| 1 | #!/bin/sh |
| 2 | |
| 3 | test_description='handling of duplicate objects in incoming packfiles' |
| 4 | |
| 5 | . ./test-lib.sh |
| 6 | . "$TEST_DIRECTORY"/lib-pack.sh |
| 7 | |
| 8 | test_expect_success 'setup' ' |
| 9 | test_oid_cache <<-EOF |
| 10 | lo_oid sha1:e68fe8129b546b101aee9510c5328e7f21ca1d18 |
| 11 | lo_oid sha256:471819e8c52bf11513f100b2810a8aa0622d5cd3d1c913758a071dd4b3bad8fe |
| 12 | |
| 13 | missing_oid sha1:e69d000000000000000000000000000000000000 |
| 14 | missing_oid sha256:4720000000000000000000000000000000000000000000000000000000000000 |
| 15 | EOF |
| 16 | ' |
| 17 | |
| 18 | # The sha1s we have in our pack. It's important that these have the same |
| 19 | # starting byte, so that they end up in the same fanout section of the index. |
| 20 | # That lets us make sure we are exercising the binary search with both sets. |
| 21 | LO_SHA1=$(test_oid lo_oid) |
| 22 | HI_SHA1=$EMPTY_BLOB |
| 23 | |
| 24 | # And here's a "missing sha1" which will produce failed lookups. It must also |
| 25 | # be in the same fanout section, and should be between the two (so that during |
| 26 | # our binary search, we are sure to end up looking at one or the other of the |
| 27 | # duplicate runs). |
| 28 | MISSING_SHA1=$(test_oid missing_oid) |
| 29 | |
| 30 | # Three distinct objects for tests where physical pack order matters. |
| 31 | A=$(test_oid packlib_7_0) |
| 32 | B=$LO_SHA1 |
| 33 | C=$HI_SHA1 |
| 34 | |
| 35 | # git will never intentionally create packfiles with |
| 36 | # duplicate objects, so we have to construct them by hand. |
| 37 | # |
| 38 | # $1 is the name of the packfile to create |
| 39 | # |
| 40 | # $2 is the number of times to duplicate each object |
| 41 | create_pack () { |
| 42 | pack_header "$((2 * $2))" >"$1" && |
| 43 | for i in $(test_seq 1 "$2"); do |
| 44 | pack_obj $LO_SHA1 && |
| 45 | pack_obj $HI_SHA1 |
| 46 | done >>"$1" && |
| 47 | pack_trailer "$1" |
| 48 | } |
| 49 | |
| 50 | # double-check that create_pack actually works |
| 51 | test_expect_success 'pack with no duplicates' ' |
| 52 | create_pack no-dups.pack 1 && |
| 53 | git index-pack --stdin <no-dups.pack |
| 54 | ' |
| 55 | |
| 56 | test_expect_success 'index-pack will allow duplicate objects by default' ' |
| 57 | clear_packs && |
| 58 | create_pack dups.pack 100 && |
| 59 | git index-pack --stdin <dups.pack |
| 60 | ' |
| 61 | |
| 62 | test_expect_success 'bitmap writer rejects duplicate objects' ' |
| 63 | pack=$(ls .git/objects/pack/pack-*.pack) && |
| 64 | test_must_fail test-tool bitmap write "$(basename "$pack")" \ |
| 65 | </dev/null 2>err && |
| 66 | test_grep "fatal: pack contains duplicate object" err |
| 67 | ' |
| 68 | |
| 69 | test_expect_success 'create batch-check test vectors' ' |
| 70 | cat >input <<-EOF && |
| 71 | $LO_SHA1 |
| 72 | $HI_SHA1 |
| 73 | $MISSING_SHA1 |
| 74 | EOF |
| 75 | cat >expect <<-EOF |
| 76 | $LO_SHA1 blob 2 |
| 77 | $HI_SHA1 blob 0 |
| 78 | $MISSING_SHA1 missing |
| 79 | EOF |
| 80 | ' |
| 81 | |
| 82 | test_expect_success 'lookup in duplicated pack' ' |
| 83 | git cat-file --batch-check <input >actual && |
| 84 | test_cmp expect actual |
| 85 | ' |
| 86 | |
| 87 | test_expect_success 'verify MIDX containing duplicated pack objects' ' |
| 88 | git multi-pack-index write && |
| 89 | test-tool read-midx --show-objects .git/objects >midx-objects && |
| 90 | midx_offset=$( |
| 91 | awk -v oid="$LO_SHA1" "\$1 == oid { print \$2 }" <midx-objects |
| 92 | ) && |
| 93 | lookup_offset=$( |
| 94 | test-tool find-pack --check-count=1 --show-offset "$LO_SHA1" |
| 95 | ) && |
| 96 | test "$midx_offset" -ne "$lookup_offset" && |
| 97 | git multi-pack-index verify && |
| 98 | git fsck --full |
| 99 | ' |
| 100 | |
| 101 | test_expect_success 'duplicate entries remain in pack reverse index' ' |
| 102 | clear_packs && |
| 103 | { |
| 104 | pack_header 4 && |
| 105 | pack_obj $A && |
| 106 | pack_obj $B && |
| 107 | pack_obj $A && |
| 108 | pack_obj $C |
| 109 | } >physical-order.pack && |
| 110 | pack_trailer physical-order.pack && |
| 111 | |
| 112 | test_must_fail git index-pack --rev-index --stdin --strict \ |
| 113 | <physical-order.pack 2>err && |
| 114 | test_grep "appears twice in the pack" err && |
| 115 | |
| 116 | git index-pack --rev-index --stdin <physical-order.pack && |
| 117 | git show-index <"$(ls .git/objects/pack/pack-*.idx)" >offsets.raw && |
| 118 | |
| 119 | sort -n offsets.raw | grep -A1 "$B" | cut -d" " -f1 >adjacent && |
| 120 | echo $(($(tail -n1 adjacent) - $(head -n1 adjacent))) >expect && |
| 121 | echo "$B" >in && |
| 122 | |
| 123 | GIT_TEST_REV_INDEX_DIE_IN_MEMORY=1 \ |
| 124 | git cat-file --batch-check="%(objectsize:disk)" \ |
| 125 | <in >actual.disk && |
| 126 | GIT_TEST_REV_INDEX_DIE_ON_DISK=1 \ |
| 127 | git -c pack.readReverseIndex=false \ |
| 128 | cat-file --batch-check="%(objectsize:disk)" \ |
| 129 | <in >actual.mem && |
| 130 | |
| 131 | test_cmp expect actual.disk && |
| 132 | test_cmp expect actual.mem |
| 133 | ' |
| 134 | |
| 135 | test_expect_success 'index-pack can reject packs with duplicates' ' |
| 136 | clear_packs && |
| 137 | create_pack dups.pack 2 && |
| 138 | test_must_fail git index-pack --strict --stdin <dups.pack && |
| 139 | test_expect_code 1 git cat-file -e $LO_SHA1 |
| 140 | ' |
| 141 | |
| 142 | test_done |