pack-revindex.c

   1 #include "cache.h"
   2 #include "pack-revindex.h"
   3 #include "object-store.h"
   4 #include "packfile.h"
   5
   6 struct revindex_entry {
   7         off_t offset;
   8         unsigned int nr;
   9 };
  10
  11 /*
  12  * Pack index for existing packs give us easy access to the offsets into
  13  * corresponding pack file where each object's data starts, but the entries
  14  * do not store the size of the compressed representation (uncompressed
  15  * size is easily available by examining the pack entry header).  It is
  16  * also rather expensive to find the sha1 for an object given its offset.
  17  *
  18  * The pack index file is sorted by object name mapping to offset;
  19  * this revindex array is a list of offset/index_nr pairs
  20  * ordered by offset, so if you know the offset of an object, next offset
  21  * is where its packed representation ends and the index_nr can be used to
  22  * get the object sha1 from the main index.
  23  */
  24
  25 /*
  26  * This is a least-significant-digit radix sort.
  27  *
  28  * It sorts each of the "n" items in "entries" by its offset field. The "max"
  29  * parameter must be at least as large as the largest offset in the array,
  30  * and lets us quit the sort early.
  31  */
  32 static void sort_revindex(struct revindex_entry *entries, unsigned n, off_t max)
  33 {
  34         /*
  35          * We use a "digit" size of 16 bits. That keeps our memory
  36          * usage reasonable, and we can generally (for a 4G or smaller
  37          * packfile) quit after two rounds of radix-sorting.
  38          */
  39 #define DIGIT_SIZE (16)
  40 #define BUCKETS (1 << DIGIT_SIZE)
  41         /*
  42          * We want to know the bucket that a[i] will go into when we are using
  43          * the digit that is N bits from the (least significant) end.
  44          */
  45 #define BUCKET_FOR(a, i, bits) (((a)[(i)].offset >> (bits)) & (BUCKETS-1))
  46
  47         /*
  48          * We need O(n) temporary storage. Rather than do an extra copy of the
  49          * partial results into "entries", we sort back and forth between the
  50          * real array and temporary storage. In each iteration of the loop, we
  51          * keep track of them with alias pointers, always sorting from "from"
  52          * to "to".
  53          */
  54         struct revindex_entry *tmp, *from, *to;
  55         int bits;
  56         unsigned *pos;
  57
  58         ALLOC_ARRAY(pos, BUCKETS);
  59         ALLOC_ARRAY(tmp, n);
  60         from = entries;
  61         to = tmp;
  62
  63         /*
  64          * If (max >> bits) is zero, then we know that the radix digit we are
  65          * on (and any higher) will be zero for all entries, and our loop will
  66          * be a no-op, as everybody lands in the same zero-th bucket.
  67          */
  68         for (bits = 0; max >> bits; bits += DIGIT_SIZE) {
  69                 unsigned i;
  70
  71                 memset(pos, 0, BUCKETS * sizeof(*pos));
  72
  73                 /*
  74                  * We want pos[i] to store the index of the last element that
  75                  * will go in bucket "i" (actually one past the last element).
  76                  * To do this, we first count the items that will go in each
  77                  * bucket, which gives us a relative offset from the last
  78                  * bucket. We can then cumulatively add the index from the
  79                  * previous bucket to get the true index.
  80                  */
  81                 for (i = 0; i < n; i++)
  82                         pos[BUCKET_FOR(from, i, bits)]++;
  83                 for (i = 1; i < BUCKETS; i++)
  84                         pos[i] += pos[i-1];
  85
  86                 /*
  87                  * Now we can drop the elements into their correct buckets (in
  88                  * our temporary array).  We iterate the pos counter backwards
  89                  * to avoid using an extra index to count up. And since we are
  90                  * going backwards there, we must also go backwards through the
  91                  * array itself, to keep the sort stable.
  92                  *
  93                  * Note that we use an unsigned iterator to make sure we can
  94                  * handle 2^32-1 objects, even on a 32-bit system. But this
  95                  * means we cannot use the more obvious "i >= 0" loop condition
  96                  * for counting backwards, and must instead check for
  97                  * wrap-around with UINT_MAX.
  98                  */
  99                 for (i = n - 1; i != UINT_MAX; i--)
 100                         to[--pos[BUCKET_FOR(from, i, bits)]] = from[i];
 101
 102                 /*
 103                  * Now "to" contains the most sorted list, so we swap "from" and
 104                  * "to" for the next iteration.
 105                  */
 106                 SWAP(from, to);
 107         }
 108
 109         /*
 110          * If we ended with our data in the original array, great. If not,
 111          * we have to move it back from the temporary storage.
 112          */
 113         if (from != entries)
 114                 COPY_ARRAY(entries, tmp, n);
 115         free(tmp);
 116         free(pos);
 117
 118 #undef BUCKET_FOR
 119 #undef BUCKETS
 120 #undef DIGIT_SIZE
 121 }
 122
 123 /*
 124  * Ordered list of offsets of objects in the pack.
 125  */
 126 static void create_pack_revindex(struct packed_git *p)
 127 {
 128         const unsigned num_ent = p->num_objects;
 129         unsigned i;
 130         const char *index = p->index_data;
 131         const unsigned hashsz = the_hash_algo->rawsz;
 132
 133         ALLOC_ARRAY(p->revindex, num_ent + 1);
 134         index += 4 * 256;
 135
 136         if (p->index_version > 1) {
 137                 const uint32_t *off_32 =
 138                         (uint32_t *)(index + 8 + (size_t)p->num_objects * (hashsz + 4));
 139                 const uint32_t *off_64 = off_32 + p->num_objects;
 140                 for (i = 0; i < num_ent; i++) {
 141                         const uint32_t off = ntohl(*off_32++);
 142                         if (!(off & 0x80000000)) {
 143                                 p->revindex[i].offset = off;
 144                         } else {
 145                                 p->revindex[i].offset = get_be64(off_64);
 146                                 off_64 += 2;
 147                         }
 148                         p->revindex[i].nr = i;
 149                 }
 150         } else {
 151                 for (i = 0; i < num_ent; i++) {
 152                         const uint32_t hl = *((uint32_t *)(index + (hashsz + 4) * i));
 153                         p->revindex[i].offset = ntohl(hl);
 154                         p->revindex[i].nr = i;
 155                 }
 156         }
 157
 158         /*
 159          * This knows the pack format -- the hash trailer
 160          * follows immediately after the last object data.
 161          */
 162         p->revindex[num_ent].offset = p->pack_size - hashsz;
 163         p->revindex[num_ent].nr = -1;
 164         sort_revindex(p->revindex, num_ent, p->pack_size);
 165 }
 166
 167 int load_pack_revindex(struct packed_git *p)
 168 {
 169         if (!p->revindex) {
 170                 if (open_pack_index(p))
 171                         return -1;
 172                 create_pack_revindex(p);
 173         }
 174         return 0;
 175 }
 176
 177 int offset_to_pack_pos(struct packed_git *p, off_t ofs, uint32_t *pos)
 178 {
 179         unsigned lo, hi;
 180
 181         if (load_pack_revindex(p) < 0)
 182                 return -1;
 183
 184         lo = 0;
 185         hi = p->num_objects + 1;
 186
 187         do {
 188                 const unsigned mi = lo + (hi - lo) / 2;
 189                 off_t got = pack_pos_to_offset(p, mi);
 190
 191                 if (got == ofs) {
 192                         *pos = mi;
 193                         return 0;
 194                 } else if (ofs < got)
 195                         hi = mi;
 196                 else
 197                         lo = mi + 1;
 198         } while (lo < hi);
 199
 200         error("bad offset for revindex");
 201         return -1;
 202 }
 203
 204 uint32_t pack_pos_to_index(struct packed_git *p, uint32_t pos)
 205 {
 206         if (!p->revindex)
 207                 BUG("pack_pos_to_index: reverse index not yet loaded");
 208         if (p->num_objects <= pos)
 209                 BUG("pack_pos_to_index: out-of-bounds object at %"PRIu32, pos);
 210         return p->revindex[pos].nr;
 211 }
 212
 213 off_t pack_pos_to_offset(struct packed_git *p, uint32_t pos)
 214 {
 215         if (!p->revindex)
 216                 BUG("pack_pos_to_index: reverse index not yet loaded");
 217         if (p->num_objects < pos)
 218                 BUG("pack_pos_to_offset: out-of-bounds object at %"PRIu32, pos);
 219         return p->revindex[pos].offset;
 220 }