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 [ $# -lt 1 ]; then
  12         echo "Usage: gc.sh projname [extra-repack-args]" >&2
  13         exit 1
  14 fi
  15
  16 # Includes
  17 _shlib_done=1
  18 unset GIROCCO_SUPPRESS_AUTO_GC_UPDATE
  19 . "$cfg_basedir/jobd/maintain-auto-gc-hack.sh"
  20 . "$cfg_basedir/jobd/generate-auto-gc-update.sh"
  21 GIROCCO_SUPPRESS_AUTO_GC_UPDATE=1 && export GIROCCO_SUPPRESS_AUTO_GC_UPDATE
  22
  23 # packing options
  24 packopts="--depth=50 --window=50 --window-memory=${var_window_memory:-1g}"
  25 quiet="-q"; [ "${show_progress:-0}" = "0" ] || quiet=
  26
  27 umask 002
  28 [ "$cfg_permission_control" != "Hooks" ] || umask 000
  29 clean_git_env
  30
  31 vcnt() {
  32         eval "$1="'$(( $# - 1 ))'
  33 }
  34
  35 pidactive() {
  36         if _result="$(kill -0 "$1" 2>&1)"; then
  37                 # process exists and we have permission to signal it
  38                 return 0
  39         fi
  40         case "$_result" in *"not permitted"*)
  41                 # we do not have permission to signal the process
  42                 return 0
  43         esac
  44         # process does not exist
  45         return 1
  46 }
  47
  48 createlock() {
  49         # A .lock file should only exist for much less than a second.
  50         # If we see a stale lock file (> 1h old), remove it and then,
  51         # just in case, wait 30 seconds for any process whose .lock
  52         # we might have just removed (it's racy) to finish doing what
  53         # should take much less than a second to do.
  54         _stalelock="$(find -L "$1.lock" -maxdepth 1 -mmin +60 -print 2>/dev/null)" || :
  55         if [ -n "$_stalelock" ]; then
  56                 rm -f "$_stalelock"
  57                 sleep 30
  58         fi
  59         for _try in p p n; do
  60                 if (set -C; >"$1.lock") 2>/dev/null; then
  61                         echo "$1.lock"
  62                         return 0
  63                 fi
  64                 # delay and try again
  65                 [ "$_try" != "p" ] || sleep 1
  66         done
  67         # cannot create lock file
  68         return 1
  69 }
  70
  71 # The pre-receive script creates one ref log file per push but we want them to
  72 # be coalesced into one ref log file per day.  We are guaranteed that any files
  73 # we find to coalesce are NOT currently being written to since they are always
  74 # written first as temporary files and then moved into place.  We attempt to
  75 # transfer the most recent modification time to the coalesced log file which
  76 # would step on its mod time if it were being written to directly, but if we
  77 # find per-process ref log files then it must be a push project and the only
  78 # thing that would write directly to the main per-day log file would be a
  79 # mirror project so there's actually no conflict.
  80 # Also, if the clock is wonky (or was futzed with) we may have both YYYYMMDD
  81 # and YYYYMMDD.gz present in which case combine them into YYYYMMDD
  82 coalesce_reflogs() {
  83         [ -d reflogs ] || return 0
  84         rm -f .gc_failed
  85         find -L reflogs -maxdepth 1 -type f -name "[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]" -print |
  86         while read -r rname; do
  87                 if [ -e "$rname.gz" ]; then
  88                         if [ -s "$rname" ]; then
  89                                 # Presumably the .gz file must have been created before the non-gz
  90                                 # file since it had to be uncompressed at some point therefore
  91                                 # we need to append the non-gz contents to it but keep the non-gz
  92                                 # contents timestamp so we rename to YYYYMMDD_ which will sort first
  93                                 # and be picked up in the next step if we are interrupted in the middle.
  94                                 # If a YYYYMMDD_ file already exists we append to it and transfer the
  95                                 # timestamp.  Finally we transfer the YYYYMMDD_ timestamp to the result
  96                                 # and remove the YYYYMMDD_ temporary file leaving the result uncompressed.
  97                                 if [ -e "${rname}_" ]; then
  98                                         cat "$rname" >>"${rname}_"
  99                                         touch -r "$rname" "${rname}_"
 100                                         rm -f "$rname"
 101                                         ! [ -e "$rname" ]
 102                                 else
 103                                         mv "$rname" "${rname}_"
 104                                 fi
 105                                 gzip -d "$rname.gz" </dev/null
 106                                 [ -e "$rname" ] && ! [ -e "$rname.gz" ]
 107                                 cat "${rname}_" >>"$rname"
 108                                 touch -r "${rname}_" "$rname"
 109                                 rm -f "${rname}_"
 110                         else
 111                                 # Just remove the empty file to resolve the problem
 112                                 rm -f "$rname"
 113                         fi
 114                 fi
 115         done
 116         find -L reflogs -maxdepth 1 -type f -name "[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]_*" -print | LC_ALL=C sort |
 117         while read -r rname; do
 118                 logname="${rname%%_*}"
 119                 # If someone's been futzing with the date, the file we want to
 120                 # append to could already have been compressed, so we just uncompress
 121                 # it here.  The previous block guarantees we do not have both a compressed
 122                 # and uncompressed version present at the same time.
 123                 if [ -e "$logname.gz" ]; then
 124                         gzip -d "$logname.gz" </dev/null
 125                         [ -e "$logname" ] && ! [ -e "$logname.gz" ]
 126                 fi
 127                 cat "$rname" >>"$logname"
 128                 touch -r "$rname" "$logname"
 129                 rm -f "$rname"
 130                 if [ -e "$rname" ]; then
 131                         >.gc_failed
 132                         echo "! [$proj] failed to remove $rname" >&2
 133                         exit 1 # will only exit subshell created by "|"
 134                 fi
 135         done
 136         ! [ -e .gc_failed ]
 137 }
 138
 139 # Remove any files in reflogs that are older than $cfg_reflogs_lifetime days
 140 prune_reflogs() {
 141         [ -d reflogs ] || return 0
 142         exp="$(( ${cfg_reflogs_lifetime:-1} * 1440 ))"
 143         [ $exp -gt 0 ] || exp=1440
 144         [ $exp -le 43200 ] || exp=43200
 145         find -L reflogs -maxdepth 1 -type f -mmin "+$exp" -exec rm -f '{}' + || :
 146 }
 147
 148 # Compact any reflogs that are not today's UTC date unless a .gz version exists
 149 compact_reflogs() {
 150         [ -d reflogs ] || return 0
 151         _td="reflogs/$(TZ=UTC date '+%Y%m%d')"
 152         find -L reflogs -maxdepth 1 -type f -name "[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]" -print |
 153         while read -r rname; do
 154                 [ "$rname" != "$_td" ] || continue
 155                 ! [ -e "$rname.gz" ] || continue
 156                 gzip -9 "$rname" </dev/null
 157         done
 158 }
 159
 160 # return true if there's more than one objects/pack-<sha>.pack file or
 161 # ANY sha-1 files in objects
 162 is_dirty() {
 163         _packs=$(find -L objects/pack -name "pre-auto-gc-[12].pack" -prune -o -name "*.pack" -type f -print | head -n 2 | LC_ALL=C wc -l)
 164         if [ $_packs != 1 ] && [ $_packs != 0 ]; then
 165                 return 0
 166         fi
 167         _objs=$(find -L objects/$octet -name "$octet19*" -type f -print 2>/dev/null | head -n 1 | LC_ALL=C wc -l)
 168         [ $_objs -ne 0 ]
 169 }
 170
 171 # make sure combine-packs uses the correct Git executable
 172 run_combine_packs() {
 173         PATH="$var_git_exec_path:$cfg_basedir/bin:$PATH" @basedir@/jobd/combine-packs.sh "$@"
 174 }
 175
 176 # duplicate the first file to the name given by the second file making sure that
 177 # the second file appears atomically all-at-once after the copy has been completed
 178 # and does not appear at all if the copy fails (in which case this function fails)
 179 # if the second file already exists this function fails with status 1
 180 # if the file names are the same this function returns immediately with success
 181 dupe_file() {
 182         [ "$1" != "$2" ] || return 0
 183         ! [ -e "$2" ] || return 1
 184         case "$2" in
 185                 *?/?*)  _tmpdir="${2%/*}";;
 186                 *)      _tmpdir=".";;
 187         esac
 188         _tmpfile="$(mktemp "${_tmpdir:-.}/packtmp-XXXXXX")" || return 1
 189         cp -fp "$1" "$_tmpfile" || return 1
 190         mv -f "$_tmpfile" "$2"
 191 }
 192
 193 # rename_pack oldnamepath newnamepath
 194 # note that .keep and .bndl files are left untouched and not moved at all!
 195 rename_pack() {
 196         [ $# -eq 2 ] && [ "$1" != "$2" ] || {
 197                 echo >&2 "[$proj] incorrect use of rename_pack function"
 198                 exit 1
 199         }
 200         # Git assumes that if the destination of the rename already exists
 201         # that it is, in fact, a copy of the same bytes so silently succeeds
 202         # without doing anything.  We duplicate that logic here.
 203         # Git checks for the .idx file first before even trying to use a pack
 204         # so it should be the last moved and the first removed.
 205         for ext in pack bitmap idx; do
 206                 [ -f "$1.$ext" ] || continue
 207                 ln "$1.$ext" "$2.$ext" >/dev/null 2>&1 ||
 208                 dupe_file "$1.$ext" "$2.$ext" >/dev/null 2>&1 ||
 209                 [ -f "$2.$ext" ] || {
 210                         echo >&2 "[$proj] unable to move $1.$ext to $2.$ext"
 211                         exit 1
 212                 }
 213         done
 214         for ext in idx pack bitmap; do
 215                 rm -f "$1.$ext"
 216         done
 217         return 0
 218 }
 219
 220 # combine the input pack(s) into a new pack (or possibly packs if packSizeLimit set)
 221 # input pack names are read from standard input one per line delimited by the first
 222 # ':', ' ' or '\n' character on the line (which allows gfi-packs to be read directly)
 223 # all arguments, if any, are passed to pack-objects as additional options
 224 # returns non-zero on failure AND creates .gc_failed in that case
 225 combine_packs() {
 226         rm -f .gc_failed
 227         find -L objects/pack -maxdepth 1 -type f -name '*.zap*' -exec rm -f '{}' + || :
 228         run_combine_packs --replace "$@" $packopts --all-progress-implied $quiet --non-empty || {
 229                 >.gc_failed
 230                 return 1
 231         }
 232         return 0
 233 }
 234
 235 # if the current directory is_gfi_mirror then repack all packs listed in gfi-packs
 236 repack_gfi_packs() {
 237         [ -n "$gfi_mirror" ] || return 0
 238         [ -d objects/pack ] || { rm -f gfi-packs; return 0; }
 239         progress "~ [$proj] redeltifying poor quality git fast-import packs"
 240         combine_packs --ignore-missing --no-reuse-delta <gfi-packs
 241         rm -f gfi-packs
 242         return 0
 243 }
 244
 245 # see if there are "lotsa" loose objects
 246 # "lotsa" is defined as the 17, 68, 71 and 86 object directories existing
 247 # and there being at least 5 total objects between them which corresponds
 248 # to an approximate average of 320 loose objects before this function starts
 249 # returning true and triggering a "mini" gc to pack up loose objects
 250 lotsa_loose_objects() {
 251         [ -d objects/17 ] && [ -d objects/68 ] && [ -d objects/71 ] && [ -d objects/86 ] || return 1
 252         _objs=$(( $(find -L objects/17 objects/68 objects/71 objects/86 -maxdepth 1 -name "$octet19*" -type f -print 2>/dev/null | LC_ALL=C wc -l) ))
 253         [ ${_objs:-0} -ge 5 ]
 254 }
 255
 256 # pack any existing loose objects into a new _l.pack file then run prune-packed
 257 # note that prune-packed is NOT run beforehand -- the caller must do that if needed
 258 # loose objects need not be part of complete commits/trees as --weak-naming is used
 259 pack_loose_objects() {
 260         _lpacks="$(run_combine_packs </dev/null --names --loose --weak-naming --non-empty --all-progress-implied ${quiet:---progress} $packopts)"
 261         if [ -n "$_lpacks" ]; then
 262                 # We need to identify these packs later so we don't combine_packs them
 263                 for _objpack in $_lpacks; do
 264                         rename_pack "objects/pack/pack-$_objpack" "objects/pack/pack-${_objpack}_l" || :
 265                 done
 266                 git prune-packed $quiet
 267         fi
 268 }
 269
 270 # combine small packs into larger pack(s)
 271 # we avoid any _[lo], keep, bndl or bitmap packs
 272 # if the optional argument is non-empty even a single small pack will be redeltad
 273 combine_small_packs() {
 274         _didprogress=
 275         _minsmallpacks=2
 276         if [ -n "$1" ] && [ -n "$noreusedeltaopt" ]; then
 277                 _minsmallpacks=1
 278         fi
 279         _lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl"
 280         _lpo="$_lpo --exclude-sfx _u --exclude-sfx _o --exclude-sfx _l"
 281         _lpo="$_lpo --quiet --object-limit $var_redelta_threshold objects/pack"
 282         while
 283                 _cnt="$(list_packs --count $_lpo)" || :
 284                 test "${_cnt:-0}" -ge $_minsmallpacks
 285         do
 286                 [ -n "$_didprogress" ] || {
 287                         progress "~ [$proj] combining small packs into a single larger pack"
 288                         _didprogress=1
 289                 }
 290                 _newp="$(list_packs $_lpo | combine_packs --names $noreusedeltaopt)"
 291                 vcnt _newc $_newp
 292                 # be paranoid and exit the loop if we haven't reduced the number of packs
 293                 [ $_newc -lt $_cnt ] || break
 294                 _minsmallpacks=2
 295         done
 296         return 0
 297 }
 298
 299 # combine small _l packs into larger pack(s) using --weak-naming
 300 # we avoid any non _l, keep, bndl or bitmap packs
 301 # if the optional 2nd argument is non-empty even a single small pack will be redeltad
 302 combine_small_loose_packs() {
 303         _didprogress=
 304         _minsmallpacks=2
 305         if [ -n "$1" ] && [ -n "$noreusedeltaopt" ]; then
 306                 _minsmallpacks=1
 307         fi
 308         _lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl"
 309         _lpo="$_lpo --exclude-no-sfx _l"
 310         _lpo="$_lpo --quiet --object-limit $var_redelta_threshold objects/pack"
 311         while
 312                 _cnt="$(list_packs --count $_lpo)" || :
 313                 test "${_cnt:-0}" -ge $_minsmallpacks
 314         do
 315                 [ -n "$_didprogress" ] || {
 316                         progress "~ [$proj] combining small loose packs into a single larger pack"
 317                         _didprogress=1
 318                 }
 319                 _newp="$(list_packs $_lpo | combine_packs --names --weak-naming $noreusedeltaopt)"
 320                 # We need to identify these packs later so we don't combine_packs them
 321                 for _objpack in $_newp; do
 322                         rename_pack "objects/pack/pack-$_objpack" "objects/pack/pack-${_objpack}_l" || :
 323                 done
 324                 vcnt _newc $_newp
 325                 # be paranoid and exit the loop if we haven't reduced the number of packs
 326                 [ $_newc -lt $_cnt ] || break
 327                 _minsmallpacks=2
 328         done
 329         return 0
 330 }
 331
 332 # Unfortunately git-svn lacks the ability to store newly fetched revisions as a pack.
 333 # However, the fetch code conveniently sets .svnpack just before it runs git-svn fetch
 334 # so that it's easy to find all the objects that have been fetched by git-svn and
 335 # combine them into a pack.  The --no-reuse-delta option is meaningless here since
 336 # everything to be packed is a loose object and therefore not a delta so deltification
 337 # will always take place.
 338 make_svn_pack() {
 339         [ -f .svnpack ] && [ -n "$svn_mirror" ] || return 0
 340         rm -f .svnpackgc
 341         mv -f .svnpack .svnpackgc
 342         progress "~ [$proj] combining loose git-svn objects into a pack"
 343         _newp="$(find -L objects/$octet -maxdepth 1 -type f -newer .svnpackgc -name "$octet19*" -print 2>/dev/null |
 344         LC_ALL=C awk -F / '{print $2 $3}' |
 345         run_combine_packs --objects --names $packopts --incremental --all-progress-implied $quiet --non-empty)" || {
 346                 mv -f .svnpackgc .svnpack
 347                 >.gc_failed
 348                 return 1
 349         }
 350         if [ -n "$_newp" ]; then
 351                 # remove the now-redundant loose objects -- this is always safe
 352                 # even during a concurrent push because a reprepare_packed_git
 353                 # will be triggered if an object that should be there is not
 354                 # found thereby finding it in the new pack instead
 355                 git prune-packed $quiet
 356         fi
 357         rm -f .svnpackgc
 358 }
 359
 360 # HEADSHA="$(pack_is_complete /full/path/to/some.pack /full/path/to/packed-refs "$(cat HEAD)")"
 361 pack_is_complete() {
 362         # Must have a matching .idx file and a non-empty packed-refs file
 363         [ -s "${1%.pack}.idx" ] || return 1
 364         [ -s "$2" ] || return 1
 365         _headsha=
 366         case "$3" in
 367                 $octet20*)
 368                         _headsha="$3"
 369                         ;;
 370                 "ref: refs/"?*|"ref:refs/"?*|"refs/"?*)
 371                         _headmatch="${3#ref:}"
 372                         _headmatch="${_headmatch# }"
 373                         _headmatchpat="$(echo "$_headmatch" | LC_ALL=C sed -e 's/\([.$]\)/\\\1/g')"
 374                         _headsha="$(LC_ALL=C grep -e "^$octet20$hexdig* $_headmatchpat\$" <"$2" |
 375                                     LC_ALL=C cut -d ' ' -f 1)"
 376                         case "$_headsha" in $octet20*) :;; *)
 377                                 return 1
 378                         esac
 379                         ;;
 380                 *)
 381                         # bad HEAD
 382                         return 1
 383         esac
 384         rm -rf pack_is_complete_test
 385         mkdir pack_is_complete_test
 386         mkdir pack_is_complete_test/refs
 387         mkdir pack_is_complete_test/objects
 388         mkdir pack_is_complete_test/objects/pack
 389         echo "$_headsha" >pack_is_complete_test/HEAD
 390         ln -s "$1" pack_is_complete_test/objects/pack/
 391         ln -s "${1%.pack}.idx" pack_is_complete_test/objects/pack/
 392         ln -s "$2" pack_is_complete_test/packed-refs
 393         _count="$(git --git-dir=pack_is_complete_test rev-list --count --all 2>/dev/null)" || :
 394         rm -rf pack_is_complete_test
 395         [ -n "$_count" ] || return 1
 396         [ "$_count" -gt 0 ] 2>/dev/null || return 1
 397         echo "$_headsha"
 398 }
 399
 400 # On return a "$lockf" will have been created that must be removed when gc is done
 401 lock_gc() {
 402         # be compatibile with gc.pid file from newer Git releases
 403         lockf=gc.pid
 404         hn="$(hostname)"
 405         active=
 406         if [ "$(createlock "$lockf")" ]; then
 407                 # If $lockf is:
 408                 #   1) less than 12 hours old
 409                 #   2) contains two fields (pid hostname) NO trailing NL
 410                 #   3) the hostname is different OR the pid is still alive
 411                 # then we exit as another active process is holding the lock
 412                 if [ "$(find -L "$lockf" -maxdepth 1 -mmin -720 -print 2>/dev/null)" ]; then
 413                         apid=
 414                         ahost=
 415                         read -r apid ahost ajunk <"$lockf" || :
 416                         if [ "$apid" ] && [ "$ahost" ]; then
 417                                 if [ "$ahost" != "$hn" ] || pidactive "$apid"; then
 418                                         active=1
 419                                 fi
 420                         fi
 421                 fi
 422         else
 423                 echo >&2 "[$proj] unable to create gc.pid.lock file"
 424                 exit 1
 425         fi
 426         if [ -n "$active" ]; then
 427                 rm -f "$lockf.lock"
 428                 echo >&2 "[$proj] gc already running on machine '$ahost' pid '$apid'"
 429                 exit 1
 430         fi
 431         printf "%s %s" "$$" "$hn" >"$lockf.lock"
 432         chmod 0664 "$lockf.lock"
 433         mv -f "$lockf.lock" "$lockf"
 434 }
 435
 436 # Create a repack subdirectory such that running repack in it will pack the
 437 # same things that a pack in the normal directory would except that the pack
 438 # is guaranteed to be generated in an optimized order by adding a suitable
 439 # synthesized ref in the refs/tags namespace (yes, pack-objects.c really does
 440 # behave differently depending on the contents of the refs/tags namespace).
 441 # Before calling this, pack-refs --all MUST be performed or the wrong pack
 442 # will end up being made.
 443 #
 444 # If a ref deletion is pushed after making the repack subdir but before the
 445 # the actual repack, the discarded objects will be packed -- no big deal,
 446 # they'll get discarded the next time gc runs.
 447 #
 448 # If a fast-forward ref update is pushed after making the repack subdir but
 449 # before the actual repack, it will be picked up and the new objects packed
 450 # (subject to the normal git repack race about picking such updates up).
 451 #
 452 # If a non-fast-forward ref update is pushed after making the repack subdir but
 453 # before the actual repack, it will be picked up like a fast-forward update but
 454 # the discarded objects will be included like a ref deletion (until the next
 455 # scheduled gc takes place).
 456 #
 457 # We retain a copy of the original packed-refs file as repack/packed-refs.orig
 458 # If ref deletions come in while we're repacking, the original packed-refs
 459 # file will be modified, but we'll still pack the deleted ref(s).
 460 # If the packed-refs.orig file is used to create the bundle header we avoid
 461 # a situation where the bundle contains a ref state that never actually
 462 # existed in reality (for example a new branch is pushed and then an old
 463 # branch deleted afterwards -- the deletion would show up in the bundle
 464 # because it will cause the original packed-refs file to be re-written, but
 465 # the new branch creation will not unless we do another pack-refs which might
 466 # lead to having in incomplete bundle).  Therefore we want to keep a copy of
 467 # the original packed-refs file around.  We do the same thing for HEAD.
 468 #
 469 # It's possible that the "objects" subdirectory is a symbolic link.
 470 # Git does support this.  However, during the repacking process, new packs
 471 # will be created in repack/alt/pack and then moved into objects/pack.
 472 # In order for this to work seemlessly, they must both be on the same
 473 # filesystem.  But when objects (or even objects/pack) is a symbolic link they
 474 # might not be.  For this reason a "repack" subdirectory is created under
 475 # objects/pack and the repack/alt/pack directory symbolicly linked to it.
 476 #
 477 # Git allows not just HEAD to be a symbolic-ref, but any ref anywhere in the
 478 # refs namespace.  We are concerned about ref name collisions and getting the
 479 # right tag set to get an optimal pack.  We can safely duplicate the ref space
 480 # under refs/heads, refs/notes and refs/remotes without any risk of unwanted
 481 # collisions and this will likely make over 99%+ of all symbolic refs found
 482 # in the wild work properly.  Girocco itself never creates any symbolic refs
 483 # inside the refs namespace; this is a nod to simultaneously using a Girocco
 484 # repository for other purposes.
 485 make_repack_dir() {
 486         ! [ -d repack ] || rm -rf repack
 487         ! [ -d repack ] || { echo >&2 "[$proj] cannot remove repack subdirectory"; exit 1; }
 488         [ -d objects/pack ] || mkdir -p objects/pack
 489         ! [ -d objects/pack/repack ] || rm -rf objects/pack/repack
 490         ! [ -d objects/pack/repack ] || { echo >&2 "[$proj] cannot remove objects/pack/repack subdirectory"; exit 1; }
 491         mkdir repack repack/refs repack/alt objects/pack/repack
 492         [ -d info ] || mkdir info
 493         ln -s ../config  repack/config
 494         ln -s ../info    repack/info
 495         ln -s ../objects repack/objects
 496         ln -s "$PWD/objects/pack/repack" repack/alt/pack
 497         ln -s ../../refs repack/refs/refs
 498         _lines=$(( $(LC_ALL=C wc -l <packed-refs) ))
 499         cat HEAD >repack/HEAD.orig
 500         >repack/packed-refs.extra
 501         _xtralines=0
 502         cat packed-refs >repack/packed-refs.orig
 503         if [ $(LC_ALL=C wc -l <repack/packed-refs.orig) -ne "$_lines" ]; then
 504                 echo >&2 "[$proj] error: make_repack_dir failed original packed-refs line count sanity check"
 505                 exit 1
 506         fi
 507         if [ "${cfg_fetch_stash_refs:-0}" = "0" ]; then
 508                 # migrate any refs/stash or refs/tgstash lines to repack/packed-refs.extra
 509                 <repack/packed-refs.orig awk -v xtra="repack/packed-refs.extra" '
 510                         BEGIN { peeling = 0 }
 511                         NR == 1 && /^#/ { print; next; }
 512                         peeling && /^\^/ { print >>xtra; next; }
 513                         /^[0-9A-Fa-f][0-9A-Fa-f][0-9A-Fa-f][0-9A-Fa-f]+ refs\/(stash|tgstash)(\/|$)/ {
 514                                 peeling = 1
 515                                 print >>xtra
 516                                 next
 517                         }
 518                         { peeling = 0; print; }
 519                 ' >repack/packed-refs.new
 520                 _xtralines="$(( $(LC_ALL=C wc -l <repack/packed-refs.extra) + 0 ))"
 521                 _newlines="$(( $(LC_ALL=C wc -l <repack/packed-refs.new) + 0 ))"
 522                 if [ "$(( $_newlines + $_xtralines ))" -ne "$_lines" ]; then
 523                         echo >&2 "[$proj] error: make_repack_dir failed packed-refs.extra line count sanity check"
 524                         exit 1
 525                 fi
 526                 mv -f repack/packed-refs.new repack/packed-refs.orig
 527                 _lines="$_newlines"
 528         fi
 529         # Note: Git v1.5.0 introduced the "# pack-refs with:" header line for the packed-refs file
 530         sed '/^# pack-refs/d; s, refs/, refs/!/,' <repack/packed-refs.orig >repack/packed-refs
 531         nohead=
 532         headref="$(git rev-parse --verify --quiet HEAD)" || :
 533         if [ -n "$headref" ]; then
 534                 echo "$headref refs/!=/HEAD" >>repack/packed-refs
 535                 echo "$headref refs/heads/!" >>repack/packed-refs
 536                 nohead='\, refs/heads/!$,d; '
 537                 _lines=$(( $_lines + 2 ))
 538         fi
 539         if [ $(( $(LC_ALL=C wc -l <repack/packed-refs) + 1 )) -ne "$_lines" ]; then
 540                 echo >&2 "[$proj] error: make_repack_dir failed packed-refs initial line count sanity check"
 541                 exit 1
 542         fi
 543         sed -n "$nohead"'\, refs/heads/,p; \, refs/notes/,p; \, refs/remotes/,p' <repack/packed-refs.orig >>repack/packed-refs
 544         _newlines="$(( $(LC_ALL=C wc -l <repack/packed-refs) ))"
 545         if [ $(( $_newlines + 1 )) -lt "$_lines" ]; then
 546                 echo >&2 "[$proj] error: make_repack_dir failed packed-refs extra line count sanity check"
 547                 exit 1
 548         fi
 549         _lines="$_newlines"
 550         optref="$(git rev-list -n 1 --all 2>/dev/null)" || :
 551         if [ -n "$optref" ]; then
 552                 echo "$optref refs/tags/!" >>repack/packed-refs
 553                 _lines=$(( $_lines + 1 ))
 554                 echo "$optref" >repack/HEAD
 555         else
 556                 cat HEAD >repack/HEAD
 557         fi
 558         if [ $(LC_ALL=C wc -l <repack/packed-refs) -ne "$_lines" ]; then
 559                 echo >&2 "[$proj] error: make_repack_dir failed packed-refs line count sanity check"
 560                 exit 1
 561         fi
 562 }
 563
 564 # Remove any crud that's been left behind by interrupted operations
 565 # that did not clean up after themselves
 566 remove_crud() {
 567         # Remove any existing FETCH_HEAD
 568         # There can only be a FETCH_HEAD if we've been fetching, not if we've been
 569         # receiving pushes (those never create a FETCH_HEAD).
 570         # And if we're fetching because we're a mirror, we know we're not fetching right
 571         # now since jobd.pl never runs a project's fetch simultaneously with its gc.
 572         # Therefore any existing FETCH_HEAD is junk.  And it may be many megabytes if
 573         # there were a lot of refs.
 574         rm -f FETCH_HEAD
 575
 576         # remove any existing pack_is_complete_test or repack subdirectories
 577         # If either exists when this function is called it's crud
 578         rm -rf pack_is_complete_test repack objects/pack/repack
 579
 580         # Remove any stale pack remnants that are more than an hour old.
 581         # Stale pack fragments are defined as any pack-<sha1>.ext where .ext is NOT
 582         # .pack AND the corresponding .pack DOES NOT exist.  A bunch of stale
 583         # pack-<sha1>.idx files without their corresponding .pack files are worthless
 584         # and just waste space.  Normally there shouldn't be any remnants but actually
 585         # this can happen when things are interrupted at just the wrong time.
 586         # Note that the objects/pack directory is created by git init and should
 587         # always exist.
 588         find -L objects/pack -maxdepth 1 -type f -mmin +60 -name "pack-$octet20*.?*" -print |
 589         LC_ALL=C sed -e 's/^objects\/pack\/pack-//; s/\..*$//' | LC_ALL=C sort -u |
 590         while read packsha; do
 591                 ! [ -e "objects/pack/pack-$packsha.pack" ] || continue
 592                 rm -f "objects/pack/pack-$packsha".?*
 593         done
 594
 595         # Remove any stale tmp reflogs files that are more than one hour old.
 596         # Since they are created only while the pre-receive hook is running and
 597         # all it does is process a bunch of refs passed to it on standard input
 598         # it's inconceivable that it would ever take as much as an hour to run.
 599         if [ -d reflogs ]; then
 600                 find -L reflogs -maxdepth 1 -type f -mmin +60 -name "tmp_*" -exec rm -f '{}' + || :
 601         fi
 602
 603         # Remove any stale object tmp_obj_* files that are more than 3 hours old.
 604         # Really these files should only exist very briefly so there shouldn't be any
 605         # but things happen that can end up leaving them behind.
 606         find -L objects/$octet -maxdepth 1 -type f -mmin +180 -name "tmp_obj_?*" -exec rm -f '{}' + 2>/dev/null || :
 607
 608         # Remove any stale pack .keep files that are more than 12 hours old.
 609         # We don't do anything to create any permanent pack .keep files, so they must
 610         # be remnants from some failed push or something.  Removing the .keep will
 611         # allow the pack to be properly repacked.
 612         find -L objects/pack -maxdepth 1 -type f -mmin +720 -name "pack-$octet20*.keep" -exec rm -f '{}' + || :
 613
 614         # Remove any stale tmp_pack_*, tmp_idx_*, tmp_bitmap_*, packtmp-* or .tmp-*-pack* files
 615         # that are more than 12 hours old.
 616         find -L objects/pack -maxdepth 1 -type f -mmin +720 \( \
 617                 -name "tmp_pack_?*" -o -name "tmp_idx_?*" -o -name "tmp_bitmap_?*" -o \
 618                 -name "packtmp-?*" -o -name ".tmp-?*-pack*" \
 619                 \) -exec rm -f '{}' + || :
 620
 621         # Remove any stale incoming-* object quarantine directories that are
 622         # more than 12 hours old.  These are new with Git >= 2.11.0.
 623         find -L objects -maxdepth 1 -type d -name 'incoming-?*' -mmin +720 \
 624                 -exec rm -rf '{}' + || :
 625
 626         # Remove any stale shallow_* files that are more than 12 hours old.
 627         # These can be left behind by Git >= 1.8.4.2 and < 2.0.0 when a client
 628         # requests a shallow clone.  Also discard stale .refs-temp* and
 629         # .refs-new* files at the same time.
 630         find -L . -maxdepth 1 -type f -mmin +720 \( \
 631                 -name "shallow_?*" -o -name ".refs-temp*" -o -name ".refs-new*" \
 632                 \) -exec rm -f '{}' + || :
 633
 634         # Remove any stale *.temp files in the objects area that are more than 12 hours old.
 635         # This can be stale sha1.temp, or stale *.pack.temp so we kill all stale *.temp.
 636         find -L objects -type f -mmin +720 -name "*.temp" -exec rm -f '{}' + || :
 637
 638         # Remove any stale *.lock files in the htmlcache area that might have been left
 639         # behind after an abnormal exit during an attempt to update a cached file and
 640         # are more than 1 hour old.
 641         ! [ -d htmlcache ] || find -L htmlcache -type f -mmin +60 -name "*.lock" -exec rm -f '{}' + || :
 642
 643         # Remove any stale git-svn temp files that are more than 12 hours old.
 644         # The git-svn process creates temp files with random 10 character names
 645         # in the root of $GIT_DIR.  Unfortunately they do not have a recognizable
 646         # prefix, so we just have to kill any files with a 10-character name.  We
 647         # do this only for git-svn mirrors.  All characters are chosen from
 648         # [A-Za-z0-9_] so we can at least check that and fortunately the only
 649         # collision is 'FETCH_HEAD' but that shouldn't matter.
 650         # There may also be temp files with a Git_ prefix as well.
 651         if [ -n "$svn_mirror" ]; then
 652                 _randchar='[A-Za-z0-9_]'
 653                 _randchar2="$_randchar$_randchar"
 654                 _randchar4="$_randchar2$_randchar2"
 655                 _randchar10="$_randchar4$_randchar4$_randchar2"
 656                 find -L . -maxdepth 1 -type f -mmin +720 -name "$_randchar10" -exec rm -f '{}' + || :
 657                 find -L . -maxdepth 1 -type f -mmin +720 -name "Git_*" -exec rm -f '{}' + || :
 658         fi
 659
 660         # Remove any stale fast_import_crash_<pid> files that are more than 3 days old.
 661         if [ -n "$gfi_mirror" ]; then
 662                 find -L . -maxdepth 1 -type f -mmin +4320 -name "fast_import_crash_?*" -exec rm -f '{}' + || :
 663         fi
 664
 665         # Remove any stale core or *.core or core.* files that are more than 3 days old.
 666         find -L . -maxdepth 1 -type f -mmin +4320 \( -name "core" -o -name "*.core" -o -name "core.*" \) \
 667                 -exec rm -f '{}' + || :
 668 }
 669
 670 #
 671 ## Garbage Collection Types
 672 ##
 673 ## There are two kinds of possible garbage collection (gc) operations:
 674 ##
 675 ## 1.  A normal, full gc
 676 ## 2.  A "mini" gc
 677 ##
 678 ## If the full garbage collection interval has expired (or gc has never been
 679 ## run), then a normal, full gc will take place.  Otherwise, a "mini" gc will
 680 ## take place if the file .needsgc exists.
 681 ##
 682 ## A "mini" gc is similar to "git gc --auto" in that it may not end up actually
 683 ## doing anything unless the right conditions are present so it's not a burden
 684 ## to run it often.  If the file .needsgc exists, a "mini" gc will occur at
 685 ## the next opportunity.
 686 ##
 687 ## See the docs/technical/gc.txt and docs/technical/gc-mini.txt files for more
 688 ## of the gory details of how garbage collection is performed.
 689 ##
 690 ## Note, however, that the .nogc file suppresses ALL gc activity (normal or mini).
 691 #
 692
 693 proj="${1%.git}"
 694 shift
 695 cd "$cfg_reporoot/$proj.git"
 696 [ -d objects/pack ] || { rm -f gfi-packs; mkdir -p objects/pack; }
 697 mirror_url="$(get_mirror_url)" || :
 698 svn_mirror=
 699 ! is_svn_mirror_url "$mirror_url" || svn_mirror=1
 700 gfi_mirror=
 701 if [ -f gfi-packs ] && [ -s gfi-packs ] && is_gfi_mirror_url "$mirror_url"; then
 702         gfi_mirror=1
 703 fi
 704
 705 # If git config --bool --get girocco.redelta is explicitly false then automatic
 706 # redelta when there are less than $var_redelta_threshold objects will be suppressed.
 707 # On the other hand, if git config --get girocco.redelta is "always" then, on a full
 708 # gc only, for the final repack, deltas will always be recomputed.
 709 # This can be set on a per-project basis to avoid unusual pathological gc behavior.
 710 # Setting this will hurt efficiency of the affected repository.
 711 # Note that fast-import packs ALWAYS get new deltas regardless of this setting.
 712 noreusedeltaopt="--no-reuse-delta"
 713 [ "$(git config --bool --get girocco.redelta 2>/dev/null || :)" != "false" ] || noreusedeltaopt=
 714 alwaysredelta=
 715 [ "$(git config --get girocco.redelta 2>/dev/null || :)" != "always" ] || alwaysredelta=1
 716
 717 # Extract any -f or -F or --no-reuse-object or --no-reuse-delta options
 718 # to be compatible with the old and new gc.sh versions and avoid ugly argument
 719 # duplication in process lists at the same time
 720 # Any options found will override the "girocco.redelta" setting
 721 recompress=
 722 idx=$#
 723 while [ $idx -gt 0 ]; do
 724         idx=$(( $idx - 1 ))
 725         opt="$1"
 726         shift
 727         case "$opt" in
 728         -f|--no-reuse-delta)
 729                 alwaysredelta=1
 730                 continue
 731                 ;;
 732         -F|--no-reuse-object)
 733                 alwaysredelta=1
 734                 recompress=1
 735                 continue
 736                 ;;
 737         -?*)
 738                 ;;
 739         *)
 740                 printf >&2 '%s\n' "bad non-option argument: $opt"
 741                 echo >&2 "(Did you perhaps intend to use a --xxx=yyy form?)"
 742                 exit 1
 743         esac
 744         [ -z "$opt" ] || set -- "$@" "$opt"
 745 done
 746 if [ -n "$alwaysredelta" ]; then
 747         noreusedeltaopt="--no-reuse-delta"
 748         [ -z "$recompress" ] || noreusedeltaopt="--no-reuse-object"
 749 fi
 750
 751 trap 'e=$?; rm -f .gc_in_progress; if [ $e != 0 ]; then echo "gc failed dir: $PWD" >&2; fi' EXIT
 752 trap 'exit 130' INT
 753 trap 'exit 143' TERM
 754
 755 # date -R is linux-only, POSIX equivalent is '+%a, %d %b %Y %T %z'
 756 datefmt='+%a, %d %b %Y %T %z'
 757
 758 isminigc=
 759 if [ "${force_gc:-0}" = "0" ] && check_interval lastgc $cfg_min_gc_interval; then
 760         if [ -e .needsgc ]; then
 761                 isminigc=1
 762         else
 763                 progress "= [$proj] garbage check skip (last at $(config_get lastgc))"
 764                 exit 0
 765         fi
 766 fi
 767 if [ -e .nogc ]; then
 768         progress "x [$proj] garbage check disabled"
 769         exit 0
 770 fi
 771 if [ -z "$isminigc" ] && [ -e .delaygc ] && [ -e .needsgc ]; then
 772         # Eligible for a full gc but .delaygc is set so it would be skipped
 773         # However .needsgc is also set so transform it into a mini instead
 774         isminigc=1
 775         progress "~ [$proj] garbage check delayed but checking mini because .needsgc"
 776 fi
 777
 778 if [ -n "$isminigc" ]; then
 779         # Perform a "mini" gc
 780         # Note that .delaygc is ignored here as that's only intended for full gc
 781         lock_gc
 782         rm -f .allowgc .needsgc
 783         rm -f objects/pack/pack-*_[rful].keep
 784         remove_crud
 785         coalesce_reflogs
 786         prune_reflogs
 787         compact_reflogs
 788         maintain_auto_gc_hack
 789         generate_auto_gc_update
 790         miniactive=
 791         if [ -f .svnpack ] && [ -n "$svn_mirror" ]; then
 792                 miniactive=1
 793                 progress "+ [$proj] mini garbage check ($(date))"
 794                 make_svn_pack
 795         fi
 796         if [ -z "$cfg_delay_gfi_redelta" ] && [ -n "$gfi_mirror" ]; then
 797                 # $Girocco::Config::delay_gfi_redelta is false, force redeltification now
 798                 if [ -z "$miniactive" ]; then
 799                         miniactive=1
 800                         progress "+ [$proj] mini garbage check ($(date))"
 801                 fi
 802                 repack_gfi_packs
 803         fi
 804         if lotsa_loose_objects; then
 805                 if [ -z "$miniactive" ]; then
 806                         miniactive=1
 807                         progress "+ [$proj] mini garbage check ($(date))"
 808                 fi
 809                 pack_loose_objects
 810         fi
 811         # If there aren't at least 10 non-keep, non-bitmap, non-bndl packs then
 812         # don't actually process them yet
 813         lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl --quiet"
 814         packcnt="$(list_packs --count $lpo objects/pack)" || :
 815         if [ "${packcnt:-0}" -ge 10 ]; then
 816                 if [ -z "$miniactive" ]; then
 817                         miniactive=1
 818                         progress "+ [$proj] mini garbage check ($(date))"
 819                 fi
 820                 if [ -n "$gfi_mirror" ]; then
 821                         repack_gfi_packs
 822                         packcnt="$(list_packs --count $lpo objects/pack)" || :
 823                 fi
 824                 # if repack_gfi_packs dropped the pack count to < 10 don't combine
 825                 if [ "${packcnt:-0}" -ge 10 ]; then
 826                         combine_small_packs
 827                         combine_small_loose_packs
 828                         packcnt="$(list_packs --count $lpo objects/pack)" || :
 829                 fi
 830                 # if we still have more than 10 packs trigger a full gc
 831                 if [ "${packcnt:-0}" -ge 10 ]; then
 832                         # We shouldn't be in a .delaygc state at this point, but if
 833                         # we are then nuke it because we really need a full gc now
 834                         rm -f .delaygc
 835                         git config --unset gitweb.lastgc
 836                         rm -f "$lockf"
 837                         git update-server-info # just in case
 838                         progress "- [$proj] mini garbage check triggering full gc too many packs ($(date))"
 839                         exit 0
 840                 fi
 841         fi
 842         rm -f "$lockf"
 843         if [ -n "$miniactive" ]; then
 844                 git update-server-info
 845                 progress "- [$proj] mini garbage check ($(date))"
 846         else
 847                 progress "= [$proj] mini garbage check nothing but crud removal to do ($(date))"
 848         fi
 849         exit 0
 850 fi
 851
 852 # Avoid unnecessary garbage collections:
 853 #   1. If lastreceive is set and is older than lastgc
 854 #   -AND-
 855 #   2. We are not a fork (! -s alternates) -OR- lastparentgc is older than lastgc
 856
 857 # If lastgc is NOT set or lastreceive is NOT set we MUST run gc
 858 # If we are a fork and lastparentgc is NOT set we MUST run gc
 859
 860 # If the repo is dirty after removing any crud we MUST run gc
 861
 862 gcstart="$(date "$datefmt")"
 863 skipgc=
 864 isfork=
 865 ! [ -s objects/info/alternates ] || isfork=1
 866 lastparentgcsecs=
 867 [ -z "$isfork" ] || lastparentgcsecs="$(config_get_date_seconds lastparentgc)" || :
 868 lastreceivesecs=
 869 if lastreceivesecs="$(config_get_date_seconds lastreceive)" &&
 870    [ "${force_gc:-0}" = "0" ] &&
 871    lastgcsecs="$(config_get_date_seconds lastgc)" &&
 872    [ $lastreceivesecs -lt $lastgcsecs ]; then
 873         # We've run gc since we last received, so maybe we can skip,
 874         # check if not fork or fork and lastparentgc < lastgc
 875         if [ -n "$isfork" ]; then
 876                 if [ -n "$lastparentgcsecs" ] &&
 877                    [ $lastparentgcsecs -lt $lastgcsecs ]; then
 878                         # We've run gc since our parent ran gc so we can skip
 879                         skipgc=1
 880                 fi
 881         else
 882                 # We don't have any alternates (we're not a forK) so we can skip
 883                 skipgc=1
 884         fi
 885 fi
 886
 887 # Prevent any other simultaneous gc operations
 888 lock_gc
 889
 890 # At this point, if .allowgc or .gc_failed exists, it's now crud to be removed
 891 rm -f .allowgc .gc_failed
 892
 893 # Ideally we would do this in post-receive, but that would mean duplicating the
 894 # logic so it's available in the chroot jail and that's highly undesirable
 895 # Instead, since the first gc will be triggered immediately following the first
 896 # push, we do the check here as it's quick and harmless if HEAD is already valid
 897 check_and_set_head || :
 898
 899 # Always get rid of crud
 900 remove_crud
 901
 902 # Always perform reflogs maintenance
 903 coalesce_reflogs
 904 prune_reflogs
 905 compact_reflogs
 906
 907 # Always maintain auto gc hack
 908 maintain_auto_gc_hack
 909 generate_auto_gc_update
 910
 911 # Run 'git svn gc' now for svn mirrors
 912 if [ -n "$svn_mirror" ]; then
 913         git svn gc || :
 914 fi
 915
 916 # Skip the actual gc if .delaygc is set
 917 if [ -e .delaygc ]; then
 918         progress "x [$proj] garbage check delayed (except for crud removal)"
 919         rm -f "$lockf"
 920         exit 0
 921 fi
 922
 923 # Do not skip gc if the repo is dirty
 924 if [ -n "$skipgc" ] && ! is_dirty; then
 925         progress "= [$proj] garbage check nothing but crud removal to do ($(date))"
 926         config_set lastgc "$gcstart"
 927         rm -f "$lockf"
 928         exit 0
 929 fi
 930
 931 bumptime=
 932 if [ -n "$isfork" ] && [ -z "$lastparentgcsecs" ]; then
 933         # set lastparentgc and then update gcstart to be at least 1 second later
 934         config_set lastparentgc "$gcstart"
 935         bumptime=1
 936 fi
 937 if [ -z "$lastreceivesecs" ]; then
 938         # set lastreceive and then update gcstart to be at least 1 second later
 939         config_set lastreceive "$gcstart"
 940         bumptime=1
 941 fi
 942 if [ -n "$bumptime" ]; then
 943         sleep 1
 944         gcstart="$(date "$datefmt")"
 945 fi
 946
 947 progress "+ [$proj] garbage check ($(date))"
 948
 949 newdeltas=
 950 [ -z "$alwaysredelta" ] || newdeltas="$noreusedeltaopt"
 951 if [ -z "$newdeltas" ] && [ -n "$gfi_mirror" ]; then
 952         if [ $(list_packs --exclude-no-idx --count objects/pack) -le \
 953              $(list_packs --exclude-no-idx --count --quiet --only gfi-packs) ]; then
 954                 # Don't bother with repack_gfi_packs since everything's being repacked
 955                 newdeltas="--no-reuse-delta"
 956         fi
 957 fi
 958 if [ -z "$newdeltas" ] && [ -n "$noreusedeltaopt" ] &&
 959    [ $(list_packs --all --exclude-no-idx --count-objects objects/pack) -le $var_redelta_threshold ]; then
 960         # There aren't enough objects to worry about so just redelta to get the best pack
 961         newdeltas="--no-reuse-delta"
 962 fi
 963 if [ -z "$newdeltas" ]; then
 964         # Since we're not going to recompute deltas overall, we need to do the
 965         # "mini" maintenance so that we can get more optimal deltas
 966         [ -z "$noreusedeltaopt" ] || make_svn_pack
 967         repack_gfi_packs
 968         force_single_pack_redelta=
 969         [ -n "$gfi_mirror" ] || [ -n "$svn_mirror" ] || force_single_pack_redelta=1
 970         [ -z "$noreusedeltaopt" ] || combine_small_packs $force_single_pack_redelta
 971         [ -z "$noreusedeltaopt" ] || combine_small_loose_packs $force_single_pack_redelta
 972 fi
 973
 974 #
 975 ## Safe Pruning In Forks
 976 ##
 977 ## We are about to perform garbage collection.  We do NOT use the "git gc" or
 978 ## the "git repack" commands directly as they do not provide enough control over
 979 ## the fine details.  However, we DO maintain a "gc.pid" file during our garbage
 980 ## collection so that a simultaneous "git gc" by an administrator will be
 981 ## blocked (and similarly we refuse to start garbage collection if we cannot
 982 ## create the "gc.pid" file).
 983 ##
 984 ## When we say "gc" in the below description we are referring to our "gc.sh"
 985 ## script, NOT the "git gc" command.
 986 ##
 987 ## If the project we are running garbage collection (gc) on has any forks we
 988 ## must be careful not to remove any objects that while no longer referenced by
 989 ## this project (the parent) are still referenced by one or more forks (the
 990 ## children) otherwise the children will become corrupt and we can't abide
 991 ## corrupt children.
 992 ##
 993 ## One way to accomplish this is to simply hard-link all currently existing
 994 ## loose objects and packs in the parent into all the children that refer to the
 995 ## parent (via a line in their objects/info/alternates file) before beginning
 996 ## the gc operation and then relying on a subsequent gc in the child to clean up
 997 ## any excess objects/packs.  We used to use this strategy but it's very
 998 ## inefficient because:
 999 ##
1000 ##   1. The disk space used by the old pack(s)/object(s) will not be reclaimed
1001 ##      until all children (and their children, if any) run gc by which time
1002 ##      it's quite possible the topmost parent will have run gc again and
1003 ##      hard-linked yet another old pack down to its children (not to mention
1004 ##      loose objects).
1005 ##
1006 ##   2. When using the "-A" option with "git repack", any new objects in the
1007 ##      parent that are not referenced by children will continually get
1008 ##      exploded out of the hard-linked pack in the children whenever the
1009 ##      children run gc.
1010 ##
1011 ##   3. To avoid suboptimal and/or unnecessarily many packs being hard-linked
1012 ##      into child forks, we must run the "mini" gc maintenance before we
1013 ##      perform the hard-linking into the children which provides yet another
1014 ##      source of inefficiency.
1015 ##
1016 ## While we were still using the "-A" option to "git repack" (that was not
1017 ## always the case) to guarantee we can access old ref values for long enough
1018 ## to send out a meaningful mail.sh notification, another, more efficient,
1019 ## option became available to prevent corruption of child forks that continue
1020 ## to refer to objects that are no longer reachable from any ref in the parent.
1021 ##
1022 ## The only things that need be copied (or hard-linked) into the child fork(s)
1023 ## are those objects that have become unreachable from any ref in the parent.
1024 ##
1025 ## When we were using the "git repack -A -d" + "git prune --expire=1.day.ago"
1026 ## technique, the only objects that could ever be removed were loose objects
1027 ## that "git prune" determined were expired.  In that case, loose objects were
1028 ## all that need be hard-linked down to child forks in order to avoid
1029 ## corruption of any child fork(s).
1030 ##
1031 ## The "git repack -A -d" + "git prune --expire=1.day.ago" + hard-linking loose
1032 ## objects to child forks technique remains fundamentally sound from the
1033 ## perspective of supporting simultaneous gc and push and keeping newly
1034 ## unreachable objects around long enough to be sure we can send out meaningful
1035 ## ref change notifications and never corrupting any child forks and never
1036 ## persisting the lifetime of large old packs containing mostly duplicate or
1037 ## unreachable objects as gc percolates through a project's entire fork tree.
1038 ##
1039 ## However, that technique suffers from one potential prodigious pitfall.
1040 ##
1041 ## Unreachable objects come flying out of their packs to splatter all over the
1042 ## objects subdirectories possibly creating a huge, inefficient mess.
1043 ##
1044 ## Often this is not an issue.  Even with a lot of rebasing going on, usually
1045 ## the only objects that will splatter are some commits, trees and the odd blob
1046 ## here and there.  Not enough to be overly concerned about.
1047 ##
1048 ## However, for the reppository that frequently experiences a lot of non-fast-
1049 ## forward updates and/or outright ref deletion, the number of objects suddenly
1050 ## popping out of their packs at "git repack -A -d" time can be overwhelming.
1051 ##
1052 ## To avoid this issue we now use a four phase pack creation strategy.
1053 ## This will result in creation of up to four packs (instead of at most one).
1054 ##
1055 ##   I. A complete pack (with bitmaps if appropriate) gets created including
1056 ##      only "reachable" objects from all refs/... refs plus HEAD.  This will
1057 ##      also serve as the virtual bundle for the repository.
1058 ##
1059 ##  II. A pack of recently-became-unreachable objects and friends is created.
1060 ##      (The "friends" are ref logs, linked working tree HEADs and indicies.)
1061 ##      Because both the pre-receive and update.sh script record all ref
1062 ##      changes we can easily choose the cut off point for "recently".
1063 ##      It is only the fact we maintain those logs in the reflogs subdirectory
1064 ##      that allows this step to be possible.
1065 ##
1066 ## III. If the repository has any forks with a non-zero length alternates file,
1067 ##      yet another pack of "--keep-unreachable" objects is generated that will
1068 ##      not actually be kept in the parent, but hard-linked into all the forks.
1069 ##
1070 ##  IV. Finally, after running "git prune-packed", any remaining loose objects
1071 ##      are migrated into a pack of their own.
1072 ##
1073 ## We then remove any non-.keep packs that existed before we started the
1074 ## process being careful to keep any same-pack pushes for the "Push Pack Redux"
1075 ## race condition (see README-GC).
1076 ##
1077 ## By using "git pack-objects" directly we are able to accomplish this with
1078 ## very little additional effort.
1079 ##
1080 ## The packs produced by (III) are treated almost like ".keep" packs by child
1081 ## forks in that the objects in them are never repacked into any other
1082 ## "--keep-unreachable" packs (but they can migrate into phase I or II packs)
1083 ## and those phase III packs are then hard-linked into any grandchild forks.
1084 ##
1085 ## This avoids the space explosion that could occur if each fork level ended
1086 ## up duplicating the "--keep-unreachable" pack space by repacking those
1087 ## objects (essentially breaking the hard-link to the single copy of those
1088 ## objects).
1089 ##
1090 ## While it is true that each level of forks could potentially add yet another
1091 ## phase III pack to be hard-linked down to its children, such packs will only
1092 ## include unreachable objects not already in any phase III packs that were
1093 ## received from the parent.
1094 ##
1095 ## The space for the phase III packs will not be reclaimed until the gc
1096 ## finishes percolating through the entire "fork tree" of a project.
1097 ##
1098 ## This is not much different than the "git repack -A -d" situation where
1099 ## all the loose objects are hard-linked down into child forks.  In that
1100 ## case forks that actually need any of those objects could gradually reduce
1101 ## the number of objects hard-linked into deeper fork levels.
1102 ##
1103 ## The difference with a phase III "--keep-unreachable" pack is that there
1104 ## cannot be any gradual reduction like that since it would require repacking
1105 ## the pack and breaking the hard-link thereby increasing storage space.  The
1106 ## storage will instead always be reclaimed all at once when all of the
1107 ## projects in the "fork tree" complete their gc.
1108 ##
1109 ## However, the belief is that the huge space win by having all the
1110 ## unreachable objects packed up together far eclipses (when many objects are
1111 ## involved, the single-pack version can end up using 1/20th or less of the
1112 ## disk space compared to having them all as loose objects) any brief minor
1113 ## space savings that might occur under the "git repack -A -d" loose object
1114 ## system prior to the gc collection completing for all the projects in the
1115 ## "fork tree".
1116 #
1117
1118 #
1119 ## utility functions
1120 #
1121
1122 make_packs_ugw() {
1123         find -L "$1" -maxdepth 1 -type f ! -perm -ug+w \
1124                 -name "pack-$octet20*.pack" -exec chmod ug+w '{}' + || :
1125 } 2>/dev/null
1126
1127 get_index_tree() {
1128         if [ -s "$1" ]; then
1129                 GIT_INDEX_FILE="$1"
1130                 export GIT_INDEX_FILE
1131                 git write-tree 2>/dev/null || :
1132                 unset GIT_INDEX_FILE
1133         fi
1134 }
1135
1136 get_detached_head() {
1137         if [ -s "$1" ] && read -r _head <"$1" 2>/dev/null; then
1138                 case "$_head" in $octet20*)
1139                         echo "$_head"
1140                 esac
1141         fi
1142 }
1143
1144 # compute_extra_reachables
1145 # create lines suitable for a packed-refs file mentioning all the
1146 # other refs we might like to keep.
1147 # the current directory MUST be set to the repository's --git-dir
1148 # the following are included:
1149 #   * refs mentioned in repack/packed-refs.extra (if it exists)
1150 #   * refs mentioned in reflogs/... files
1151 #   * tree(s) created from index file(s)
1152 #   * detached linked working tree heads
1153 # Resulting objects are tested for existence and uniqified then output
1154 # one per line under a refs/z* namespace
1155 compute_extra_reachables() {
1156         {
1157                 if [ -s repack/packed-refs.extra ]; then
1158                         LC_ALL=C sed <repack/packed-refs.extra -n \
1159                                 -e 's/^\([0-9A-Fa-f][0-9A-Fa-f]*\).*$/\1/p' \
1160                                 -e 's/^\^\([0-9A-Fa-f][0-9A-Fa-f]*\).*$/\1/p'
1161                 fi
1162                 digits8='[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]'
1163                 find -L reflogs -mindepth 1 -maxdepth 1 -type f -name "$digits8*" -exec gzip -c -d -f '{}' + |
1164                 awk '{print $2; print $3}'
1165                 ! [ -f index ] || get_index_tree index
1166                 if [ -d worktrees ]; then
1167                         find -L worktrees -mindepth 2 -maxdepth 2 -name HEAD -type f -print |
1168                         while read -r lwth; do
1169                                 get_detached_head "$lwth"
1170                                 get_index_tree "${lwth%HEAD}index"
1171                         done
1172                 fi
1173         } | LC_ALL=C sort -u |
1174         git cat-file ${var_have_git_260:+--buffer} --batch-check"${var_have_git_185:+=%(objectname)}" |
1175         awk '!/missing/ {num++; print $1 " " "refs/" substr("zzzzzzzzzzzz", 1, length(num)) "/" num}'
1176 }
1177
1178 #
1179 ## main gc logic
1180 #
1181
1182 # Everything else is more efficient if we do this first
1183 # The "--prune" option is the default since v1.5.0 but it serves as "documentation" here
1184 git pack-refs --all --prune
1185 [ -e packed-refs ] || >>packed-refs # should never happen...
1186
1187 # If we have a logs directory or a worktrees directory expire the ref logs now
1188 # Note that Git itself does not use either --rewrite or --updateref, so neither do we
1189 ! [ -d logs ] && ! [ -d worktrees ] || eval git reflog expire --all "${quiet:+>/dev/null 2>&1}" || :
1190
1191 make_repack_dir
1192 ! [ -e .gc_failed ] || exit 1
1193 rm -f .gc_in_progress # make sure
1194 touch .gc_in_progress # it's truly fresh
1195 rm -f bundles/* objects/pack/pack-*.bndl
1196 # These only exist for a brief time before the packs loose their _f suffix
1197 # "Push Pack Redux" does not apply to these since they were only ever present with _f
1198 rm -f objects/pack/pack-*_f.keep
1199 # This is perhaps a bit aggressive in that if we're suffering from "Push Pack Redux"
1200 # and somehow we get run again immediately after the run where "Push Pack Redux" happened
1201 # and we have garbage collection forced, there's just the barest, almost negligible,
1202 # possibility that the "Push Pack Redux" ref updates _still_ have not happened and we
1203 # should not be removing _r .keep files.  None of the normal Girocco processing can
1204 # cause this.  The second run of this script would have to use the force gc option
1205 # for it to even be possible in the first place.  What's much more likely is that
1206 # the initial run of this script was somehow interrupted in the middle before it
1207 # could get rid of the _r .keep file itself in which case it's better to get rid of
1208 # it now to avoid keeping something around that would perturb our nice and neat gc
1209 rm -f objects/pack/pack-*_r.keep
1210 # We will add .keep files for _u and _l packs if and when we run phase III
1211 # Otherwise they need to not have any .keep files during phases I and II
1212 rm -f objects/pack/pack-*_[ul].keep
1213
1214 # We need to make sure that any non-Girocco (barely tolerated) Git object creation
1215 # activity will be able to "freshen" the pack containing a pre-existing object
1216 # that's being written.  This really should not be necessary as the pre-receive
1217 # hook should make sure this takes place for any incoming pushes.
1218 # However, do it here anyway just in case.
1219 make_packs_ugw objects/pack
1220
1221 # This is only effective with Git v2.3.5 and later and it will only matter when
1222 # we are using one of the "internal_rev_list" modes of pack-objects
1223 # (the combine-packs.sh script never uses any of those modes)
1224 # The "git repack" and "git prune" commands always set this internally themselves
1225 # It makes no difference if there's no repository corruption
1226 GIT_REF_PARANOIA=1 && export GIT_REF_PARANOIA
1227
1228 # All of the options we might want to use with pack-objects were supported
1229 # at some point prior to Git version v1.6.6 which is the minimum version that
1230 # Girocco now requires.  Except for one (--use-bitmap-index).  Several of them
1231 # are "boiler plate" options we always want to use so we bundle them up here.
1232 pkopt="--delta-base-offset --keep-true-parents --non-empty --all-progress-implied"
1233 # We want to use --include-tag, but before Git v2.10.1 it would leave out
1234 # "middle" tags (e.g. a tag of a tag of a commit would omit the tagged tag)
1235 # See http://repo.or.cz/git.git/b773ddea2cd3b08c for details
1236 # ("pack-objects: walk tag chains for --include-tag", 2016-09-07, v2.10.1)
1237 # This is not a free check as it matches all refs against refs/tags/ then
1238 # peels all the annotated tags and checks for inclusion.  The situation in
1239 # which it would add a tag that was not already included by a reachability
1240 # trace that included tag starting points can only occur if a new tag gets
1241 # pushed during gc pointing to something that would have been packed anyway.
1242 # But, it could happen and, really, compared to gc as a whole it's not that
1243 # expensive to perform (provided we do not get an unconnected pack).
1244 [ -z "$var_have_git_2101" ] || pkopt="$pkopt --include-tag"
1245 pkopt="$pkopt ${quiet:---progress} $packopts"
1246
1247 # The git pack-objects command only supports bitmaps if all objects are being
1248 # packed (the "--all" option) and the "--stdout" option is NOT being used.
1249 # Additionally, while packing, if any encountered reachable objects are
1250 # determined to be "not wanted" then no bitmap index will be written anyway.
1251 # While it is theoretically possible that a project with a non-empty alternates
1252 # file ends up packing all objects (because it does not actually use any of the
1253 # objects found in the alternates), it's very unlikely.  And, in the unlikely
1254 # event that did occur, clients would see a message about only using one bitmap
1255 # because Git can only use one bitmap at a time and at least one of the
1256 # alternates is bound to have a bitmap.  Therefore if we see a non-empty
1257 # alternates file, we disable writing bitmaps which avoids the warning and any
1258 # possibility of a client warning as well.  Also if we are running anything
1259 # before Git v2.1.0 (the effective version for repack.writeBitmaps=true) then
1260 # we also always disable bitmap writing.
1261 wbmopt=
1262 [ -z "$var_have_git_210" ] || wbmopt="--write-bitmap-index"
1263 # More recent versions of pack-objects have optimizations when not using the
1264 # --local option.  If we do not have any alternates it's a pointless option.
1265 # If we do have alternates we need to skip writing a bitmap and we cannot
1266 # have a bundle since it must contain all objects.
1267 if [ -n "$isfork" ]; then
1268         lclopt="--local"
1269         wbmopt=
1270         makebndl=
1271 else
1272         lclopt=
1273         makebndl=1
1274 fi
1275
1276 #
1277 ## Phase I
1278 #
1279
1280 wbmstr=
1281 [ -n "$wbmopt" ] || wbmstr=" (bitmaps disabled)"
1282 progress "~ [$proj] running primary full gc pack-objects$wbmstr ($(date))"
1283
1284 gotforks=
1285 ! has_forks_with_alternates "$proj" || gotforks=1
1286
1287 # To avoid "Push Pack Redux" (see README-GC), after collecting the initial
1288 # preexisting non-keep pack list, we rename them so that an incoming push
1289 # pack cannot possibly experience a pack name collision.  Git does not require
1290 # use of the "default" pack names, simply that the proper extensions are used.
1291 # We rename to insert an "_r" just before the extension to avoid "Push Pack Redux"
1292 # name collisions.  Later on we may create an "unreachable" pack for hard-linking
1293 # down into forks and it will have an "_u" inserted just before its extension.
1294 packlist="$(list_packs -C objects/pack --all --exclude-no-idx --exclude-keep --quiet .)" || :
1295 oldpacks=
1296 for oldpack in $packlist; do
1297         oldpack="${oldpack%.pack}"
1298         [ -f "objects/pack/$oldpack.pack" ] || {
1299                 echo >&2 "[$proj] unable to list old pack files"
1300                 exit 1
1301         }
1302         case "$oldpack" in pre-auto-gc-[12])
1303                 # we never disturb pre-auto-gc-1 or pre-auto-gc-2 packs
1304                 continue
1305         esac
1306         oldpackhex="${oldpack#pack-}"
1307         if [ "${oldpackhex#*[!0-9a-fA-F]}" != "$oldpackhex" ]; then
1308                 # names not exclusively hexadecimal do not need renaming
1309                 case "$oldpack" in
1310                 pack-$octet20*_l)
1311                         # _l packs are treated like still-unpacked loose objects
1312                         continue;;
1313                 *_f)
1314                         # _f packs can only be left over from a previously interrupted gc;
1315                         # they need to be renamed to _r now so they're not confused with
1316                         # any freshly generated "final" packs (and we already removed
1317                         # any pre-existing *_f.keep files so we're good to go)
1318                         ;;
1319                 *)
1320                         oldpacks="${oldpacks:+$oldpacks }$oldpack"
1321                         continue;;
1322                 esac
1323         fi
1324         rename_pack "objects/pack/$oldpack" "objects/pack/${oldpack%_f}_r" || {
1325                 echo >&2 "[$proj] unable to rename old pack files"
1326                 exit 1
1327         }
1328         # If the oldpack has a .keep now it means a "Push Pack Redux" is actually
1329         # in progress at this moment and we need to .keep the renamed pack,
1330         # otherwise no "Push Pack Redux" has started yet or it has already finished.
1331         # In either case we're okay because if it's just finished then all ref
1332         # changes have already been made so we don't need a .keep and we will
1333         # see the ref changes and grab all the objects via a reachability trace.
1334         # If it hasn't started yet that's okay because we're done moving that
1335         # name so a complete pack will appear under the old name that we'll
1336         # leave alone.
1337         if [ -f "objects/pack/$oldpack.keep" ]; then
1338                 echo "Push Pack Redux" >"objects/pack/${oldpack%_f}_r.keep"
1339         else
1340                 oldpacks="${oldpacks:+$oldpacks }${oldpack%_f}_r"
1341         fi
1342 done
1343
1344 # We wish to keep deltas from our last full pack so if we're not redeltaing
1345 # then make sure the .pack associated with the .bitmap has a newer mod time
1346 # (If there is no .bitmap then touch the pack with the most objects instead.)
1347 if [ -z "$newdeltas" ]; then
1348         bmpack="$(list_packs --exclude-no-bitmap --exclude-no-idx --max-matches 1 objects/pack)"
1349         [ -n "$bmpack" ] || bmpack="$(list_packs --exclude-no-idx --max-matches 1 --object-limit -1 --include-boundary objects/pack)"
1350         if [ -n "$bmpack" ] && [ -f "$bmpack" ] && [ -s "$bmpack" ]; then
1351                 sleep 1
1352                 touch -c "$bmpack" 2>/dev/null || :
1353                 # We must touch .gc_in_progress here to avoid $bmpack looking
1354                 # like it's been "freshened" when redundant packs are removed
1355                 # It's okay if they have the same mod time, but POSIX does not
1356                 # guarantee an ordering for the "touching" that occurs which is
1357                 # why this must be a separate command but needs no "sleep 1"
1358                 touch .gc_in_progress
1359         fi
1360 fi
1361
1362 # Now we need to make sure that any "freshening" that takes place will actually
1363 # result in a "newer" modification time than the .gc_in_progress file now has
1364 sleep 1
1365
1366 # We run git pack-objects from the repack subdirectory so we can force
1367 # optimized packs to be generated even for repositories that do not have any
1368 # tagged commits
1369 packs="$(git --git-dir=repack pack-objects </dev/null \
1370         $pkopt --all $newdeltas $lclopt ${wbmopt:---honor-pack-keep} "$@" repack/alt/pack/pack)"
1371 vcnt packcnt $packs
1372 [ $packcnt -eq 1 ] || makebndl=
1373
1374 #
1375 ## Phase II
1376 #
1377
1378 progress "~ [$proj] running supplementary gc pack-objects ($(date))"
1379
1380 # Add the "supplementary" refs
1381 compute_extra_reachables >>repack/packed-refs
1382
1383 # Subtract the primary refs
1384 GIT_ALTERNATE_OBJECT_DIRECTORIES="$PWD/repack/alt"
1385 export GIT_ALTERNATE_OBJECT_DIRECTORIES
1386
1387 # For this one we MUST use --local and MUST NOT use --write-bitmap-index
1388 # However, if there is a "logs" subdirectory we need to use --reflog
1389 # We do add it, just in case, if the linked working trees dir is present
1390 # We do not add --indexed-objects as that requires v2.2.0 and it's unclear
1391 # if it properly includes linked working tree index files or not.  The
1392 # above compute_extra_reachables has already included all index trees (thereby
1393 # providing proper --indexed-objects support for all Git versions) making the
1394 # option completely unnecessary.
1395 rflopt=
1396 ! [ -d logs ] && ! [ -d worktrees ] || rflopt=--reflog
1397 spacks="$(git --git-dir=repack pack-objects </dev/null \
1398         $pkopt --honor-pack-keep --all $rflopt $newdeltas --local "$@" repack/alt/pack/pack)"
1399
1400 #
1401 ## Phase III
1402 #
1403
1404 # There's nothing to do for Phase III unless we have forks that refer to our
1405 # project from their alternates file
1406 hlpacks=
1407 upacks=
1408 if [ -n "$gotforks" ]; then
1409
1410         progress "~ [$proj] running keep-unreachable gc pack-objects for forks ($(date))"
1411
1412         # If we are a fork, any pre-existing _u packs need to have a .keep
1413         # for this phase and be added to the hlpacks list otherwise (we are
1414         # not a fork) pre-existing _u packs are anomalies to be treated like
1415         # regular non-_u packs
1416         if [ -n "$isfork" ]; then
1417                 for upack in $(find -L objects/pack -mindepth 1 -maxdepth 1 -name "pack-$octet20*_[ul].pack" -print); do
1418                         upack="${upack%.pack}"
1419                         [ -e "$upack.keep" ] || echo "unreachable" >"$upack.keep"
1420                         case "$upack" in *_l);;*)
1421                                 hlpacks="${hlpacks:+$hlpacks }${upack#objects/pack/pack-}"
1422                         esac
1423                 done
1424         fi
1425         # Using either --no-reuse-delta or --no-reuse-object together with the
1426         # --keep-unreachable option is a very, very, very bad idea when good
1427         # packs are the desired outcome.  If newdeltas are being generated
1428         # then we pack to a temp name, and use combine-packs.sh to get a better
1429         # pack as the result to avoid making a bad --keep-unreachable pack
1430         pfx=
1431         [ -z "$newdeltas" ] || pfx="ku"
1432         upacks="$(git --git-dir=repack pack-objects </dev/null \
1433                 $pkopt --honor-pack-keep --all $rflopt --keep-unreachable --local "$@" repack/alt/pack/${pfx}pack)"
1434         if [ -n "$upacks" ] && [ -n "$newdeltas" ]; then
1435                 progress "~ [$proj] rebuilding keep-unreachable pack deltas"
1436                 oldupacks="$upacks"
1437                 upacks="$(
1438                         printf "repack/alt/pack/${pfx}pack-%s.pack\n" $oldupacks |
1439                         run_combine_packs --names --weak-naming --non-empty --all-progress-implied ${quiet:---progress} \
1440                                 $packopts $newdeltas "$@" repack/alt/pack/pack)"
1441                 eval rm -f "$(printf \""repack/alt/pack/${pfx}pack-%s.*"\"" " $oldupacks)"
1442         fi
1443         for upack in $upacks; do
1444                 rename_pack "repack/alt/pack/pack-$upack" "repack/alt/pack/pack-${upack}_u"
1445         done
1446         rm -f objects/pack/pack-*_[ul].keep
1447         [ -z "$hlpacks" ] && [ -z "$upacks" ] ||
1448         progress "~ [$proj] hard-linking keep-unreachable pack(s) into immediate child forks"
1449
1450         # We have to update the lastparentgc time in the child forks even if they do not get any
1451         # new "unreachable packs" because they need to run gc just in case the parent now has some
1452         # objects that used to only be in the child so they can be removed from the child.
1453         # For example, a "patch" might be developed first in a fork and then later accepted into
1454         # the parent in which case the objects making up the patch in the child fork are now
1455         # redundant (since they're now in the parent as well) and need to be removed from the
1456         # child fork which can only happen if the child fork runs gc.
1457         lastparentgc="$(date "$datefmt")"
1458
1459         # It is enough to copy objects just one level down and get_repo_list
1460         # takes a regular expression (which is automatically prefixed with '^')
1461         # so we can easily match forks exactly one level down from this project
1462         forkdir="$proj"
1463         get_repo_list "$forkdir/[^/:][^/:]*:" |
1464         while read fork; do
1465                 # Ignore forks that do not exist or are symbolic links
1466                 ! [ -L "$cfg_reporoot/$fork.git" ] && [ -d "$cfg_reporoot/$fork.git" ] ||
1467                         continue
1468                 # Or do not have a non-zero length alternates file
1469                 [ -s "$cfg_reporoot/$fork.git/objects/info/alternates" ] ||
1470                         continue
1471                 runupdate=
1472                 # Match hlpacks in parent project if any
1473                 if [ -n "$hlpacks" ]; then
1474                         mkdir -p "$cfg_reporoot/$fork.git/objects/pack"
1475                         eval ln -f "$(printf '"objects/pack/pack-%s.pack" ' $hlpacks)" \
1476                                 "$(printf '"objects/pack/pack-%s.idx" ' $hlpacks)" \
1477                                 '"$cfg_reporoot/$fork.git/objects/pack/"'
1478                         runupdate=1
1479                 fi
1480                 # Match upacks in repack/alt area if any
1481                 if [ -n "$upacks" ]; then
1482                         mkdir -p "$cfg_reporoot/$fork.git/objects/pack"
1483                         eval ln -f "$(printf '"repack/alt/pack/pack-%s_u.pack" ' $upacks)" \
1484                                 "$(printf '"repack/alt/pack/pack-%s_u.idx" ' $upacks)" \
1485                                 '"$cfg_reporoot/$fork.git/objects/pack/"'
1486                         runupdate=1
1487                 fi
1488                 if ! [ -e "$cfg_reporoot/$fork.git/.needsgc" ]; then
1489                         # Trigger a mini gc in the fork if it now has too many packs
1490                         packs="$(list_packs --quiet --count --exclude-no-idx --exclude-keep "$cfg_reporoot/$fork.git/objects/pack")" || :
1491                         if [ -n "$packs" ] && [ "$packs" -ge 20 ]; then
1492                                 >"$cfg_reporoot/$fork.git/.needsgc"
1493                         fi
1494                 fi
1495                 [ -z "$runupdate" ] || git --git-dir="$cfg_reporoot/$fork.git" update-server-info
1496                 # Update the fork's lastparentgc date (must be more recent than $gcstart)
1497                 git --git-dir="$cfg_reporoot/$fork.git" config gitweb.lastparentgc "$lastparentgc"
1498         done
1499 fi
1500
1501 # Now move any primary/supplementary packs back into objects/pack
1502 # then drop any "unfreshened" redundant packs and clear repack/alt
1503
1504 # First make sure the primary pack(s) have the most recent mod time
1505 if [ -n "$packs" ]; then
1506         [ -z "$spacks" ] || sleep 1
1507         printf 'repack/alt/pack/pack-%s.pack\n' $packs | xargs touch -c 2>/dev/null || :
1508 fi
1509
1510 # Move the packs into place but with a _f suffix and a .keep file for now
1511 for pack in $packs $spacks; do
1512         rename_pack "repack/alt/pack/pack-$pack" "objects/pack/pack-${pack}_f"
1513         [ -e "objects/pack/pack-${pack}_f.keep" ] ||
1514         echo "final" >"objects/pack/pack-${pack}_f.keep"
1515 done
1516
1517 # It's possible that one of the $oldpacks had a .bitmap, got renamed (along
1518 # with its .bitmap) and then got "freshened" causing us to not remove it
1519 # However, if $wbmopt is set we most likely now have TWO .bitmap packs!
1520 # This can produce ugly warnings we don't want and possibly get the wrong
1521 # bitmap used since only one .bitmap file can ever be used by Git.
1522 # If this has happened, the .bitmap we want to discard will always have
1523 # an _r suffix so we can just zap any such now since it will leave the pack.
1524 [ -z "$wbmopt" ] || rm -f objects/pack/pack-*_r.bitmap || :
1525
1526 # Remove the redundant packs that have not since been "freshened"
1527 # This does not completely eliminate the race condition window (Girocco's own
1528 # activites -- gc/fetch/receive are immune to the race) but it substantially
1529 # shrinks it down to just the time after the find but before the following rm
1530 >repack/oldpacks
1531 [ -z "$oldpacks" ] ||
1532 printf 'objects/pack/%s.pack\n' $oldpacks |
1533 LC_ALL=C sort >repack/oldpacks
1534 find -L objects/pack -maxdepth 1 -type f -name "pack-$octet20*.pack" -newer .gc_in_progress -print |
1535 LC_ALL=C sort >repack/freshened
1536 deadpacks="$(LC_ALL=C join -v 1 repack/oldpacks repack/freshened | LC_ALL=C sed 's/\.pack$//')"
1537 [ -z "$deadpacks" ] ||
1538 eval echo "$(printf '"%s".* ' $deadpacks)" | xargs rm -f || :
1539
1540 # No need for this anymore
1541 rm -rf repack/alt objects/pack/repack
1542 unset GIT_ALTERNATE_OBJECT_DIRECTORIES
1543
1544 #
1545 ## Phase IV
1546 #
1547
1548 progress "~ [$proj] running gc prune-packed"
1549
1550 # We do not want the redundant packs or any new "--keep-unreachable" pack(s) to be
1551 # present while running prune-packed.  We try to guarantee that any loose object
1552 # (or any object present in a pack with an _l suffix which was created by mini gc)
1553 # that's unreachable persists for at least one $Girocco::Config::min_gc_interval
1554 # (not withstanding administrator interference to force earlier gc to occur).
1555 # If we were to include the redundant/keep-unreachable pack(s) when running
1556 # prune-packed and a loose unreachable object happened to be duplicated in one
1557 # of them we would end up removing it too soon and void our guarantee.
1558 git prune-packed $quiet
1559
1560 progress "~ [$proj] running loose objects gc pack-objects ($(date))"
1561
1562 # Although Git v2.10.0 and later support a --pack-loose-unreachable option,
1563 # we MUST NOT use it for these reasons:
1564 #  1) We're not interested in expensive "unreachable" at this point, only "loose"
1565 #  2) It produces simply horrid packs about 3.8x times larger than they should be
1566 #  3) We don't require anything more than Git v1.6.6
1567 # The only way we could see any _o pack files at this point is if one got
1568 # "freshened" while we were running gc.  If that happens then it gets to live on
1569 # until the next full gc and we need to include it in the loose repack here.
1570 lpacks="$(list_packs --exclude-no-idx --exclude-no-sfx _l --exclude-no-sfx _o --quiet objects/pack |
1571         run_combine_packs --replace --names --loose --weak-naming --non-empty --honor-pack-keep \
1572         --all-progress-implied ${quiet:---progress} $packopts $newdeltas "$@")"
1573
1574 if [ -n "$lpacks" ]; then
1575         # Make sure any primary pack(s) have a more recent mod time than "unreachable" objects packs
1576         if [ -n "$packs" ]; then
1577                 sleep 1
1578                 printf 'objects/pack/pack-%s_f.pack\n' $packs | xargs touch -c 2>/dev/null || :
1579         fi
1580         # We need to identify these packs later so we don't combine_packs them
1581         for objpack in $lpacks; do
1582                 rename_pack "objects/pack/pack-$objpack" "objects/pack/pack-${objpack}_o" || :
1583         done
1584 fi
1585
1586 # Polish up the final packs now
1587 rm -f objects/pack/pack-*_f.keep
1588 for pack in $packs $spacks; do
1589         rename_pack "objects/pack/pack-${pack}_f" "objects/pack/pack-$pack"
1590 done
1591
1592 if [ -n "$lpacks" ]; then
1593         # Finally zap the corresponding loose objects
1594         progress "~ [$proj] running packed loose objects gc prune-packed"
1595         git prune-packed $quiet
1596 fi
1597
1598 ! [ -e .gc_failed ] || exit 1
1599 # These, if they exist, are now meaningless and need to be removed
1600 rm -f gfi-packs .needsgc .svnpack .svnpackgc
1601
1602 # Make sure this stays up to date
1603 git update-server-info
1604
1605 # We must make loose objects group writable so that they
1606 # can be freshened by other pushers.  Technically we need only do this for
1607 # push projects but to enable mirror projects to be more easily converted to
1608 # push projects, we go ahead and do it for all projects.
1609 # By the time we get here we really shouldn't have any of these, but just in case.
1610 { find -L objects/$octet -type f -name "$octet19*" -exec chmod ug+w '{}' + || :; } 2>/dev/null
1611
1612 # darcs:// mirrors have a xxx.log file that will grow endlessly
1613 # if this is a mirror and the file exists, shorten it to 10000 lines
1614 # also take this opportunity to optimize the darcs repo
1615 if ! [ -e .nofetch ] && [ -n "$cfg_mirror" ]; then
1616         url="$(config_get baseurl)" || :
1617         case "$url" in darcs://*)
1618                 if [ -n "$cfg_mirror_darcs" ]; then
1619                         url="${url%/}"
1620                         basedarcs="$(basename "${url#darcs:/}")"
1621                         if [ -f "$basedarcs.log" ]; then
1622                                 tail -n 10000 "$basedarcs.log" >"$basedarcs.log.$$"
1623                                 mv -f "$basedarcs.log.$$" "$basedarcs.log"
1624                         fi
1625                         if [ -d "$basedarcs.darcs" ]; then
1626                                 (
1627                                         cd "$basedarcs.darcs"
1628                                         # without show_progress suppress non-error output
1629                                         [ "${show_progress:-0}" != "0" ] || exec >/dev/null
1630                                         # Note that this does not optimize _darcs/inventories/ :(
1631                                         darcs optimize || :
1632                                 )
1633                         fi
1634                 fi
1635         esac
1636 fi
1637
1638 # Create a matching .bndl header file for the all-in-one pack we just created
1639 # but only if we're not a fork (otherwise the bundle would not be complete)
1640 # and we are running at least Git version 1.7.2 (pack_is_complete always fails otherwise)
1641 if [ -n "$makebndl" ] && [ -n "$var_have_git_172" ]; then
1642         # There should only be one pack in $packs but do some checking...
1643         # The one we just created will have a .idx and will NOT have a .keep
1644         progress "~ [$proj] creating downloadable bundle header"
1645         pkbase=
1646         pkhead=
1647         IFS= read -r curhead <repack/HEAD.orig || :
1648         if
1649                 [ -s "objects/pack/pack-$packs.pack" ] &&
1650                 [ -s "objects/pack/pack-$packs.idx" ] &&
1651                 ! [ -e "objects/pack/pack-$packs.keep" ] &&
1652                 pkhead="$(pack_is_complete "$PWD/objects/pack/pack-$packs.pack" \
1653                         "$PWD/repack/packed-refs.orig" "$curhead")"
1654         then
1655                 pkbase="objects/pack/pack-$packs"
1656         fi
1657         if [ -n "$pkbase" ] && [ -n "$pkhead" ]; then
1658                 {
1659                         symref=
1660                         case "$curhead" in "ref: refs/"?*|"ref:refs/"?*|"refs/"?*)
1661                                 symref="${curhead#ref:}"
1662                                 symref="${symref# }"
1663                         esac
1664                         bndlurl=
1665                         [ -z "$cfg_httpbundleurl" ] || bndlurl=" url=$cfg_httpbundleurl/$proj.git/clone.bundle"
1666                         echo "# v2 git bundle"
1667                         LC_ALL=C sed -ne "/^$octet20$hexdig* refs\/[^ $tab]*\$/ p" <repack/packed-refs.orig
1668                         if [ -n "$symref" ]; then
1669                                 printf "$pkhead HEAD\0symref=HEAD:%s%s\n" "$symref" "$bndlurl"
1670                         else
1671                                 if [ -n "$bndlurl" ]; then
1672                                         printf "$pkhead HEAD\0%s\n" "${bndlurl:# }"
1673                                 else
1674                                         echo "$pkhead HEAD"
1675                                 fi
1676                         fi
1677                         echo ""
1678                 } >"$pkbase.bndl"
1679                 bndletag="$("$cfg_basedir/bin/rangecgi" --etag -m 1 "$pkbase.bndl" "$pkbase.pack")" || :
1680                 bndlsha="$(printf '%s' "$bndletag" | git hash-object --stdin)" || :
1681                 if [ -n "$bndletag" ]; then
1682                         case "$bndlsha" in $octet20*)
1683                                 bndlshatrailer="${bndlsha#????????}"
1684                                 bndlshaprefix="${bndlsha%$bndlshatrailer}"
1685                                 bndlname="$(TZ=UTC date +%Y%m%d_%H%M%S)-${bndlshaprefix:-0}"
1686                                 [ -d bundles ] || mkdir bundles
1687                                 echo "${pkbase#objects/pack/}.bndl" >"bundles/$bndlname"
1688                                 echo "${pkbase#objects/pack/}.pack" >>"bundles/$bndlname"
1689                                 ln -s -f -n "$bndlname" bundles/latest
1690                         esac
1691                 fi
1692         fi
1693 fi
1694
1695 # Record the size of this repo as the sum of its clone packed-refs + *.pack sizes as 1024-byte blocks
1696 eval "reposizek=$(( $(
1697         echo 0 $(du -k repack/packed-refs.orig $(printf 'objects/pack/pack-%s.pack ' $packs) 2>/dev/null |
1698                 LC_ALL=C awk '{print $1}') |
1699         LC_ALL=C sed -e 's/ / + /g') ))"
1700 config_set_raw girocco.reposizek "${reposizek:-0}"
1701
1702 # Now we're finally done with this
1703 rm -rf repack
1704
1705 # We didn't used to do anything about rerere or worktrees but we're
1706 # trying to make nice with linked working trees these days :)
1707 # Maybe even non-bare repositories too, but *shush* about those ;)
1708 if [ -n "$var_have_git_250" ] && [ -d worktrees ]; then
1709         # The value "3.months.ago" is hard-coded into gc.c rather than
1710         # having the default be in worktree.c so we must provide it if
1711         # we get nothing out of the gc.worktreePruneExpire config item
1712         # Prior to Git v2.6.0 the config item was gc.pruneworktreesexpire
1713         # however we just always use the newer name no matter what Git version
1714         expiry="$(git config --get gc.worktreePruneExpire 2>/dev/null)" || :
1715         eval git worktree prune --expire '"${expiry:-3.months.ago}"' "${quiet:+>/dev/null 2>&1}" || :
1716 fi
1717 # git rerere does it right and handles its own default/config'd expiration values
1718 ! [ -d rr-cache ] || eval git rerere gc "${quiet:+>/dev/null 2>&1}" || :
1719
1720 # We use $gcstart here to avoid a race where a push occurs during the gc itself
1721 # and the next future gc could be incorrectly skipped if we used the current
1722 # timestamp here instead
1723 config_set lastgc "$gcstart"
1724 rm -f "$lockf"
1725
1726 progress "- [$proj] garbage check ($(date))"