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 static int cmp_offset(const void *a_, const void *b_)
  63 {
  64         const struct revindex_entry *a = a_;
  65         const struct revindex_entry *b = b_;
  66         return (a->offset < b->offset) ? -1 : (a->offset > b->offset) ? 1 : 0;
  67 }
  68
  69 /*
  70  * Ordered list of offsets of objects in the pack.
  71  */
  72 static void create_pack_revindex(struct pack_revindex *rix)
  73 {
  74         struct packed_git *p = rix->p;
  75         int num_ent = p->num_objects;
  76         int i;
  77         const char *index = p->index_data;
  78
  79         rix->revindex = xmalloc(sizeof(*rix->revindex) * (num_ent + 1));
  80         index += 4 * 256;
  81
  82         if (p->index_version > 1) {
  83                 const uint32_t *off_32 =
  84                         (uint32_t *)(index + 8 + p->num_objects * (20 + 4));
  85                 const uint32_t *off_64 = off_32 + p->num_objects;
  86                 for (i = 0; i < num_ent; i++) {
  87                         uint32_t off = ntohl(*off_32++);
  88                         if (!(off & 0x80000000)) {
  89                                 rix->revindex[i].offset = off;
  90                         } else {
  91                                 rix->revindex[i].offset =
  92                                         ((uint64_t)ntohl(*off_64++)) << 32;
  93                                 rix->revindex[i].offset |=
  94                                         ntohl(*off_64++);
  95                         }
  96                         rix->revindex[i].nr = i;
  97                 }
  98         } else {
  99                 for (i = 0; i < num_ent; i++) {
 100                         uint32_t hl = *((uint32_t *)(index + 24 * i));
 101                         rix->revindex[i].offset = ntohl(hl);
 102                         rix->revindex[i].nr = i;
 103                 }
 104         }
 105
 106         /* This knows the pack format -- the 20-byte trailer
 107          * follows immediately after the last object data.
 108          */
 109         rix->revindex[num_ent].offset = p->pack_size - 20;
 110         rix->revindex[num_ent].nr = -1;
 111         qsort(rix->revindex, num_ent, sizeof(*rix->revindex), cmp_offset);
 112 }
 113
 114 struct revindex_entry *find_pack_revindex(struct packed_git *p, off_t ofs)
 115 {
 116         int num;
 117         int lo, hi;
 118         struct pack_revindex *rix;
 119         struct revindex_entry *revindex;
 120
 121         if (!pack_revindex_hashsz)
 122                 init_pack_revindex();
 123         num = pack_revindex_ix(p);
 124         if (num < 0)
 125                 die("internal error: pack revindex fubar");
 126
 127         rix = &pack_revindex[num];
 128         if (!rix->revindex)
 129                 create_pack_revindex(rix);
 130         revindex = rix->revindex;
 131
 132         lo = 0;
 133         hi = p->num_objects + 1;
 134         do {
 135                 int mi = (lo + hi) / 2;
 136                 if (revindex[mi].offset == ofs) {
 137                         return revindex + mi;
 138                 } else if (ofs < revindex[mi].offset)
 139                         hi = mi;
 140                 else
 141                         lo = mi + 1;
 142         } while (lo < hi);
 143         die("internal error: pack revindex corrupt");
 144 }