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 # if the optional argument is non-empty even a single small pack will be redeltad
 102 combine_small_packs() {
 103         _minsmallpacks=2
 104         if [ -n "$1" ] && [ -n "$noreusedeltaopt" ]; then
 105                 _minsmallpacks=1
 106         fi
 107         _lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl --quiet"
 108         _lpo="$_lpo --object-limit $var_redelta_threshold objects/pack"
 109         while
 110                 _cnt="$(list_packs --count $_lpo || :)"
 111                 test "${_cnt:-0}" -ge $_minsmallpacks
 112         do
 113                 _newp="$(list_packs $_lpo | combine_packs --names $noreusedeltaopt)"
 114                 _newc="$(echo $(echo "$_newp" | LC_ALL=C wc -w))"
 115                 # be paranoid and exit the loop if we haven't reduced the number of packs
 116                 [ $_newc -lt $_cnt ] || break
 117                 _minsmallpacks=2
 118         done
 119         return 0
 120 }
 121
 122 # Unfortunately git-svn lacks the ability to store newly fetched revisions as a pack.
 123 # However, the fetch code conveniently sets .svnpack just before it runs git-svn fetch
 124 # so that it's easy to find all the objects that have been fetched by git-svn and
 125 # combine them into a pack.  The --no-reuse-delta option is meaningless here since
 126 # everything to be packed is a loose object and therefore not a delta so deltification
 127 # will always take place.
 128 make_svn_pack() {
 129         [ -f .svnpack ] && is_svn_mirror || return 0
 130         rm -f .svnpackgc
 131         mv -f .svnpack .svnpackgc
 132         _newp="$(find objects/$octet -maxdepth 1 -type f -newer .svnpackgc -name "$octet19" -print 2>/dev/null |
 133         LC_ALL=C awk -F / '{print $2 $3}' |
 134         run_combine_packs --objects --names $packopts --incremental --all-progress-implied $quiet --non-empty)" || {
 135                 mv -f .svnpackgc .svnpack
 136                 >.gc_failed
 137                 return 1
 138         }
 139         if [ -n "$_newp" ]; then
 140                 # remove the now-redundant loose objects -- this is always safe
 141                 # even during a concurrent push because a reprepare_packed_git
 142                 # will be triggered if an object that should be there is not
 143                 # found thereby finding it in the new pack instead
 144                 git prune-packed $quiet
 145         fi
 146         rm -f .svnpackgc
 147 }
 148
 149 # HEADSHA="$(pack_is_complete /full/path/to/some.pack /full/path/to/packed-refs "$(cat HEAD)")"
 150 pack_is_complete() {
 151         # Must have a matching .idx file and a non-empty packed-refs file
 152         [ -s "${1%.pack}.idx" ] || return 1
 153         [ -s "$2" ] || return 1
 154         _headsha=
 155         case "$3" in
 156                 $octet20)
 157                         _headsha="$3"
 158                         ;;
 159                 "ref: refs/"?*|"ref:refs/"?*|"refs/"?*)
 160                         _headmatch="${3#ref:}"
 161                         _headmatch="${_headmatch# }"
 162                         _headmatchpat="$(echo "$_headmatch" | LC_ALL=C sed -e 's/\([.$]\)/\\\1/g')"
 163                         _headsha="$(LC_ALL=C grep -e "^$octet20 $_headmatchpat\$" < "$2" | \
 164                                     LC_ALL=C cut -d ' ' -f 1)"
 165                         case "$_headsha" in $octet20) :;; *)
 166                                 return 1
 167                         esac
 168                         ;;
 169                 *)
 170                         # bad HEAD
 171                         return 1
 172         esac
 173         rm -rf pack_is_complete_test
 174         mkdir pack_is_complete_test
 175         mkdir pack_is_complete_test/refs
 176         mkdir pack_is_complete_test/objects
 177         mkdir pack_is_complete_test/objects/pack
 178         echo "$_headsha" > pack_is_complete_test/HEAD
 179         ln -s "$1" pack_is_complete_test/objects/pack/
 180         ln -s "${1%.pack}.idx" pack_is_complete_test/objects/pack/
 181         ln -s "$2" pack_is_complete_test/packed-refs
 182         _count="$(git --git-dir=pack_is_complete_test rev-list --count --all 2>/dev/null || :)"
 183         rm -rf pack_is_complete_test
 184         [ -n "$_count" ] || return 1
 185         [ "$_count" -gt 0 ] 2>/dev/null || return 1
 186         echo "$_headsha"
 187 }
 188
 189 # On return a "$lockf" will have been created that must be removed when gc is done
 190 lock_gc() {
 191         # be compatibile with gc.pid file from newer Git releases
 192         lockf=gc.pid
 193         hn="$(hostname)"
 194         active=
 195         if [ "$(createlock "$lockf")" ]; then
 196                 # If $lockf is:
 197                 #   1) less than 12 hours old
 198                 #   2) contains two fields (pid hostname) NO trailing NL
 199                 #   3) the hostname is different OR the pid is still alive
 200                 # then we exit as another active process is holding the lock
 201                 if [ "$(find "$lockf" -maxdepth 1 -mmin -720 -print 2>/dev/null)" ]; then
 202                         apid=
 203                         ahost=
 204                         read -r apid ahost ajunk < "$lockf" || :
 205                         if [ "$apid" ] && [ "$ahost" ]; then
 206                                 if [ "$ahost" != "$hn" ] || pidactive "$apid"; then
 207                                         active=1
 208                                 fi
 209                         fi
 210                 fi
 211         else
 212                 echo >&2 "[$proj] unable to create gc.pid.lock file"
 213                 exit 1
 214         fi
 215         if [ -n "$active" ]; then
 216                 rm -f "$lockf.lock"
 217                 echo >&2 "[$proj] gc already running on machine '$ahost' pid '$apid'"
 218                 exit 1
 219         fi
 220         printf "%s %s" "$$" "$hn" > "$lockf.lock"
 221         chmod 0664 "$lockf.lock"
 222         mv -f "$lockf.lock" "$lockf"
 223 }
 224
 225 # Remove any crud that's been left behind by interrupted operations
 226 # that did not clean up after themselves
 227 remove_crud() {
 228         # Remove any existing FETCH_HEAD
 229         # There can only be a FETCH_HEAD if we've been fetching, not if we've been
 230         # receiving pushes (those never create a FETCH_HEAD).
 231         # And if we're fetching because we're a mirror, we know we're not fetching right
 232         # now since jobd.pl never runs a project's fetch simultaneously with its gc.
 233         # Therefore any existing FETCH_HEAD is junk.  And it may be many megabytes if
 234         # there were a lot of refs.
 235         rm -f FETCH_HEAD
 236
 237         # Remove any stale pack remnants that are more than an hour old.
 238         # Stale pack fragments are defined as any pack-<sha1>.ext where .ext is NOT
 239         # .pack AND the corresponding .pack DOES NOT exist.  A bunch of stale
 240         # pack-<sha1>.idx files without their corresponding .pack files are worthless
 241         # and just waste space.  Normally there shouldn't be any remnants but actually
 242         # this can happen when things are interrupted at just the wrong time.
 243         # Note that the objects/pack directory is created by git init and should
 244         # always exist.
 245         find objects/pack -maxdepth 1 -type f -mmin +60 -name "pack-$octet20.?*" -print | \
 246         LC_ALL=C sed -e 's/^objects\/pack\/pack-//; s/\..*$//' | LC_ALL=C sort -u | \
 247         while read packsha; do
 248                 [ ! -e "objects/pack/pack-$packsha.pack" ] || continue
 249                 rm -f "objects/pack/pack-$packsha".?*
 250         done
 251
 252         # Remove any stale pack .keep files that are more than 12 hours old.
 253         # We don't do anything to create any permanent pack .keep files, so they must
 254         # be remnants from some failed push or something.  Removing the .keep will
 255         # allow the pack to be properly repacked.
 256         find objects/pack -maxdepth 1 -type f -mmin +720 -name "pack-$octet20.keep" -print0 | xargs -0 rm -f
 257
 258         # Remove any stale tmp_pack_*, tmp_idx_*, tmp_bitmap_*, packtmp-* or .tmp-*-pack files
 259         # that are more than 12 hours old.
 260         find objects/pack -maxdepth 1 -type f -mmin +720 \( \
 261                 -name "tmp_pack_?*" -o -name "tmp_idx_?*" -o -name "tmp_bitmap_?*" -o \
 262                 -name "packtmp-?*" -o -name ".tmp-?*-pack" \
 263                 \) -print0 | xargs -0 rm -f
 264
 265         # Remove any stale shallow_* files that are more than 12 hours old.
 266         # These can be left behind by Git >= 1.8.4.2 and < 2.0.0 when a client
 267         # requests a shallow clone.
 268         find . -maxdepth 1 -type f -mmin +720 -name "shallow_?*" -print0 | xargs -0 rm -f
 269
 270         # Remove any stale *.temp files in the objects area that are more than 12 hours old.
 271         # This can be stale sha1.temp, or stale *.pack.temp so we kill all stale *.temp.
 272         find objects -type f -mmin +720 -name "*.temp" -print0 | xargs -0 rm -f
 273
 274         # Remove any stale *.lock files in the htmlcache area that might have been left
 275         # behind after an abnormal exit during an attempt to update a cached file and
 276         # are more than 1 hour old.
 277         ! [ -d htmlcache ] || find htmlcache -type f -mmin +60 -name "*.lock" -print0 | xargs -0 rm -f
 278
 279         # Remove any stale git-svn temp files that are more than 12 hours old.
 280         # The git-svn process creates temp files with random 10 character names
 281         # in the root of $GIT_DIR.  Unfortunately they do not have a recognizable
 282         # prefix, so we just have to kill any files with a 10-character name.  We
 283         # do this only for git-svn mirrors.  All characters are chosen from
 284         # [A-Za-z0-9_] so we can at least check that and fortunately the only
 285         # collision is 'FETCH_HEAD' but that shouldn't matter.
 286         # There may also be temp files with a Git_ prefix as well.
 287         if is_svn_mirror; then
 288                 _randchar='[A-Za-z0-9_]'
 289                 _randchar2="$_randchar$_randchar"
 290                 _randchar4="$_randchar2$_randchar2"
 291                 _randchar10="$_randchar4$_randchar4$_randchar2"
 292                 find . -maxdepth 1 -type f -mmin +720 -name "$_randchar10" -print0 | xargs -0 rm -f
 293                 find . -maxdepth 1 -type f -mmin +720 -name "Git_*" -print0 | xargs -0 rm -f
 294         fi
 295
 296         # Remove any stale fast_import_crash_<pid> files that are more than 3 days old.
 297         if is_gfi_mirror; then
 298                 find . -maxdepth 1 -type f -mmin +4320 -name "fast_import_crash_?*" -print0 | xargs -0 rm -f
 299         fi
 300 }
 301
 302 #
 303 ## Garbage Collection Types
 304 ##
 305 ## There are two kinds of possible garbage collection (gc) operations:
 306 ##
 307 ## 1.  A normal, full gc
 308 ## 2.  A "mini" gc
 309 ##
 310 ## If the full garbage collection interval has expired (or gc has never been
 311 ## run), then a normal, full gc will take place.  Otherwise, a "mini" gc will
 312 ## take place if the file .needsgc exists.
 313 ##
 314 ## A "mini" gc is similar to "git gc --auto" in that it may not end up actually
 315 ## doing anything unless the right conditions are present so it's not a burden
 316 ## to run it often.  If the file .needsgc exists, a "mini" gc will occur at
 317 ## the next opportunity.
 318 ##
 319 ## Note, however, that the .nogc file suppresses ALL gc activity (normal or mini).
 320 #
 321
 322 proj="${1%.git}"
 323 shift
 324 cd "$cfg_reporoot/$proj.git"
 325
 326 # If git config --bool --get girocco.redelta is explicitly false then automatic
 327 # redelta when there are less than $var_redelta_threshold objects will be suppressed.
 328 # On the other hand, if git config --get girocco.redelta is "always" then, on a full
 329 # gc only, for the final repack, deltas will always be recomputed.
 330 # This can be set on a per-project basis to avoid unusual pathological gc behavior.
 331 # Setting this will hurt efficiency of the affected repository.
 332 # Note that fast-import packs ALWAYS get new deltas regardless of this setting.
 333 noreusedeltaopt="--no-reuse-delta"
 334 [ "$(git config --bool --get girocco.redelta 2>/dev/null || :)" != "false" ] || noreusedeltaopt=
 335 alwaysredelta=
 336 [ "$(git config --get girocco.redelta 2>/dev/null || :)" != "always" ] || alwaysredelta=1
 337
 338 trap 'e=$?; rm -f .gc_in_progress; if [ $e != 0 ]; then echo "gc failed dir: $PWD" >&2; fi' EXIT
 339 trap 'exit 130' INT
 340 trap 'exit 143' TERM
 341
 342 # date -R is linux-only, POSIX equivalent is '+%a, %d %b %Y %T %z'
 343 datefmt='+%a, %d %b %Y %T %z'
 344
 345 isminigc=
 346 if check_interval lastgc $cfg_min_gc_interval; then
 347         if [ -e .needsgc ]; then
 348                 isminigc=1
 349         else
 350                 progress "= [$proj] garbage check skip (last at $(config_get lastgc))"
 351                 exit 0
 352         fi
 353 fi
 354 if [ -e .nogc ]; then
 355         progress "x [$proj] garbage check disabled"
 356         exit 0
 357 fi
 358
 359 if [ -n "$isminigc" ]; then
 360         # Perform a "mini" gc
 361         # Note that .delaygc is ignored here as that's only intended for full gc
 362         lock_gc
 363         rm -f .allowgc .needsgc
 364         remove_crud
 365         miniactive=
 366         if [ -f .svnpack ] && is_svn_mirror; then
 367                 miniactive=1
 368                 progress "+ [$proj] mini garbage check (`date`)"
 369                 make_svn_pack
 370         fi
 371         # If there aren't at least 10 non-keep, non-bitmap, non-bndl packs then
 372         # don't actually process them yet
 373         lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl --quiet"
 374         packcnt="$(list_packs --count $lpo objects/pack || :)"
 375         if [ "${packcnt:-0}" -ge 10 ]; then
 376                 if [ -z "$miniactive" ]; then
 377                         miniactive=1
 378                         progress "+ [$proj] mini garbage check (`date`)"
 379                 fi
 380                 if [ -f gfi-packs -a -s gfi-packs ] && is_gfi_mirror; then
 381                         repack_gfi_packs
 382                         packcnt="$(list_packs --count $lpo objects/pack || :)"
 383                 fi
 384                 # if repack_gfi_packs dropped the pack count to < 10 don't combine
 385                 if [ "${packcnt:-0}" -ge 10 ]; then
 386                         combine_small_packs
 387                         packcnt="$(list_packs --count $lpo objects/pack || :)"
 388                 fi
 389                 # if we still have more than 10 packs trigger a full gc
 390                 if [ "${packcnt:-0}" -ge 10 ]; then
 391                         # We shouldn't be in a .delaygc state at this point, but if
 392                         # we are then nuke it because we really need a full gc now
 393                         rm -f .delaygc
 394                         git config --unset gitweb.lastgc
 395                         rm -f "$lockf"
 396                         progress "- [$proj] mini garbage check triggering full gc too many packs (`date`)"
 397                         exit 0
 398                 fi
 399         fi
 400         rm -f "$lockf"
 401         if [ -n "$miniactive" ]; then
 402                 git update-server-info
 403                 progress "- [$proj] mini garbage check (`date`)"
 404         else
 405                 progress "= [$proj] mini garbage check nothing but crud removal to do (`date`)"
 406         fi
 407         exit 0
 408 fi
 409
 410 # Avoid unnecessary garbage collections:
 411 #   1. If lastreceive is set and is older than lastgc
 412 #   -AND-
 413 #   2. We are not a fork (! -s alternates) -OR- lastparentgc is older than lastgc
 414
 415 # If lastgc is NOT set or lastreceive is NOT set we MUST run gc
 416 # If we are a fork and lastparentgc is NOT set we MUST run gc
 417
 418 # If the repo is dirty after removing any crud we MUST run gc
 419
 420 gcstart="$(date "$datefmt")"
 421 skipgc=
 422 isfork=
 423 [ -s objects/info/alternates ] && isfork=1
 424 lastparentgcsecs=
 425 [ -n "$isfork" ] && lastparentgcsecs="$(config_get_date_seconds lastparentgc || :)"
 426 lastreceivesecs=
 427 if lastreceivesecs="$(config_get_date_seconds lastreceive)" && \
 428    lastgcsecs="$(config_get_date_seconds lastgc)" && \
 429    [ $lastreceivesecs -lt $lastgcsecs ]; then
 430         # We've run gc since we last received, so maybe we can skip,
 431         # check if not fork or fork and lastparentgc < lastgc
 432         if [ -n "$isfork" ]; then
 433                 if [ -n "$lastparentgcsecs" ] && \
 434                    [ $lastparentgcsecs -lt $lastgcsecs ]; then
 435                         # We've run gc since our parent ran gc so we can skip
 436                         skipgc=1
 437                 fi
 438         else
 439                 # We don't have any alternates (we're not a forK) so we can skip
 440                 skipgc=1
 441         fi
 442 fi
 443
 444 # Prevent any other simultaneous gc operations
 445 lock_gc
 446
 447 # At this point, if .allowgc exists, it's now crud to be removed
 448 rm -f .allowgc
 449
 450 # Always get rid of crud
 451 remove_crud
 452
 453 # Run 'git svn gc' now for svn mirrors
 454 if is_svn_mirror; then
 455         git svn gc || :
 456 fi
 457
 458 # Skip the actual gc if .delaygc is set
 459 if [ -e .delaygc ]; then
 460         progress "x [$proj] garbage check delayed (except for crud removal)"
 461         rm -f "$lockf"
 462         exit 0
 463 fi
 464
 465 # Do not skip gc if the repo is dirty
 466 if [ -n "$skipgc" ] && ! is_dirty; then
 467         progress "= [$proj] garbage check nothing but crud removal to do (`date`)"
 468         config_set lastgc "$gcstart"
 469         rm -f "$lockf"
 470         exit 0
 471 fi
 472
 473 bumptime=
 474 if [ -n "$isfork" ] && [ -z "$lastparentgcsecs" ]; then
 475         # set lastparentgc and then update gcstart to be at least 1 second later
 476         config_set lastparentgc "$gcstart"
 477         bumptime=1
 478 fi
 479 if [ -z "$lastreceivesecs" ]; then
 480         # set lastreceive and then update gcstart to be at least 1 second later
 481         config_set lastreceive "$gcstart"
 482         bumptime=1
 483 fi
 484 if [ -n "$bumptime" ]; then
 485         sleep 1
 486         gcstart="$(date "$datefmt")"
 487 fi
 488
 489 progress "+ [$proj] garbage check (`date`)"
 490
 491 newdeltas=
 492 [ -z "$alwaysredelta" ] || newdeltas=-f
 493 if [ -z "$newdeltas" ] && [ -f gfi-packs -a -s gfi-packs ] && is_gfi_mirror; then
 494         if [ $(list_packs --exclude-no-idx --count objects/pack) -le \
 495              $(list_packs --exclude-no-idx --count --quiet --only gfi-packs) ]; then
 496                 # Don't bother with repack_gfi_packs since everything's being repacked
 497                 newdeltas=-f
 498         fi
 499 fi
 500 if [ -z "$newdeltas" ] && [ -n "$noreusedeltaopt" ] && \
 501    [ $(list_packs --exclude-no-idx --count-objects objects/pack) -le $var_redelta_threshold ]; then
 502         # There aren't enough objects to worry about so just redelta to get the best pack
 503         newdeltas=-f
 504 fi
 505 if [ -z "$newdeltas" ] || has_forks "$proj"; then
 506         # Since we're not going to recompute deltas overall, we need to do the "mini"
 507         # maintenance and by doing it before we copy objects down to forks we reduce
 508         # the amount that gets sprayed into the forks' objects directories.
 509         # If we have forks we always need to do the "mini" maintenance, even if we are
 510         # recomputing all deltas, in order to avoid having suboptimal packs in the forks.
 511         make_svn_pack
 512         repack_gfi_packs
 513         combine_small_packs 1
 514 fi
 515
 516 # safe pruning: we put all our objects in all forks, then we can
 517 # safely get rid of extra ones; repacks in forks will get rid of
 518 # the redundant ones again then; we carefully grab only loose
 519 # objects and pack .idx and .pack files
 520 forkdir="$proj"
 521 if [ -d "../${forkdir##*/}" ]; then
 522         # It is enough to copy objects just one level down and get_repo_list
 523         # takes a regular expression (which is automatically prefixed with '^')
 524         # so we can easily match forks exactly one level down from this project
 525         get_repo_list "$forkdir/[^/]*:" |
 526                 while read fork; do
 527                         # Ignore forks that do not exist or are symbolic links
 528                         [ ! -L "$cfg_reporoot/$fork.git" -a -d "$cfg_reporoot/$fork.git" ] || \
 529                                 continue
 530                         # Or do not have a non-zero length alternates file
 531                         [ -s "$cfg_reporoot/$fork.git/objects/info/alternates" ] || \
 532                                 continue
 533                         # Match objects in parent project
 534                         for d in objects/??; do
 535                                 [ "$d" != "objects/??" ] || continue
 536                                 mkdir -p "$cfg_reporoot/$fork.git/$d"
 537                                 ln -f "$d"/* "$cfg_reporoot/$fork.git/$d" || :
 538                         done
 539                         # Match packs in parent project
 540                         mkdir -p "$cfg_reporoot/$fork.git/objects/pack"
 541                         if [ "$(echo objects/pack/pack-*.idx)" != \
 542                                     "objects/pack/pack-*.idx" ]; then
 543                                 ln -f objects/pack/pack-*.pack "$cfg_reporoot/$fork.git/objects/pack" || :
 544                                 ln -f objects/pack/pack-*.idx "$cfg_reporoot/$fork.git/objects/pack" || :
 545                                 if ! [ -e "$cfg_reporoot/$fork.git/.needsgc" ]; then
 546                                         # Trigger a mini gc in the fork if it now has too many packs
 547                                         packs="$(list_packs --quiet --count --exclude-no-idx "$cfg_reporoot/$fork.git/objects/pack" || :)"
 548                                         if [ -n "$packs" ] && [ "$packs" -ge 20 ]; then
 549                                                 >"$cfg_reporoot/$fork.git/.needsgc"
 550                                         fi
 551                                 fi
 552                                 git --git-dir="$cfg_reporoot/$fork.git" update-server-info
 553                         fi
 554                         # Update the fork's lastparentgc date (must be current, not $gcstart)
 555                         git --git-dir="$cfg_reporoot/$fork.git" config \
 556                                 gitweb.lastparentgc "$(date "$datefmt")"
 557                 done
 558 fi
 559
 560 git pack-refs --all
 561 touch .gc_in_progress
 562 rm -f .gc_failed bundles/*
 563 rm -f objects/pack/pack-*.bndl
 564 # We use the -A option with git repack so that unreachable objects can live
 565 # on for a time as loose objects.  This is particularly helpful if we just
 566 # happen to be in the process of sending out a ref update for a ref that was
 567 # force updated and the old ref value would have otherwise been removed by
 568 # repack because it was now unreachable.  Admittedly the window for gc to run
 569 # and do that before we manage to send out the ref update is not large, but
 570 # it would not be difficult to create such a situation.  Unfortunately, when
 571 # Git unpacks these unreachable objects it will give them the modification
 572 # time of the *.pack file they came out of.  This could be very, very old.
 573 # If that happens, the subsequent git prune --expire some_time_ago will still
 574 # remove the object(s) and our pending ref update will still lose out.
 575 # To prevent this from happening and to get the behavior we want, we now
 576 # touch the modification time of all pack-<sha>.pack files so that any
 577 # loosened objects get a current time.  Git does not provide any other
 578 # mechanism to do this.  We do not want to just touch all loose objects
 579 # left after the repack because that would cause objects that were loosened
 580 # previously to live on which we definitely do not want.
 581 list_packs --exclude-no-idx objects/pack | xargs touch -c 2>/dev/null || :
 582 # We wish to keep deltas from our last full pack so if we're not redeltaing
 583 # then make sure the .pack associated with the .bitmap has a newer mod time
 584 # (If there is no .bitmap then touch the pack with the most objects instead.)
 585 if [ -z "$newdeltas" ]; then
 586         bmpack="$(list_packs --exclude-no-bitmap --exclude-no-idx --max-matches 1 objects/pack)"
 587         [ -n "$bmpack" ] || bmpack="$(list_packs --exclude-no-idx --max-matches 1 --object-limit -1 --include-boundary objects/pack)"
 588         if [ -n "$bmpack" ] && [ -f "$bmpack" -a -s "$bmpack" ]; then
 589                 sleep 1
 590                 touch -c "$bmpack" 2>/dev/null || :
 591         fi
 592 fi
 593 # The git repack command may issue a 'disabling bitmap' warning for some
 594 # repositories.  This is perfectly normal and should be suppressed unless
 595 # show_progress is set.  Unfortunately that means we have to grep -v the
 596 # output.  And furthermore, since it's a translated message, we have to
 597 # force the language to english to be sure we do it.
 598 repackcmd="git repack $packopts -A -d -l $quiet $newdeltas $@"
 599 [ -n "$show_progress" ] || \
 600 repackcmd="{ LC_ALL=C $repackcmd 2>&1 || touch .gc_failed; } | LC_ALL=C grep -v 'disabling bitmap' || :"
 601 eval "$repackcmd"
 602 [ ! -e .gc_failed ] || exit 1
 603 # These, if they exist, are now meaningless and need to be removed
 604 rm -f gfi-packs .needsgc .svnpack .svnpackgc
 605 allpacks="$(echo objects/pack/pack-$octet20.pack)"
 606 curhead="$(cat HEAD)"
 607 pkrf=
 608 [ ! -e packed-refs ] || pkrf=packed-refs
 609 eval "reposizek=$(( $(echo 0 $(du -k $pkrf $allpacks 2>/dev/null | LC_ALL=C awk '{print $1}') | \
 610         LC_ALL=C sed -e 's/ / + /g') ))"
 611 # The -A option to `git repack` may have caused some loose objects to pop
 612 # out of their packs.  We must make these objects group writable so that they
 613 # can be freshened by other pushers.  Technically we need only do this for
 614 # push projects but to enable mirror projects to be more easily converted to
 615 # push projects, we go ahead and do it for all projects.
 616 { find objects/$octet -type f -name "$octet19" -print0 | xargs -0 chmod ug+w || :; } 2>/dev/null
 617 # The git prune command does not take a -q or --quiet but started outputting
 618 # 'Checking connectivity' progress messages in v1.7.9.  However, we can
 619 # suppress those by piping through cat as it only activates the progress
 620 # messages when stderr is a tty.  We only expire loose objects older than one
 621 # day just in case there's some pending action (such as sending out a ref
 622 # update) in progress that might want to examine them.  This may leave us with
 623 # loose objects.  That's okay because at the next gc interval, we will always
 624 # run gc if we see any loose objects regardless of whether or not we've seen
 625 # any updates or we've received new linked objects from our parent.  Note that
 626 # in order to keep loose objects that just recently became unreferenced but
 627 # have a very old modification date around we rely on some help from both the
 628 # update.sh and hooks/pre-receive scripts.  Furthermore, since Git v2.2.0
 629 # (d3038d22 prune: keep objects reachable from recent objects) an unreachable
 630 # object that would otherwise be pruned (because it's too old) will be kept
 631 # alive by an unreachable object that refers to it that's not old enough to
 632 # be pruned yet.
 633 prunecmd='git prune --expire 1_day_ago'
 634 [ -n "$show_progress" ] || \
 635 prunecmd="{ $prunecmd 2>&1 || touch .gc_failed; } | cat"
 636 eval "$prunecmd"
 637 [ ! -e .gc_failed ] || exit 1
 638 git update-server-info
 639
 640 # darcs:// mirrors have a xxx.log file that will grow endlessly
 641 # if this is a mirror and the file exists, shorten it to 10000 lines
 642 # also take this opportunity to optimize the darcs repo
 643 if [ ! -e .nofetch ] && [ -n "$cfg_mirror" ]; then
 644         url="$(config_get baseurl || :)"
 645         case "$url" in darcs://*)
 646                 if [ -n "$cfg_mirror_darcs" ]; then
 647                         url="${url%/}"
 648                         basedarcs="$(basename "${url#darcs:/}")"
 649                         if [ -f "$basedarcs.log" ]; then
 650                                 tail -n 10000 "$basedarcs.log" > "$basedarcs.log.$$"
 651                                 mv -f "$basedarcs.log.$$" "$basedarcs.log"
 652                         fi
 653                         if [ -d "$basedarcs.darcs" ]; then
 654                                 (
 655                                         cd "$basedarcs.darcs"
 656                                         # without show_progress suppress non-error output
 657                                         [ -n "$show_progress" ] || exec >/dev/null
 658                                         # Note that this does not optimize _darcs/inventories/ :(
 659                                         darcs optimize || :
 660                                 )
 661                         fi
 662                 fi
 663         esac
 664 fi
 665
 666 # Create a matching .bndl header file for the all-in-one pack we just created
 667 # but only if we're not a fork (otherwise the bundle would not be complete)
 668 # and we are running at least Git version 1.7.2 (pack_is_complete always fails otherwise)
 669 if [ ! -s objects/info/alternates ] && [ -n "$var_have_git_172" ]; then
 670         # There should only be one pack in $allpacks but if there was a
 671         # simultaneous push...
 672         # The one we just created will have a .idx and will NOT have a .keep
 673         pkfound=
 674         pkhead=
 675         for pk in $allpacks; do
 676                 [ -s "$pk" ] || continue
 677                 pkbase="${pk%.pack}"
 678                 [ -s "$pkbase.idx" ] || continue
 679                 [ ! -e "$pkbase.keep" ] || continue
 680                 if pkhead="$(pack_is_complete "$PWD/$pk" "$PWD/packed-refs" "$curhead")"; then
 681                         pkfound="$pkbase"
 682                         break;
 683                 fi
 684         done
 685         if [ -n "$pkfound" -a -n "$pkhead" ]; then
 686                 {
 687                         echo "# v2 git bundle"
 688                         LC_ALL=C sed -ne "/^$octet20 refs\/[^ $tab]*\$/ p" < packed-refs
 689                         echo "$pkhead HEAD"
 690                         echo ""
 691                 } > "$pkbase.bndl"
 692                 bndletag="$("$cfg_basedir/bin/rangecgi" --etag -m 1 "$pkbase.bndl" "$pkbase.pack" || :)"
 693                 bndlsha="$(printf '%s' "$bndletag" | git hash-object --stdin || :)"
 694                 if [ -n "$bndletag" ]; then
 695                         case "$bndlsha" in $octet20)
 696                                 bndlshatrailer="${bndlsha#????????}"
 697                                 bndlshaprefix="${bndlsha%$bndlshatrailer}"
 698                                 bndlname="$(TZ=UTC date +%Y%m%d_%H%M%S)-${bndlshaprefix:-0}"
 699                                 [ -d bundles ] || mkdir bundles
 700                                 echo "${pkbase#objects/pack/}.bndl" > "bundles/$bndlname"
 701                                 echo "${pkbase#objects/pack/}.pack" >> "bundles/$bndlname"
 702                                 ln -s -f -n "$bndlname" bundles/latest
 703                         esac
 704                 fi
 705         fi
 706 fi
 707
 708 # Record the size of this repo as the sum of its *.pack sizes as 1024-byte blocks
 709 config_set_raw girocco.reposizek "${reposizek:-0}"
 710
 711 # We use $gcstart here to avoid a race where a push occurs during the gc itself
 712 # and the next future gc could be incorrectly skipped if we used the current
 713 # timestamp here instead
 714 config_set lastgc "$gcstart"
 715 rm -f "$lockf"
 716
 717 progress "- [$proj] garbage check (`date`)"