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