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