint32_t hl = *((uint32_t *)(index + 24 * i));
 191                         rix->revindex[i].offset = ntohl(hl);
 192                         rix->revindex[i].nr = i;
 193                 }
 194         }
 195
 196         /* This knows the pack format -- the 20-byte trailer
 197          * follows immediately after the last object data.
 198          */
 199         rix->revindex[num_ent].offset = p->pack_size - 20;
 200         rix->revindex[num_ent].nr = -1;
 201         sort_revindex(rix->revindex, num_ent, p->pack_size);
 202 }
 203
 204 struct revindex_entry *find_pack_revindex(struct packed_git *p, off_t ofs)
 205 {
 206         int num;
 207         unsigned lo, hi;
 208         struct pack_revindex *rix;
 209         struct revindex_entry *revindex;
 210
 211         if (!pack_revindex_hashsz)
 212                 init_pack_revindex();
 213         num = pack_revindex_ix(p);
 214         if (num < 0)
 215                 die("internal error: pack revindex fubar");
 216
 217         rix = &pack_revindex[num];
 218         if (!rix->revindex)
 219                 create_pack_revindex(rix);
 220         revindex = rix->revindex;
 221
 222         lo = 0;
 223         hi = p->num_objects + 1;
 224         do {
 225                 unsigned mi = lo + (hi - lo) / 2;
 226                 if (revindex[mi].offset == ofs) {
 227                         return revindex + mi;
 228                 } else if (ofs < revindex[mi].offset)
 229                         hi = mi;
 230                 else
 231                         lo = mi + 1;
 232         } while (lo < hi);
 233         error("bad offset for revindex");
 234         return NULL;
 235 }
 236
 237 void discard_revindex(void)
 238 {
 239         if (pack_revindex_hashsz) {
 240                 int i;
 241                 for (i = 0; i < pack_revindex_hashsz; i++)
 242                         free(pack_revindex[i].revindex);
 243                 free(pack_revindex);
 244                 pack_revindex_hashsz = 0;
 245         }
 246 }