jobd/gc.sh

   1 #!/bin/sh
   2
   3 # NOTE: additional options can be passed to git repack by specifying
   4 #       them after the project name, for example:
   5 #         gc.sh my-project -f
   6
   7 . @basedir@/shlib.sh
   8
   9 set -e
  10
  11 if [ $# -lt 1 ]; then
  12         echo "Usage: gc.sh projname [extra-repack-args]" >&2
  13         exit 1
  14 fi
  15
  16 # packing options
  17 packopts="--depth=50 --window=50 --window-memory=${var_window_memory:-1g}"
  18 quiet=; [ -n "$show_progress" ] || quiet=-q
  19
  20 umask 002
  21 [ "$cfg_permission_control" != "Hooks" ] || umask 000
  22 clean_git_env
  23
  24 pidactive() {
  25         if _result="$(kill -0 "$1" 2>&1)"; then
  26                 # process exists and we have permission to signal it
  27                 return 0
  28         fi
  29         case "$_result" in *"not permitted"*)
  30                 # we do not have permission to signal the process
  31                 return 0
  32         esac
  33         # process does not exist
  34         return 1
  35 }
  36
  37 createlock() {
  38         # A .lock file should only exist for much less than a second.
  39         # If we see a stale lock file (> 1h old), remove it and then,
  40         # just in case, wait 30 seconds for any process whose .lock
  41         # we might have just removed (it's racy) to finish doing what
  42         # should take much less than a second to do.
  43         _stalelock="$(find "$1.lock" -maxdepth 1 -mmin +60 -print 2>/dev/null || :)"
  44         if [ -n "$_stalelock" ]; then
  45                 rm -f "$_stalelock"
  46                 sleep 30
  47         fi
  48         for _try in p p n; do
  49                 if (set -C; > "$1.lock") 2>/dev/null; then
  50                         echo "$1.lock"
  51                         return 0
  52                 fi
  53                 # delay and try again
  54                 [ "$_try" != "p" ] || sleep 1
  55         done
  56         # cannot create lock file
  57         return 1
  58 }
  59
  60 # return true if there's more than one objects/pack-<sha>.pack file or
  61 # ANY sha-1 files in objects
  62 is_dirty() {
  63         _packs=$(find objects/pack -type f -name "pack-$octet20.pack" -print | head -n 2 | LC_ALL=C wc -l)
  64         if [ $_packs != 1 ] && [ $_packs != 0 ]; then
  65                 return 0
  66         fi
  67         _objs=$(find objects/$octet -type f -name "$octet19" -print 2>/dev/null | head -n 1 | LC_ALL=C wc -l)
  68         [ $_objs -ne 0 ]
  69 }
  70
  71 # make sure combine-packs uses the correct Git executable
  72 run_combine_packs() {
  73         PATH="$var_git_exec_path:$cfg_basedir/bin:$PATH" @basedir@/jobd/combine-packs.sh "$@"
  74 }
  75
  76 # combine the input pack(s) into a new pack (or possibly packs if packSizeLimit set)
  77 # input pack names are read from standard input one per line delimited by the first
  78 # ':', ' ' or '\n' character on the line (which allows gfi-packs to be read directly)
  79 # all arguments, if any, are passed to pack-objects as additional options
  80 # returns non-zero on failure AND creates .gc_failed in that case
  81 combine_packs() {
  82         rm -f .gc_failed
  83         find objects/pack -maxdepth 1 -type f -name '*.zap*' -print0 | xargs -0 rm -f
  84         run_combine_packs --replace "$@" $packopts --all-progress-implied $quiet --non-empty || {
  85                 >.gc_failed
  86                 return 1
  87         }
  88         return 0
  89 }
  90
  91 # if the current directory is_gfi_mirror then repack all packs listed in gfi-packs
  92 repack_gfi_packs() {
  93         [ -n "$gfi_mirror" ] || return 0
  94         [ -d objects/pack ] || { rm -f gfi-packs; return 0; }
  95         progress "~ [$proj] redeltifying poor quality git fast-import packs"
  96         combine_packs --ignore-missing --no-reuse-delta <gfi-packs
  97         rm -f gfi-packs
  98         return 0
  99 }
 100
 101 # combine small packs into larger pack(s)
 102 # we avoid any keep, bndl or bitmap packs
 103 # if the optional argument is non-empty even a single small pack will be redeltad
 104 combine_small_packs() {
 105         _didprogress=
 106         _minsmallpacks=2
 107         if [ -n "$1" ] && [ -n "$noreusedeltaopt" ]; then
 108                 _minsmallpacks=1
 109         fi
 110         _lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl --quiet"
 111         _lpo="$_lpo --object-limit $var_redelta_threshold objects/pack"
 112         while
 113                 _cnt="$(list_packs --count $_lpo || :)"
 114                 test "${_cnt:-0}" -ge $_minsmallpacks
 115         do
 116                 [ -n "$_didprogress" ] || {
 117                         progress "~ [$proj] combining small packs into a single larger pack"
 118                         _didprogress=1
 119                 }
 120                 _newp="$(list_packs $_lpo | combine_packs --names $noreusedeltaopt)"
 121                 _newc="$(echo $(echo "$_newp" | LC_ALL=C wc -w))"
 122                 # be paranoid and exit the loop if we haven't reduced the number of packs
 123                 [ $_newc -lt $_cnt ] || break
 124                 _minsmallpacks=2
 125         done
 126         return 0
 127 }
 128
 129 # Unfortunately git-svn lacks the ability to store newly fetched revisions as a pack.
 130 # However, the fetch code conveniently sets .svnpack just before it runs git-svn fetch
 131 # so that it's easy to find all the objects that have been fetched by git-svn and
 132 # combine them into a pack.  The --no-reuse-delta option is meaningless here since
 133 # everything to be packed is a loose object and therefore not a delta so deltification
 134 # will always take place.
 135 make_svn_pack() {
 136         [ -f .svnpack ] && [ -n "$svn_mirror" ] || return 0
 137         rm -f .svnpackgc
 138         mv -f .svnpack .svnpackgc
 139         progress "~ [$proj] combining loose git-svn objects into a pack"
 140         _newp="$(find objects/$octet -maxdepth 1 -type f -newer .svnpackgc -name "$octet19" -print 2>/dev/null |
 141         LC_ALL=C awk -F / '{print $2 $3}' |
 142         run_combine_packs --objects --names $packopts --incremental --all-progress-implied $quiet --non-empty)" || {
 143                 mv -f .svnpackgc .svnpack
 144                 >.gc_failed
 145                 return 1
 146         }
 147         if [ -n "$_newp" ]; then
 148                 # remove the now-redundant loose objects -- this is always safe
 149                 # even during a concurrent push because a reprepare_packed_git
 150                 # will be triggered if an object that should be there is not
 151                 # found thereby finding it in the new pack instead
 152                 git prune-packed $quiet
 153         fi
 154         rm -f .svnpackgc
 155 }
 156
 157 # HEADSHA="$(pack_is_complete /full/path/to/some.pack /full/path/to/packed-refs "$(cat HEAD)")"
 158 pack_is_complete() {
 159         # Must have a matching .idx file and a non-empty packed-refs file
 160         [ -s "${1%.pack}.idx" ] || return 1
 161         [ -s "$2" ] || return 1
 162         _headsha=
 163         case "$3" in
 164                 $octet20)
 165                         _headsha="$3"
 166                         ;;
 167                 "ref: refs/"?*|"ref:refs/"?*|"refs/"?*)
 168                         _headmatch="${3#ref:}"
 169                         _headmatch="${_headmatch# }"
 170                         _headmatchpat="$(echo "$_headmatch" | LC_ALL=C sed -e 's/\([.$]\)/\\\1/g')"
 171                         _headsha="$(LC_ALL=C grep -e "^$octet20 $_headmatchpat\$" < "$2" | \
 172                                     LC_ALL=C cut -d ' ' -f 1)"
 173                         case "$_headsha" in $octet20) :;; *)
 174                                 return 1
 175                         esac
 176                         ;;
 177                 *)
 178                         # bad HEAD
 179                         return 1
 180         esac
 181         rm -rf pack_is_complete_test
 182         mkdir pack_is_complete_test
 183         mkdir pack_is_complete_test/refs
 184         mkdir pack_is_complete_test/objects
 185         mkdir pack_is_complete_test/objects/pack
 186         echo "$_headsha" > pack_is_complete_test/HEAD
 187         ln -s "$1" pack_is_complete_test/objects/pack/
 188         ln -s "${1%.pack}.idx" pack_is_complete_test/objects/pack/
 189         ln -s "$2" pack_is_complete_test/packed-refs
 190         _count="$(git --git-dir=pack_is_complete_test rev-list --count --all 2>/dev/null || :)"
 191         rm -rf pack_is_complete_test
 192         [ -n "$_count" ] || return 1
 193         [ "$_count" -gt 0 ] 2>/dev/null || return 1
 194         echo "$_headsha"
 195 }
 196
 197 # On return a "$lockf" will have been created that must be removed when gc is done
 198 lock_gc() {
 199         # be compatibile with gc.pid file from newer Git releases
 200         lockf=gc.pid
 201         hn="$(hostname)"
 202         active=
 203         if [ "$(createlock "$lockf")" ]; then
 204                 # If $lockf is:
 205                 #   1) less than 12 hours old
 206                 #   2) contains two fields (pid hostname) NO trailing NL
 207                 #   3) the hostname is different OR the pid is still alive
 208                 # then we exit as another active process is holding the lock
 209                 if [ "$(find "$lockf" -maxdepth 1 -mmin -720 -print 2>/dev/null)" ]; then
 210                         apid=
 211                         ahost=
 212                         read -r apid ahost ajunk < "$lockf" || :
 213                         if [ "$apid" ] && [ "$ahost" ]; then
 214                                 if [ "$ahost" != "$hn" ] || pidactive "$apid"; then
 215                                         active=1
 216                                 fi
 217                         fi
 218                 fi
 219         else
 220                 echo >&2 "[$proj] unable to create gc.pid.lock file"
 221                 exit 1
 222         fi
 223         if [ -n "$active" ]; then
 224                 rm -f "$lockf.lock"
 225                 echo >&2 "[$proj] gc already running on machine '$ahost' pid '$apid'"
 226                 exit 1
 227         fi
 228         printf "%s %s" "$$" "$hn" > "$lockf.lock"
 229         chmod 0664 "$lockf.lock"
 230         mv -f "$lockf.lock" "$lockf"
 231 }
 232
 233 # Remove any crud that's been left behind by interrupted operations
 234 # that did not clean up after themselves
 235 remove_crud() {
 236         # Remove any existing FETCH_HEAD
 237         # There can only be a FETCH_HEAD if we've been fetching, not if we've been
 238         # receiving pushes (those never create a FETCH_HEAD).
 239         # And if we're fetching because we're a mirror, we know we're not fetching right
 240         # now since jobd.pl never runs a project's fetch simultaneously with its gc.
 241         # Therefore any existing FETCH_HEAD is junk.  And it may be many megabytes if
 242         # there were a lot of refs.
 243         rm -f FETCH_HEAD
 244
 245         # Remove any stale pack remnants that are more than an hour old.
 246         # Stale pack fragments are defined as any pack-<sha1>.ext where .ext is NOT
 247         # .pack AND the corresponding .pack DOES NOT exist.  A bunch of stale
 248         # pack-<sha1>.idx files without their corresponding .pack files are worthless
 249         # and just waste space.  Normally there shouldn't be any remnants but actually
 250         # this can happen when things are interrupted at just the wrong time.
 251         # Note that the objects/pack directory is created by git init and should
 252         # always exist.
 253         find objects/pack -maxdepth 1 -type f -mmin +60 -name "pack-$octet20.?*" -print | \
 254         LC_ALL=C sed -e 's/^objects\/pack\/pack-//; s/\..*$//' | LC_ALL=C sort -u | \
 255         while read packsha; do
 256                 [ ! -e "objects/pack/pack-$packsha.pack" ] || continue
 257                 rm -f "objects/pack/pack-$packsha".?*
 258         done
 259
 260         # Remove any stale pack .keep files that are more than 12 hours old.
 261         # We don't do anything to create any permanent pack .keep files, so they must
 262         # be remnants from some failed push or something.  Removing the .keep will
 263         # allow the pack to be properly repacked.
 264         find objects/pack -maxdepth 1 -type f -mmin +720 -name "pack-$octet20.keep" -print0 | xargs -0 rm -f
 265
 266         # Remove any stale tmp_pack_*, tmp_idx_*, tmp_bitmap_*, packtmp-* or .tmp-*-pack files
 267         # that are more than 12 hours old.
 268         find objects/pack -maxdepth 1 -type f -mmin +720 \( \
 269                 -name "tmp_pack_?*" -o -name "tmp_idx_?*" -o -name "tmp_bitmap_?*" -o \
 270                 -name "packtmp-?*" -o -name ".tmp-?*-pack" \
 271                 \) -print0 | xargs -0 rm -f
 272
 273         # Remove any stale shallow_* files that are more than 12 hours old.
 274         # These can be left behind by Git >= 1.8.4.2 and < 2.0.0 when a client
 275         # requests a shallow clone.
 276         find . -maxdepth 1 -type f -mmin +720 -name "shallow_?*" -print0 | xargs -0 rm -f
 277
 278         # Remove any stale *.temp files in the objects area that are more than 12 hours old.
 279         # This can be stale sha1.temp, or stale *.pack.temp so we kill all stale *.temp.
 280         find objects -type f -mmin +720 -name "*.temp" -print0 | xargs -0 rm -f
 281
 282         # Remove any stale *.lock files in the htmlcache area that might have been left
 283         # behind after an abnormal exit during an attempt to update a cached file and
 284         # are more than 1 hour old.
 285         ! [ -d htmlcache ] || find htmlcache -type f -mmin +60 -name "*.lock" -print0 | xargs -0 rm -f
 286
 287         # Remove any stale git-svn temp files that are more than 12 hours old.
 288         # The git-svn process creates temp files with random 10 character names
 289         # in the root of $GIT_DIR.  Unfortunately they do not have a recognizable
 290         # prefix, so we just have to kill any files with a 10-character name.  We
 291         # do this only for git-svn mirrors.  All characters are chosen from
 292         # [A-Za-z0-9_] so we can at least check that and fortunately the only
 293         # collision is 'FETCH_HEAD' but that shouldn't matter.
 294         # There may also be temp files with a Git_ prefix as well.
 295         if [ -n "$svn_mirror" ]; then
 296                 _randchar='[A-Za-z0-9_]'
 297                 _randchar2="$_randchar$_randchar"
 298                 _randchar4="$_randchar2$_randchar2"
 299                 _randchar10="$_randchar4$_randchar4$_randchar2"
 300                 find . -maxdepth 1 -type f -mmin +720 -name "$_randchar10" -print0 | xargs -0 rm -f
 301                 find . -maxdepth 1 -type f -mmin +720 -name "Git_*" -print0 | xargs -0 rm -f
 302         fi
 303
 304         # Remove any stale fast_import_crash_<pid> files that are more than 3 days old.
 305         if [ -n "$gfi_mirror" ]; then
 306                 find . -maxdepth 1 -type f -mmin +4320 -name "fast_import_crash_?*" -print0 | xargs -0 rm -f
 307         fi
 308 }
 309
 310 #
 311 ## Garbage Collection Types
 312 ##
 313 ## There are two kinds of possible garbage collection (gc) operations:
 314 ##
 315 ## 1.  A normal, full gc
 316 ## 2.  A "mini" gc
 317 ##
 318 ## If the full garbage collection interval has expired (or gc has never been
 319 ## run), then a normal, full gc will take place.  Otherwise, a "mini" gc will
 320 ## take place if the file .needsgc exists.
 321 ##
 322 ## A "mini" gc is similar to "git gc --auto" in that it may not end up actually
 323 ## doing anything unless the right conditions are present so it's not a burden
 324 ## to run it often.  If the file .needsgc exists, a "mini" gc will occur at
 325 ## the next opportunity.
 326 ##
 327 ## Note, however, that the .nogc file suppresses ALL gc activity (normal or mini).
 328 #
 329
 330 proj="${1%.git}"
 331 shift
 332 cd "$cfg_reporoot/$proj.git"
 333 [ -d objects/pack ] || { rm -f gfi-packs; mkdir -p objects/pack; }
 334 mirror_url="$(get_mirror_url)"
 335 svn_mirror=
 336 ! is_svn_mirror_url "$mirror_url" || svn_mirror=1
 337 gfi_mirror=
 338 if [ -f gfi-packs -a -s gfi-packs ] && is_gfi_mirror_url "$mirror_url"; then
 339         gfi_mirror=1
 340 fi
 341
 342 # If git config --bool --get girocco.redelta is explicitly false then automatic
 343 # redelta when there are less than $var_redelta_threshold objects will be suppressed.
 344 # On the other hand, if git config --get girocco.redelta is "always" then, on a full
 345 # gc only, for the final repack, deltas will always be recomputed.
 346 # This can be set on a per-project basis to avoid unusual pathological gc behavior.
 347 # Setting this will hurt efficiency of the affected repository.
 348 # Note that fast-import packs ALWAYS get new deltas regardless of this setting.
 349 noreusedeltaopt="--no-reuse-delta"
 350 [ "$(git config --bool --get girocco.redelta 2>/dev/null || :)" != "false" ] || noreusedeltaopt=
 351 alwaysredelta=
 352 [ "$(git config --get girocco.redelta 2>/dev/null || :)" != "always" ] || alwaysredelta=1
 353
 354 trap 'e=$?; rm -f .gc_in_progress; if [ $e != 0 ]; then echo "gc failed dir: $PWD" >&2; fi' EXIT
 355 trap 'exit 130' INT
 356 trap 'exit 143' TERM
 357
 358 # date -R is linux-only, POSIX equivalent is '+%a, %d %b %Y %T %z'
 359 datefmt='+%a, %d %b %Y %T %z'
 360
 361 isminigc=
 362 if check_interval lastgc $cfg_min_gc_interval; then
 363         if [ -e .needsgc ]; then
 364                 isminigc=1
 365         else
 366                 progress "= [$proj] garbage check skip (last at $(config_get lastgc))"
 367                 exit 0
 368         fi
 369 fi
 370 if [ -e .nogc ]; then
 371         progress "x [$proj] garbage check disabled"
 372         exit 0
 373 fi
 374
 375 if [ -n "$isminigc" ]; then
 376         # Perform a "mini" gc
 377         # Note that .delaygc is ignored here as that's only intended for full gc
 378         lock_gc
 379         rm -f .allowgc .needsgc
 380         remove_crud
 381         miniactive=
 382         if [ -f .svnpack ] && [ -n "$svn_mirror" ]; then
 383                 miniactive=1
 384                 progress "+ [$proj] mini garbage check (`date`)"
 385                 make_svn_pack
 386         fi
 387         if [ -z "$cfg_delay_gfi_redelta" ] && [ -n "$gfi_mirror" ]; then
 388                 # $Girocco::Config::delay_gfi_redelta is false, force redeltification now
 389                 if [ -z "$miniactive" ]; then
 390                         miniactive=1
 391                         progress "+ [$proj] mini garbage check (`date`)"
 392                 fi
 393                 repack_gfi_packs
 394         fi
 395         # If there aren't at least 10 non-keep, non-bitmap, non-bndl packs then
 396         # don't actually process them yet
 397         lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl --quiet"
 398         packcnt="$(list_packs --count $lpo objects/pack || :)"
 399         if [ "${packcnt:-0}" -ge 10 ]; then
 400                 if [ -z "$miniactive" ]; then
 401                         miniactive=1
 402                         progress "+ [$proj] mini garbage check (`date`)"
 403                 fi
 404                 if [ -n "$gfi_mirror" ]; then
 405                         repack_gfi_packs
 406                         packcnt="$(list_packs --count $lpo objects/pack || :)"
 407                 fi
 408                 # if repack_gfi_packs dropped the pack count to < 10 don't combine
 409                 if [ "${packcnt:-0}" -ge 10 ]; then
 410                         combine_small_packs
 411                         packcnt="$(list_packs --count $lpo objects/pack || :)"
 412                 fi
 413                 # if we still have more than 10 packs trigger a full gc
 414                 if [ "${packcnt:-0}" -ge 10 ]; then
 415                         # We shouldn't be in a .delaygc state at this point, but if
 416                         # we are then nuke it because we really need a full gc now
 417                         rm -f .delaygc
 418                         git config --unset gitweb.lastgc
 419                         rm -f "$lockf"
 420                         progress "- [$proj] mini garbage check triggering full gc too many packs (`date`)"
 421                         exit 0
 422                 fi
 423         fi
 424         rm -f "$lockf"
 425         if [ -n "$miniactive" ]; then
 426                 git update-server-info
 427                 progress "- [$proj] mini garbage check (`date`)"
 428         else
 429                 progress "= [$proj] mini garbage check nothing but crud removal to do (`date`)"
 430         fi
 431         exit 0
 432 fi
 433
 434 # Avoid unnecessary garbage collections:
 435 #   1. If lastreceive is set and is older than lastgc
 436 #   -AND-
 437 #   2. We are not a fork (! -s alternates) -OR- lastparentgc is older than lastgc
 438
 439 # If lastgc is NOT set or lastreceive is NOT set we MUST run gc
 440 # If we are a fork and lastparentgc is NOT set we MUST run gc
 441
 442 # If the repo is dirty after removing any crud we MUST run gc
 443
 444 gcstart="$(date "$datefmt")"
 445 skipgc=
 446 isfork=
 447 [ -s objects/info/alternates ] && isfork=1
 448 lastparentgcsecs=
 449 [ -n "$isfork" ] && lastparentgcsecs="$(config_get_date_seconds lastparentgc || :)"
 450 lastreceivesecs=
 451 if lastreceivesecs="$(config_get_date_seconds lastreceive)" && \
 452    lastgcsecs="$(config_get_date_seconds lastgc)" && \
 453    [ $lastreceivesecs -lt $lastgcsecs ]; then
 454         # We've run gc since we last received, so maybe we can skip,
 455         # check if not fork or fork and lastparentgc < lastgc
 456         if [ -n "$isfork" ]; then
 457                 if [ -n "$lastparentgcsecs" ] && \
 458                    [ $lastparentgcsecs -lt $lastgcsecs ]; then
 459                         # We've run gc since our parent ran gc so we can skip
 460                         skipgc=1
 461                 fi
 462         else
 463                 # We don't have any alternates (we're not a forK) so we can skip
 464                 skipgc=1
 465         fi
 466 fi
 467
 468 # Prevent any other simultaneous gc operations
 469 lock_gc
 470
 471 # At this point, if .allowgc exists, it's now crud to be removed
 472 rm -f .allowgc
 473
 474 # Ideally we would do this in post-receive, but that would mean duplicating the
 475 # logic so it's available in the chroot jail and that's highly undesirable
 476 # Instead, since the first gc will be triggered immediately following the first
 477 # push, we do the check here as it's quick and harmless if HEAD is already valid
 478 check_and_set_head || :
 479
 480 # Always get rid of crud
 481 remove_crud
 482
 483 # Run 'git svn gc' now for svn mirrors
 484 if [ -n "$svn_mirror" ]; then
 485         git svn gc || :
 486 fi
 487
 488 # Skip the actual gc if .delaygc is set
 489 if [ -e .delaygc ]; then
 490         progress "x [$proj] garbage check delayed (except for crud removal)"
 491         rm -f "$lockf"
 492         exit 0
 493 fi
 494
 495 # Do not skip gc if the repo is dirty
 496 if [ -n "$skipgc" ] && ! is_dirty; then
 497         progress "= [$proj] garbage check nothing but crud removal to do (`date`)"
 498         config_set lastgc "$gcstart"
 499         rm -f "$lockf"
 500         exit 0
 501 fi
 502
 503 bumptime=
 504 if [ -n "$isfork" ] && [ -z "$lastparentgcsecs" ]; then
 505         # set lastparentgc and then update gcstart to be at least 1 second later
 506         config_set lastparentgc "$gcstart"
 507         bumptime=1
 508 fi
 509 if [ -z "$lastreceivesecs" ]; then
 510         # set lastreceive and then update gcstart to be at least 1 second later
 511         config_set lastreceive "$gcstart"
 512         bumptime=1
 513 fi
 514 if [ -n "$bumptime" ]; then
 515         sleep 1
 516         gcstart="$(date "$datefmt")"
 517 fi
 518
 519 progress "+ [$proj] garbage check (`date`)"
 520
 521 newdeltas=
 522 [ -z "$alwaysredelta" ] || newdeltas=-f
 523 if [ -z "$newdeltas" ] && [ -n "$gfi_mirror" ]; then
 524         if [ $(list_packs --exclude-no-idx --count objects/pack) -le \
 525              $(list_packs --exclude-no-idx --count --quiet --only gfi-packs) ]; then
 526                 # Don't bother with repack_gfi_packs since everything's being repacked
 527                 newdeltas=-f
 528         fi
 529 fi
 530 if [ -z "$newdeltas" ] && [ -n "$noreusedeltaopt" ] && \
 531    [ $(list_packs --exclude-no-idx --count-objects objects/pack) -le $var_redelta_threshold ]; then
 532         # There aren't enough objects to worry about so just redelta to get the best pack
 533         newdeltas=-f
 534 fi
 535 if [ -z "$newdeltas" ]; then
 536         # Since we're not going to recompute deltas overall, we need to do the
 537         # "mini" maintenance so that we can get more optimal deltas
 538         [ -z "$noreusedeltaopt" ] || make_svn_pack
 539         repack_gfi_packs
 540         force_single_pack_redelta=
 541         [ -n "$gfi_mirror" ] || [ -n "$svn_mirror" ] || force_single_pack_redelta=1
 542         [ -z "$noreusedeltaopt" ] || combine_small_packs $force_single_pack_redelta
 543 fi
 544
 545 #
 546 ## Safe Pruning In Forks
 547 ##
 548 ## We are about to perform garbage collection.  We do NOT use the "git gc"
 549 ## command directly as it does not provide enough control over the fine details
 550 ## that we require.  However, we DO maintain a "gc.pid" file during our garbage
 551 ## collection so that a simultaneous "git gc" by an administrator will be
 552 ## blocked (and similarly we refuse to start garbage collection if we cannot
 553 ## create the "gc.pid" file).  When we say "gc" in the below description we are
 554 ## referring to our "gc.sh" script, NOT the "git gc" command.
 555 ##
 556 ## If the project we are running garbage collection (gc) on has any forks we
 557 ## must be careful not to remove any objects that while no longer referenced by
 558 ## this project (the parent) are still referenced by one or more forks (the
 559 ## children) otherwise the children will become corrupt and we can't abide
 560 ## corrupt children.
 561 ##
 562 ## One way to accomplish this is to simply hard-link all currently existing
 563 ## loose objects and packs in the parent into all the children that refer to the
 564 ## parent (via a line in their objects/info/alternates file) before beginning
 565 ## the gc operation and then relying on a subsequent gc in the child to clean up
 566 ## any excess objects/packs.  We used to use this strategy but it's very
 567 ## inefficient because:
 568 ##
 569 ##   1. The disk space used by the old pack(s)/object(s) will not be reclaimed
 570 ##      until all children (and their children, if any) run gc by which time
 571 ##      it's quite possible the topmost parent will have run gc again and
 572 ##      hard-linked yet another old pack down to its children (not to mention
 573 ##      loose objects).
 574 ##
 575 ##   2. As we are now using the "-A" option with "git repack", any new objects
 576 ##      in the parent that are not referenced by children will continually get
 577 ##      exploded out of the hard-linked pack in the children whenever the
 578 ##      children run gc.
 579 ##
 580 ##   3. To avoid suboptimal and/or unnecessarily many packs being hard-linked
 581 ##      into child forks, we must run the "mini" gc maintenance before we
 582 ##      perform the hard-linking into the children which provides yet another
 583 ##      source of inefficiency.
 584 ##
 585 ## Since we are using the "-A" option to "git repack" (that was not always the
 586 ## case) to guarantee we can access old ref values for long enough to send out
 587 ## a meaningful mail.sh notification, we now have another, more efficient,
 588 ## option available to prevent corruption of child forks that continue to refer
 589 ## to objects that are no longer reachable from any ref in the parent.
 590 ##
 591 ## The only things that need be copied (or hard-linked) into the child fork(s)
 592 ## are those objects that have become unreachable from any ref in the parent.
 593 ## They are the only things that could ever be removed by "git prune" and
 594 ## therefore the only things we need to prevent the loss of in order to avoid
 595 ## corruption of the child fork(s).
 596 ##
 597 ## Therefore we now use the following strategy instead to avoid excessive disk
 598 ## use and lots of unnecessary loose objects in child forks:
 599 ##
 600 ##   1. Run "git repack -A -d -l" in the parent BEFORE doing anything about
 601 ##      child forks.
 602 ##
 603 ##   2. Hard-link all remaining existing loose objects in the parent into the
 604 ##      immediate child forks.
 605 ##
 606 ##   3. Now run "git prune" in the parent.
 607 ##
 608 ## With this new strategy we avoid the need to run any "mini" gc maintenance
 609 ## before copying (or hard-linking) anything down to the child forks.
 610 ## Furthermore, only when the parent performs a non-fast-forward update will
 611 ## anything ever be transferred to the children leaving them unperturbed in the
 612 ## vast majority of cases.  Finally, even if the parent references objects the
 613 ## children do not, those objects will no longer continually end up in the
 614 ## children as unreachable loose objects after the children run gc.
 615 #
 616
 617 git pack-refs --all
 618 touch .gc_in_progress
 619 rm -f .gc_failed bundles/*
 620 rm -f objects/pack/pack-*.bndl
 621 # We use the -A option with git repack so that unreachable objects can live
 622 # on for a time as loose objects.  This is particularly helpful if we just
 623 # happen to be in the process of sending out a ref update for a ref that was
 624 # force updated and the old ref value would have otherwise been removed by
 625 # repack because it was now unreachable.  Admittedly the window for gc to run
 626 # and do that before we manage to send out the ref update is not large, but
 627 # it would not be difficult to create such a situation.  Unfortunately, when
 628 # Git unpacks these unreachable objects it will give them the modification
 629 # time of the *.pack file they came out of.  This could be very, very old.
 630 # If that happens, the subsequent git prune --expire some_time_ago will still
 631 # remove the object(s) and our pending ref update will still lose out.
 632 # To prevent this from happening and to get the behavior we want, we now
 633 # touch the modification time of all pack-<sha>.pack files so that any
 634 # loosened objects get a current time.  Git does not provide any other
 635 # mechanism to do this.  We do not want to just touch all loose objects
 636 # left after the repack because that would cause objects that were loosened
 637 # previously to live on which we definitely do not want.
 638 list_packs --exclude-no-idx objects/pack | xargs touch -c 2>/dev/null || :
 639 # We wish to keep deltas from our last full pack so if we're not redeltaing
 640 # then make sure the .pack associated with the .bitmap has a newer mod time
 641 # (If there is no .bitmap then touch the pack with the most objects instead.)
 642 if [ -z "$newdeltas" ]; then
 643         bmpack="$(list_packs --exclude-no-bitmap --exclude-no-idx --max-matches 1 objects/pack)"
 644         [ -n "$bmpack" ] || bmpack="$(list_packs --exclude-no-idx --max-matches 1 --object-limit -1 --include-boundary objects/pack)"
 645         if [ -n "$bmpack" ] && [ -f "$bmpack" -a -s "$bmpack" ]; then
 646                 sleep 1
 647                 touch -c "$bmpack" 2>/dev/null || :
 648         fi
 649 fi
 650 # The git repack command only supports bitmaps if all objects are being packed.
 651 # While it is theoretically possible that a project with a non-empty alternates
 652 # file ends up packing all objects (because it does not actually use any of the
 653 # objects found in the alternates), it's very unlikely.  And, in the unlikely
 654 # event that did occur, clients would see a message about only using one bitmap
 655 # because Git can only use one bitmap at a time and at least one of the
 656 # alternates is bound to have a bitmap.  Therefore if we see a non-empty
 657 # alternates file, we disable writing bitmaps which avoids the warning and any
 658 # possibility of a client warning as well.
 659 nobm=
 660 [ -z "$var_have_git_172" ] || ! [ -s objects/info/alternates ] || \
 661         nobm='-c repack.writebitmaps=false -c pack.writebitmaps=false'
 662 progress "~ [$proj] running full gc repack${nobm:+ (bitmaps disabled)}"
 663 git $nobm repack $packopts -A -d -l $quiet $newdeltas $@
 664 [ ! -e .gc_failed ] || exit 1
 665 # These, if they exist, are now meaningless and need to be removed
 666 rm -f gfi-packs .needsgc .svnpack .svnpackgc
 667 allpacks="$(echo objects/pack/pack-$octet20.pack)"
 668 curhead="$(cat HEAD)"
 669 pkrf=
 670 [ ! -e packed-refs ] || pkrf=packed-refs
 671 eval "reposizek=$(( $(echo 0 $(du -k $pkrf $allpacks 2>/dev/null | LC_ALL=C awk '{print $1}') | \
 672         LC_ALL=C sed -e 's/ / + /g') ))"
 673 git update-server-info
 674 # The -A option to `git repack` may have caused some loose objects to pop
 675 # out of their packs.  We must make these objects group writable so that they
 676 # can be freshened by other pushers.  Technically we need only do this for
 677 # push projects but to enable mirror projects to be more easily converted to
 678 # push projects, we go ahead and do it for all projects.
 679 { find objects/$octet -type f -name "$octet19" -print0 | xargs -0 chmod ug+w || :; } 2>/dev/null
 680
 681 if has_forks "$proj"; then
 682         progress "~ [$proj] hard-linking loose objects into immediate child forks"
 683         # We have to update the lastparentgc time in the child forks even if they do not get any
 684         # new "loose objects" because they need to run gc just in case the parent now has some
 685         # objects that used to only be in the child so they can be removed from the child.
 686         # For example, a "patch" might be developed first in a fork and then later accepted into
 687         # the parent in which case the objects making up the patch in the child fork are now
 688         # redundant (since they're now in the parent as well) and need to be removed from the
 689         # child fork which can only happen if the child fork runs gc.
 690         shbin="${cfg_posix_sh_bin:-/bin/sh}"
 691         forkdir="$proj"
 692         # It is enough to copy objects just one level down and get_repo_list
 693         # takes a regular expression (which is automatically prefixed with '^')
 694         # so we can easily match forks exactly one level down from this project
 695         get_repo_list "$forkdir/[^/]*:" |
 696                 while read fork; do
 697                         # Ignore forks that do not exist or are symbolic links
 698                         [ ! -L "$cfg_reporoot/$fork.git" -a -d "$cfg_reporoot/$fork.git" ] || \
 699                                 continue
 700                         # Or do not have a non-zero length alternates file
 701                         [ -s "$cfg_reporoot/$fork.git/objects/info/alternates" ] || \
 702                                 continue
 703                         # Match objects in parent project
 704                         for d in objects/$octet; do
 705                                 [ "$d" != "objects/$octet" ] || continue
 706                                 mkdir -p "$cfg_reporoot/$fork.git/$d"
 707                                 find "$d" -maxdepth 1 -type f -name "$octet19" -print0 |
 708                                         xargs -0 "$shbin" -c 'ln -f "$@" '"'$cfg_reporoot/$fork.git/$d/'" sh || :
 709                         done
 710                         # Update the fork's lastparentgc date (must be current, not $gcstart)
 711                         git --git-dir="$cfg_reporoot/$fork.git" config \
 712                                 gitweb.lastparentgc "$(date "$datefmt")"
 713                 done
 714 fi
 715
 716 # The git prune command does not take a -q or --quiet but started outputting
 717 # 'Checking connectivity' progress messages in v1.7.9.  However, we can
 718 # suppress those by piping through cat as it only activates the progress
 719 # messages when stderr is a tty.  We only expire loose objects older than one
 720 # day just in case there's some pending action (such as sending out a ref
 721 # update) in progress that might want to examine them.  This may leave us with
 722 # loose objects.  That's okay because at the next gc interval, we will always
 723 # run gc if we see any loose objects regardless of whether or not we've seen
 724 # any updates or we've received new linked objects from our parent.  Note that
 725 # in order to keep loose objects that just recently became unreferenced but
 726 # have a very old modification date around we rely on some help from both the
 727 # update.sh and hooks/pre-receive scripts.  Furthermore, since Git v2.2.0
 728 # (d3038d22 prune: keep objects reachable from recent objects) an unreachable
 729 # object that would otherwise be pruned (because it's too old) will be kept
 730 # alive by an unreachable object that refers to it that's not old enough to
 731 # be pruned yet.
 732 prunecmd='git prune --expire 1_day_ago'
 733 [ -n "$show_progress" ] || \
 734 prunecmd="{ $prunecmd 2>&1 || touch .gc_failed; } | cat"
 735 progress "~ [$proj] pruning expired unreachable loose objects"
 736 eval "$prunecmd"
 737 [ ! -e .gc_failed ] || exit 1
 738
 739 # darcs:// mirrors have a xxx.log file that will grow endlessly
 740 # if this is a mirror and the file exists, shorten it to 10000 lines
 741 # also take this opportunity to optimize the darcs repo
 742 if [ ! -e .nofetch ] && [ -n "$cfg_mirror" ]; then
 743         url="$(config_get baseurl || :)"
 744         case "$url" in darcs://*)
 745                 if [ -n "$cfg_mirror_darcs" ]; then
 746                         url="${url%/}"
 747                         basedarcs="$(basename "${url#darcs:/}")"
 748                         if [ -f "$basedarcs.log" ]; then
 749                                 tail -n 10000 "$basedarcs.log" > "$basedarcs.log.$$"
 750                                 mv -f "$basedarcs.log.$$" "$basedarcs.log"
 751                         fi
 752                         if [ -d "$basedarcs.darcs" ]; then
 753                                 (
 754                                         cd "$basedarcs.darcs"
 755                                         # without show_progress suppress non-error output
 756                                         [ -n "$show_progress" ] || exec >/dev/null
 757                                         # Note that this does not optimize _darcs/inventories/ :(
 758                                         darcs optimize || :
 759                                 )
 760                         fi
 761                 fi
 762         esac
 763 fi
 764
 765 # Create a matching .bndl header file for the all-in-one pack we just created
 766 # but only if we're not a fork (otherwise the bundle would not be complete)
 767 # and we are running at least Git version 1.7.2 (pack_is_complete always fails otherwise)
 768 if [ ! -s objects/info/alternates ] && [ -n "$var_have_git_172" ]; then
 769         # There should only be one pack in $allpacks but if there was a
 770         # simultaneous push...
 771         # The one we just created will have a .idx and will NOT have a .keep
 772         progress "~ [$proj] creating downloadble bundle header"
 773         pkfound=
 774         pkhead=
 775         for pk in $allpacks; do
 776                 [ -s "$pk" ] || continue
 777                 pkbase="${pk%.pack}"
 778                 [ -s "$pkbase.idx" ] || continue
 779                 [ ! -e "$pkbase.keep" ] || continue
 780                 if pkhead="$(pack_is_complete "$PWD/$pk" "$PWD/packed-refs" "$curhead")"; then
 781                         pkfound="$pkbase"
 782                         break;
 783                 fi
 784         done
 785         if [ -n "$pkfound" -a -n "$pkhead" ]; then
 786                 {
 787                         echo "# v2 git bundle"
 788                         LC_ALL=C sed -ne "/^$octet20 refs\/[^ $tab]*\$/ p" < packed-refs
 789                         echo "$pkhead HEAD"
 790                         echo ""
 791                 } > "$pkbase.bndl"
 792                 bndletag="$("$cfg_basedir/bin/rangecgi" --etag -m 1 "$pkbase.bndl" "$pkbase.pack" || :)"
 793                 bndlsha="$(printf '%s' "$bndletag" | git hash-object --stdin || :)"
 794                 if [ -n "$bndletag" ]; then
 795                         case "$bndlsha" in $octet20)
 796                                 bndlshatrailer="${bndlsha#????????}"
 797                                 bndlshaprefix="${bndlsha%$bndlshatrailer}"
 798                                 bndlname="$(TZ=UTC date +%Y%m%d_%H%M%S)-${bndlshaprefix:-0}"
 799                                 [ -d bundles ] || mkdir bundles
 800                                 echo "${pkbase#objects/pack/}.bndl" > "bundles/$bndlname"
 801                                 echo "${pkbase#objects/pack/}.pack" >> "bundles/$bndlname"
 802                                 ln -s -f -n "$bndlname" bundles/latest
 803                         esac
 804                 fi
 805         fi
 806 fi
 807
 808 # Record the size of this repo as the sum of its *.pack sizes as 1024-byte blocks
 809 config_set_raw girocco.reposizek "${reposizek:-0}"
 810
 811 # We use $gcstart here to avoid a race where a push occurs during the gc itself
 812 # and the next future gc could be incorrectly skipped if we used the current
 813 # timestamp here instead
 814 config_set lastgc "$gcstart"
 815 rm -f "$lockf"
 816
 817 progress "- [$proj] garbage check (`date`)"