t/t5310-pack-bitmaps.sh

   1 #!/bin/sh
   2
   3 test_description='exercise basic bitmap functionality'
   4 . ./test-lib.sh
   5
   6 objpath () {
   7         echo ".git/objects/$(echo "$1" | sed -e 's|\(..\)|\1/|')"
   8 }
   9
  10 # show objects present in pack ($1 should be associated *.idx)
  11 list_packed_objects () {
  12         git show-index <"$1" >object-list &&
  13         cut -d' ' -f2 object-list
  14 }
  15
  16 # has_any pattern-file content-file
  17 # tests whether content-file has any entry from pattern-file with entries being
  18 # whole lines.
  19 has_any () {
  20         grep -Ff "$1" "$2"
  21 }
  22
  23 # To ensure the logic for "maximal commits" is exercised, make
  24 # the repository a bit more complicated.
  25 #
  26 #    other                         second
  27 #      *                             *
  28 # (99 commits)                  (99 commits)
  29 #      *                             *
  30 #      |\                           /|
  31 #      | * octo-other  octo-second * |
  32 #      |/|\_________  ____________/|\|
  33 #      | \          \/  __________/  |
  34 #      |  | ________/\ /             |
  35 #      *  |/          * merge-right  *
  36 #      | _|__________/ \____________ |
  37 #      |/ |                         \|
  38 # (l1) *  * merge-left               * (r1)
  39 #      | / \________________________ |
  40 #      |/                           \|
  41 # (l2) *                             * (r2)
  42 #       \___________________________ |
  43 #                                   \|
  44 #                                    * (base)
  45 #
  46 # We only push bits down the first-parent history, which
  47 # makes some of these commits unimportant!
  48 #
  49 # The important part for the maximal commit algorithm is how
  50 # the bitmasks are extended. Assuming starting bit positions
  51 # for second (bit 0) and other (bit 1), the bitmasks at the
  52 # end should be:
  53 #
  54 #      second: 1       (maximal, selected)
  55 #       other: 01      (maximal, selected)
  56 #      (base): 11 (maximal)
  57 #
  58 # This complicated history was important for a previous
  59 # version of the walk that guarantees never walking a
  60 # commit multiple times. That goal might be important
  61 # again, so preserve this complicated case. For now, this
  62 # test will guarantee that the bitmaps are computed
  63 # correctly, even with the repeat calculations.
  64
  65 test_expect_success 'setup repo with moderate-sized history' '
  66         test_commit_bulk --id=file 10 &&
  67         git branch -M second &&
  68         git checkout -b other HEAD~5 &&
  69         test_commit_bulk --id=side 10 &&
  70
  71         # add complicated history setup, including merges and
  72         # ambiguous merge-bases
  73
  74         git checkout -b merge-left other~2 &&
  75         git merge second~2 -m "merge-left" &&
  76
  77         git checkout -b merge-right second~1 &&
  78         git merge other~1 -m "merge-right" &&
  79
  80         git checkout -b octo-second second &&
  81         git merge merge-left merge-right -m "octopus-second" &&
  82
  83         git checkout -b octo-other other &&
  84         git merge merge-left merge-right -m "octopus-other" &&
  85
  86         git checkout other &&
  87         git merge octo-other -m "pull octopus" &&
  88
  89         git checkout second &&
  90         git merge octo-second -m "pull octopus" &&
  91
  92         # Remove these branches so they are not selected
  93         # as bitmap tips
  94         git branch -D merge-left &&
  95         git branch -D merge-right &&
  96         git branch -D octo-other &&
  97         git branch -D octo-second &&
  98
  99         # add padding to make these merges less interesting
 100         # and avoid having them selected for bitmaps
 101         test_commit_bulk --id=file 100 &&
 102         git checkout other &&
 103         test_commit_bulk --id=side 100 &&
 104         git checkout second &&
 105
 106         bitmaptip=$(git rev-parse second) &&
 107         blob=$(echo tagged-blob | git hash-object -w --stdin) &&
 108         git tag tagged-blob $blob &&
 109         git config repack.writebitmaps true
 110 '
 111
 112 test_expect_success 'full repack creates bitmaps' '
 113         GIT_TRACE2_EVENT_NESTING=4 GIT_TRACE2_EVENT="$(pwd)/trace" \
 114                 git repack -ad &&
 115         ls .git/objects/pack/ | grep bitmap >output &&
 116         test_line_count = 1 output &&
 117         grep "\"key\":\"num_selected_commits\",\"value\":\"106\"" trace &&
 118         grep "\"key\":\"num_maximal_commits\",\"value\":\"107\"" trace
 119 '
 120
 121 test_expect_success 'rev-list --test-bitmap verifies bitmaps' '
 122         git rev-list --test-bitmap HEAD
 123 '
 124
 125 rev_list_tests_head () {
 126         test_expect_success "counting commits via bitmap ($state, $branch)" '
 127                 git rev-list --count $branch >expect &&
 128                 git rev-list --use-bitmap-index --count $branch >actual &&
 129                 test_cmp expect actual
 130         '
 131
 132         test_expect_success "counting partial commits via bitmap ($state, $branch)" '
 133                 git rev-list --count $branch~5..$branch >expect &&
 134                 git rev-list --use-bitmap-index --count $branch~5..$branch >actual &&
 135                 test_cmp expect actual
 136         '
 137
 138         test_expect_success "counting commits with limit ($state, $branch)" '
 139                 git rev-list --count -n 1 $branch >expect &&
 140                 git rev-list --use-bitmap-index --count -n 1 $branch >actual &&
 141                 test_cmp expect actual
 142         '
 143
 144         test_expect_success "counting non-linear history ($state, $branch)" '
 145                 git rev-list --count other...second >expect &&
 146                 git rev-list --use-bitmap-index --count other...second >actual &&
 147                 test_cmp expect actual
 148         '
 149
 150         test_expect_success "counting commits with limiting ($state, $branch)" '
 151                 git rev-list --count $branch -- 1.t >expect &&
 152                 git rev-list --use-bitmap-index --count $branch -- 1.t >actual &&
 153                 test_cmp expect actual
 154         '
 155
 156         test_expect_success "counting objects via bitmap ($state, $branch)" '
 157                 git rev-list --count --objects $branch >expect &&
 158                 git rev-list --use-bitmap-index --count --objects $branch >actual &&
 159                 test_cmp expect actual
 160         '
 161
 162         test_expect_success "enumerate commits ($state, $branch)" '
 163                 git rev-list --use-bitmap-index $branch >actual &&
 164                 git rev-list $branch >expect &&
 165                 test_bitmap_traversal --no-confirm-bitmaps expect actual
 166         '
 167
 168         test_expect_success "enumerate --objects ($state, $branch)" '
 169                 git rev-list --objects --use-bitmap-index $branch >actual &&
 170                 git rev-list --objects $branch >expect &&
 171                 test_bitmap_traversal expect actual
 172         '
 173
 174         test_expect_success "bitmap --objects handles non-commit objects ($state, $branch)" '
 175                 git rev-list --objects --use-bitmap-index $branch tagged-blob >actual &&
 176                 grep $blob actual
 177         '
 178 }
 179
 180 rev_list_tests () {
 181         state=$1
 182
 183         for branch in "second" "other"
 184         do
 185                 rev_list_tests_head
 186         done
 187 }
 188
 189 rev_list_tests 'full bitmap'
 190
 191 test_expect_success 'clone from bitmapped repository' '
 192         git clone --no-local --bare . clone.git &&
 193         git rev-parse HEAD >expect &&
 194         git --git-dir=clone.git rev-parse HEAD >actual &&
 195         test_cmp expect actual
 196 '
 197
 198 test_expect_success 'partial clone from bitmapped repository' '
 199         test_config uploadpack.allowfilter true &&
 200         git clone --no-local --bare --filter=blob:none . partial-clone.git &&
 201         (
 202                 cd partial-clone.git &&
 203                 pack=$(echo objects/pack/*.pack) &&
 204                 git verify-pack -v "$pack" >have &&
 205                 awk "/blob/ { print \$1 }" <have >blobs &&
 206                 # we expect this single blob because of the direct ref
 207                 git rev-parse refs/tags/tagged-blob >expect &&
 208                 test_cmp expect blobs
 209         )
 210 '
 211
 212 test_expect_success 'setup further non-bitmapped commits' '
 213         test_commit_bulk --id=further 10
 214 '
 215
 216 rev_list_tests 'partial bitmap'
 217
 218 test_expect_success 'fetch (partial bitmap)' '
 219         git --git-dir=clone.git fetch origin second:second &&
 220         git rev-parse HEAD >expect &&
 221         git --git-dir=clone.git rev-parse HEAD >actual &&
 222         test_cmp expect actual
 223 '
 224
 225 test_expect_success 'incremental repack fails when bitmaps are requested' '
 226         test_commit more-1 &&
 227         test_must_fail git repack -d 2>err &&
 228         test_i18ngrep "Incremental repacks are incompatible with bitmap" err
 229 '
 230
 231 test_expect_success 'incremental repack can disable bitmaps' '
 232         test_commit more-2 &&
 233         git repack -d --no-write-bitmap-index
 234 '
 235
 236 test_expect_success 'pack-objects respects --local (non-local loose)' '
 237         git init --bare alt.git &&
 238         echo $(pwd)/alt.git/objects >.git/objects/info/alternates &&
 239         echo content1 >file1 &&
 240         # non-local loose object which is not present in bitmapped pack
 241         altblob=$(GIT_DIR=alt.git git hash-object -w file1) &&
 242         # non-local loose object which is also present in bitmapped pack
 243         git cat-file blob $blob | GIT_DIR=alt.git git hash-object -w --stdin &&
 244         git add file1 &&
 245         test_tick &&
 246         git commit -m commit_file1 &&
 247         echo HEAD | git pack-objects --local --stdout --revs >1.pack &&
 248         git index-pack 1.pack &&
 249         list_packed_objects 1.idx >1.objects &&
 250         printf "%s\n" "$altblob" "$blob" >nonlocal-loose &&
 251         ! has_any nonlocal-loose 1.objects
 252 '
 253
 254 test_expect_success 'pack-objects respects --honor-pack-keep (local non-bitmapped pack)' '
 255         echo content2 >file2 &&
 256         blob2=$(git hash-object -w file2) &&
 257         git add file2 &&
 258         test_tick &&
 259         git commit -m commit_file2 &&
 260         printf "%s\n" "$blob2" "$bitmaptip" >keepobjects &&
 261         pack2=$(git pack-objects pack2 <keepobjects) &&
 262         mv pack2-$pack2.* .git/objects/pack/ &&
 263         >.git/objects/pack/pack2-$pack2.keep &&
 264         rm $(objpath $blob2) &&
 265         echo HEAD | git pack-objects --honor-pack-keep --stdout --revs >2a.pack &&
 266         git index-pack 2a.pack &&
 267         list_packed_objects 2a.idx >2a.objects &&
 268         ! has_any keepobjects 2a.objects
 269 '
 270
 271 test_expect_success 'pack-objects respects --local (non-local pack)' '
 272         mv .git/objects/pack/pack2-$pack2.* alt.git/objects/pack/ &&
 273         echo HEAD | git pack-objects --local --stdout --revs >2b.pack &&
 274         git index-pack 2b.pack &&
 275         list_packed_objects 2b.idx >2b.objects &&
 276         ! has_any keepobjects 2b.objects
 277 '
 278
 279 test_expect_success 'pack-objects respects --honor-pack-keep (local bitmapped pack)' '
 280         ls .git/objects/pack/ | grep bitmap >output &&
 281         test_line_count = 1 output &&
 282         packbitmap=$(basename $(cat output) .bitmap) &&
 283         list_packed_objects .git/objects/pack/$packbitmap.idx >packbitmap.objects &&
 284         test_when_finished "rm -f .git/objects/pack/$packbitmap.keep" &&
 285         >.git/objects/pack/$packbitmap.keep &&
 286         echo HEAD | git pack-objects --honor-pack-keep --stdout --revs >3a.pack &&
 287         git index-pack 3a.pack &&
 288         list_packed_objects 3a.idx >3a.objects &&
 289         ! has_any packbitmap.objects 3a.objects
 290 '
 291
 292 test_expect_success 'pack-objects respects --local (non-local bitmapped pack)' '
 293         mv .git/objects/pack/$packbitmap.* alt.git/objects/pack/ &&
 294         rm -f .git/objects/pack/multi-pack-index &&
 295         test_when_finished "mv alt.git/objects/pack/$packbitmap.* .git/objects/pack/" &&
 296         echo HEAD | git pack-objects --local --stdout --revs >3b.pack &&
 297         git index-pack 3b.pack &&
 298         list_packed_objects 3b.idx >3b.objects &&
 299         ! has_any packbitmap.objects 3b.objects
 300 '
 301
 302 test_expect_success 'pack-objects to file can use bitmap' '
 303         # make sure we still have 1 bitmap index from previous tests
 304         ls .git/objects/pack/ | grep bitmap >output &&
 305         test_line_count = 1 output &&
 306         # verify equivalent packs are generated with/without using bitmap index
 307         packasha1=$(git pack-objects --no-use-bitmap-index --all packa </dev/null) &&
 308         packbsha1=$(git pack-objects --use-bitmap-index --all packb </dev/null) &&
 309         list_packed_objects packa-$packasha1.idx >packa.objects &&
 310         list_packed_objects packb-$packbsha1.idx >packb.objects &&
 311         test_cmp packa.objects packb.objects
 312 '
 313
 314 test_expect_success 'full repack, reusing previous bitmaps' '
 315         git repack -ad &&
 316         ls .git/objects/pack/ | grep bitmap >output &&
 317         test_line_count = 1 output
 318 '
 319
 320 test_expect_success 'fetch (full bitmap)' '
 321         git --git-dir=clone.git fetch origin second:second &&
 322         git rev-parse HEAD >expect &&
 323         git --git-dir=clone.git rev-parse HEAD >actual &&
 324         test_cmp expect actual
 325 '
 326
 327 test_expect_success 'create objects for missing-HAVE tests' '
 328         blob=$(echo "missing have" | git hash-object -w --stdin) &&
 329         tree=$(printf "100644 blob $blob\tfile\n" | git mktree) &&
 330         parent=$(echo parent | git commit-tree $tree) &&
 331         commit=$(echo commit | git commit-tree $tree -p $parent) &&
 332         cat >revs <<-EOF
 333         HEAD
 334         ^HEAD^
 335         ^$commit
 336         EOF
 337 '
 338
 339 test_expect_success 'pack-objects respects --incremental' '
 340         cat >revs2 <<-EOF &&
 341         HEAD
 342         $commit
 343         EOF
 344         git pack-objects --incremental --stdout --revs <revs2 >4.pack &&
 345         git index-pack 4.pack &&
 346         list_packed_objects 4.idx >4.objects &&
 347         test_line_count = 4 4.objects &&
 348         git rev-list --objects $commit >revlist &&
 349         cut -d" " -f1 revlist |sort >objects &&
 350         test_cmp 4.objects objects
 351 '
 352
 353 test_expect_success 'pack with missing blob' '
 354         rm $(objpath $blob) &&
 355         git pack-objects --stdout --revs <revs >/dev/null
 356 '
 357
 358 test_expect_success 'pack with missing tree' '
 359         rm $(objpath $tree) &&
 360         git pack-objects --stdout --revs <revs >/dev/null
 361 '
 362
 363 test_expect_success 'pack with missing parent' '
 364         rm $(objpath $parent) &&
 365         git pack-objects --stdout --revs <revs >/dev/null
 366 '
 367
 368 test_expect_success JGIT,SHA1 'we can read jgit bitmaps' '
 369         git clone --bare . compat-jgit.git &&
 370         (
 371                 cd compat-jgit.git &&
 372                 rm -f objects/pack/*.bitmap &&
 373                 jgit gc &&
 374                 git rev-list --test-bitmap HEAD
 375         )
 376 '
 377
 378 test_expect_success JGIT,SHA1 'jgit can read our bitmaps' '
 379         git clone --bare . compat-us.git &&
 380         (
 381                 cd compat-us.git &&
 382                 git repack -adb &&
 383                 # jgit gc will barf if it does not like our bitmaps
 384                 jgit gc
 385         )
 386 '
 387
 388 test_expect_success 'splitting packs does not generate bogus bitmaps' '
 389         test-tool genrandom foo $((1024 * 1024)) >rand &&
 390         git add rand &&
 391         git commit -m "commit with big file" &&
 392         git -c pack.packSizeLimit=500k repack -adb &&
 393         git init --bare no-bitmaps.git &&
 394         git -C no-bitmaps.git fetch .. HEAD
 395 '
 396
 397 test_expect_success 'set up reusable pack' '
 398         rm -f .git/objects/pack/*.keep &&
 399         git repack -adb &&
 400         reusable_pack () {
 401                 git for-each-ref --format="%(objectname)" |
 402                 git pack-objects --delta-base-offset --revs --stdout "$@"
 403         }
 404 '
 405
 406 test_expect_success 'pack reuse respects --honor-pack-keep' '
 407         test_when_finished "rm -f .git/objects/pack/*.keep" &&
 408         for i in .git/objects/pack/*.pack
 409         do
 410                 >${i%.pack}.keep
 411         done &&
 412         reusable_pack --honor-pack-keep >empty.pack &&
 413         git index-pack empty.pack &&
 414         git show-index <empty.idx >actual &&
 415         test_must_be_empty actual
 416 '
 417
 418 test_expect_success 'pack reuse respects --local' '
 419         mv .git/objects/pack/* alt.git/objects/pack/ &&
 420         test_when_finished "mv alt.git/objects/pack/* .git/objects/pack/" &&
 421         reusable_pack --local >empty.pack &&
 422         git index-pack empty.pack &&
 423         git show-index <empty.idx >actual &&
 424         test_must_be_empty actual
 425 '
 426
 427 test_expect_success 'pack reuse respects --incremental' '
 428         reusable_pack --incremental >empty.pack &&
 429         git index-pack empty.pack &&
 430         git show-index <empty.idx >actual &&
 431         test_must_be_empty actual
 432 '
 433
 434 test_expect_success 'truncated bitmap fails gracefully (ewah)' '
 435         test_config pack.writebitmaphashcache false &&
 436         git repack -ad &&
 437         git rev-list --use-bitmap-index --count --all >expect &&
 438         bitmap=$(ls .git/objects/pack/*.bitmap) &&
 439         test_when_finished "rm -f $bitmap" &&
 440         test_copy_bytes 256 <$bitmap >$bitmap.tmp &&
 441         mv -f $bitmap.tmp $bitmap &&
 442         git rev-list --use-bitmap-index --count --all >actual 2>stderr &&
 443         test_cmp expect actual &&
 444         test_i18ngrep corrupt.ewah.bitmap stderr
 445 '
 446
 447 test_expect_success 'truncated bitmap fails gracefully (cache)' '
 448         git repack -ad &&
 449         git rev-list --use-bitmap-index --count --all >expect &&
 450         bitmap=$(ls .git/objects/pack/*.bitmap) &&
 451         test_when_finished "rm -f $bitmap" &&
 452         test_copy_bytes 512 <$bitmap >$bitmap.tmp &&
 453         mv -f $bitmap.tmp $bitmap &&
 454         git rev-list --use-bitmap-index --count --all >actual 2>stderr &&
 455         test_cmp expect actual &&
 456         test_i18ngrep corrupted.bitmap.index stderr
 457 '
 458
 459 # have_delta <obj> <expected_base>
 460 #
 461 # Note that because this relies on cat-file, it might find _any_ copy of an
 462 # object in the repository. The caller is responsible for making sure
 463 # there's only one (e.g., via "repack -ad", or having just fetched a copy).
 464 have_delta () {
 465         echo $2 >expect &&
 466         echo $1 | git cat-file --batch-check="%(deltabase)" >actual &&
 467         test_cmp expect actual
 468 }
 469
 470 # Create a state of history with these properties:
 471 #
 472 #  - refs that allow a client to fetch some new history, while sharing some old
 473 #    history with the server; we use branches delta-reuse-old and
 474 #    delta-reuse-new here
 475 #
 476 #  - the new history contains an object that is stored on the server as a delta
 477 #    against a base that is in the old history
 478 #
 479 #  - the base object is not immediately reachable from the tip of the old
 480 #    history; finding it would involve digging down through history we know the
 481 #    other side has
 482 #
 483 # This should result in a state where fetching from old->new would not
 484 # traditionally reuse the on-disk delta (because we'd have to dig to realize
 485 # that the client has it), but we will do so if bitmaps can tell us cheaply
 486 # that the other side has it.
 487 test_expect_success 'set up thin delta-reuse parent' '
 488         # This first commit contains the buried base object.
 489         test-tool genrandom delta 16384 >file &&
 490         git add file &&
 491         git commit -m "delta base" &&
 492         base=$(git rev-parse --verify HEAD:file) &&
 493
 494         # These intermediate commits bury the base back in history.
 495         # This becomes the "old" state.
 496         for i in 1 2 3 4 5
 497         do
 498                 echo $i >file &&
 499                 git commit -am "intermediate $i" || return 1
 500         done &&
 501         git branch delta-reuse-old &&
 502
 503         # And now our new history has a delta against the buried base. Note
 504         # that this must be smaller than the original file, since pack-objects
 505         # prefers to create deltas from smaller objects to larger.
 506         test-tool genrandom delta 16300 >file &&
 507         git commit -am "delta result" &&
 508         delta=$(git rev-parse --verify HEAD:file) &&
 509         git branch delta-reuse-new &&
 510
 511         # Repack with bitmaps and double check that we have the expected delta
 512         # relationship.
 513         git repack -adb &&
 514         have_delta $delta $base
 515 '
 516
 517 # Now we can sanity-check the non-bitmap behavior (that the server is not able
 518 # to reuse the delta). This isn't strictly something we care about, so this
 519 # test could be scrapped in the future. But it makes sure that the next test is
 520 # actually triggering the feature we want.
 521 #
 522 # Note that our tools for working with on-the-wire "thin" packs are limited. So
 523 # we actually perform the fetch, retain the resulting pack, and inspect the
 524 # result.
 525 test_expect_success 'fetch without bitmaps ignores delta against old base' '
 526         test_config pack.usebitmaps false &&
 527         test_when_finished "rm -rf client.git" &&
 528         git init --bare client.git &&
 529         (
 530                 cd client.git &&
 531                 git config transfer.unpackLimit 1 &&
 532                 git fetch .. delta-reuse-old:delta-reuse-old &&
 533                 git fetch .. delta-reuse-new:delta-reuse-new &&
 534                 have_delta $delta $ZERO_OID
 535         )
 536 '
 537
 538 # And do the same for the bitmap case, where we do expect to find the delta.
 539 test_expect_success 'fetch with bitmaps can reuse old base' '
 540         test_config pack.usebitmaps true &&
 541         test_when_finished "rm -rf client.git" &&
 542         git init --bare client.git &&
 543         (
 544                 cd client.git &&
 545                 git config transfer.unpackLimit 1 &&
 546                 git fetch .. delta-reuse-old:delta-reuse-old &&
 547                 git fetch .. delta-reuse-new:delta-reuse-new &&
 548                 have_delta $delta $base
 549         )
 550 '
 551
 552 test_done