jobd/gc.sh

   1 #!/bin/sh
   2
   3 # NOTE: additional options can be passed to git repack by specifying
   4 #       them after the project name, for example:
   5 #         gc.sh my-project -f
   6
   7 . @basedir@/shlib.sh
   8
   9 set -e
  10
  11 if [ $# -lt 1 ]; then
  12         echo "Usage: gc.sh projname [extra-repack-args]" >&2
  13         exit 1
  14 fi
  15
  16 # packing options
  17 packopts="--depth=50 --window=50 --window-memory=${var_window_memory:-1g}"
  18 quiet=; [ -n "$show_progress" ] || quiet=-q
  19
  20 umask 002
  21 [ "$cfg_permission_control" != "Hooks" ] || umask 000
  22 clean_git_env
  23
  24 pidactive() {
  25         if _result="$(kill -0 "$1" 2>&1)"; then
  26                 # process exists and we have permission to signal it
  27                 return 0
  28         fi
  29         case "$_result" in *"not permitted"*)
  30                 # we do not have permission to signal the process
  31                 return 0
  32         esac
  33         # process does not exist
  34         return 1
  35 }
  36
  37 createlock() {
  38         # A .lock file should only exist for much less than a second.
  39         # If we see a stale lock file (> 1h old), remove it and then,
  40         # just in case, wait 30 seconds for any process whose .lock
  41         # we might have just removed (it's racy) to finish doing what
  42         # should take much less than a second to do.
  43         _stalelock="$(find "$1.lock" -maxdepth 1 -mmin +60 -print 2>/dev/null || :)"
  44         if [ -n "$_stalelock" ]; then
  45                 rm -f "$_stalelock"
  46                 sleep 30
  47         fi
  48         for _try in p p n; do
  49                 if (set -C; > "$1.lock") 2>/dev/null; then
  50                         echo "$1.lock"
  51                         return 0
  52                 fi
  53                 # delay and try again
  54                 [ "$_try" != "p" ] || sleep 1
  55         done
  56         # cannot create lock file
  57         return 1
  58 }
  59
  60 # The pre-receive script creates one ref log file per push but we want them to
  61 # be coalesced into one ref log file per day.  We are guaranteed that any files
  62 # we find to coalesce are NOT currently being written to since they are always
  63 # written first as temporary files and then moved into place.  We attempt to
  64 # transfer the most recent modification time to the coalesced log file which
  65 # would step on its mod time if it were being written to directly, but if we
  66 # find per-process ref log files then it must be a push project and the only
  67 # thing that would write directly to the main per-day log file would be a
  68 # mirror project so there's actually no conflict.
  69 # Also, if the clock is wonky (or was futzed with) we may have both YYYYMMDD
  70 # and YYYYMMDD.gz present in which case combine them into YYYYMMDD
  71 coalesce_reflogs() {
  72         [ -d reflogs ] || return 0
  73         rm -f .gc_failed
  74         find reflogs -maxdepth 1 -type f -name "[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]" -print |
  75         while read -r rname; do
  76                 if [ -e "$rname.gz" ]; then
  77                         if [ -s "$rname" ]; then
  78                                 # Presumably the .gz file must have been created before the non-gz
  79                                 # file since it had to be uncompressed at some point therefore
  80                                 # we need to append the non-gz contents to it but keep the non-gz
  81                                 # contents timestamp so we rename to YYYYMMDD_ which will sort first
  82                                 # and be picked up in the next step if we are interrupted in the middle.
  83                                 # If a YYYYMMDD_ file already exists we append to it and transfer the
  84                                 # timestamp.  Finally we transfer the YYYYMMDD_ timestamp to the result
  85                                 # and remove the YYYYMMDD_ temporary file leaving the result uncompressed.
  86                                 if [ -e "${rname}_" ]; then
  87                                         cat "$rname" >>"${rname}_"
  88                                         touch -r "$rname" "${rname}_"
  89                                         rm -f "$rname"
  90                                         ! [ -e "$rname" ]
  91                                 else
  92                                         mv "$rname" "${rname}_"
  93                                 fi
  94                                 gzip -d "$rname.gz" </dev/null
  95                                 [ -e "$rname" ] && ! [ -e "$rname.gz" ]
  96                                 cat "${rname}_" >>"$rname"
  97                                 touch -r "${rname}_" "$rname"
  98                                 rm -f "${rname}_"
  99                         else
 100                                 # Just remove the empty file to resolve the problem
 101                                 rm -f "$rname"
 102                         fi
 103                 fi
 104         done
 105         find 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 |
 106         while read -r rname; do
 107                 logname="${rname%%_*}"
 108                 # If someone's been futzing with the date, the file we want to
 109                 # append to could already have been compressed, so we just uncompress
 110                 # it here.  The previous block guarantees we do not have both a compressed
 111                 # and uncompressed version present at the same time.
 112                 if [ -e "$logname.gz" ]; then
 113                         gzip -d "$logname.gz" </dev/null
 114                         [ -e "$logname" ] && ! [ -e "$logname.gz" ]
 115                 fi
 116                 cat "$rname" >>"$logname"
 117                 touch -r "$rname" "$logname"
 118                 rm -f "$rname"
 119                 if [ -e "$rname" ]; then
 120                         >.gc_failed
 121                         echo "! [$proj] failed to remove $rname" >&2
 122                         exit 1 # will only exit subshell created by "|"
 123                 fi
 124         done
 125         ! [ -e .gc_failed ]
 126 }
 127
 128 # Remove any files in reflogs that are older than $cfg_reflogs_lifetime days
 129 prune_reflogs() {
 130         [ -d reflogs ] || return 0
 131         exp="$(( ${cfg_reflogs_lifetime:-1} * 1440 ))"
 132         [ $exp -gt 0 ] || exp=1440
 133         [ $exp -le 43200 ] || exp=43200
 134         find reflogs -maxdepth 1 -type f -mmin "+$exp" -print0 | xargs -0 rm -f
 135 }
 136
 137 # Compact any reflogs that are not today's UTC date unless a .gz version exists
 138 compact_reflogs() {
 139         [ -d reflogs ] || return 0
 140         _td="reflogs/$(TZ=UTC date '+%Y%m%d')"
 141         find reflogs -maxdepth 1 -type f -name "[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]" -print |
 142         while read -r rname; do
 143                 [ "$rname" != "$_td" ] || continue
 144                 ! [ -e "$rname.gz" ] || continue
 145                 gzip -9 "$rname" </dev/null
 146         done
 147 }
 148
 149 # return true if there's more than one objects/pack-<sha>.pack file or
 150 # ANY sha-1 files in objects
 151 is_dirty() {
 152         _packs=$(find objects/pack -type f -name "pack-$octet20.pack" -print | head -n 2 | LC_ALL=C wc -l)
 153         if [ $_packs != 1 ] && [ $_packs != 0 ]; then
 154                 return 0
 155         fi
 156         _objs=$(find objects/$octet -type f -name "$octet19" -print 2>/dev/null | head -n 1 | LC_ALL=C wc -l)
 157         [ $_objs -ne 0 ]
 158 }
 159
 160 # make sure combine-packs uses the correct Git executable
 161 run_combine_packs() {
 162         PATH="$var_git_exec_path:$cfg_basedir/bin:$PATH" @basedir@/jobd/combine-packs.sh "$@"
 163 }
 164
 165 # combine the input pack(s) into a new pack (or possibly packs if packSizeLimit set)
 166 # input pack names are read from standard input one per line delimited by the first
 167 # ':', ' ' or '\n' character on the line (which allows gfi-packs to be read directly)
 168 # all arguments, if any, are passed to pack-objects as additional options
 169 # returns non-zero on failure AND creates .gc_failed in that case
 170 combine_packs() {
 171         rm -f .gc_failed
 172         find objects/pack -maxdepth 1 -type f -name '*.zap*' -print0 | xargs -0 rm -f
 173         run_combine_packs --replace "$@" $packopts --all-progress-implied $quiet --non-empty || {
 174                 >.gc_failed
 175                 return 1
 176         }
 177         return 0
 178 }
 179
 180 # if the current directory is_gfi_mirror then repack all packs listed in gfi-packs
 181 repack_gfi_packs() {
 182         [ -n "$gfi_mirror" ] || return 0
 183         [ -d objects/pack ] || { rm -f gfi-packs; return 0; }
 184         progress "~ [$proj] redeltifying poor quality git fast-import packs"
 185         combine_packs --ignore-missing --no-reuse-delta <gfi-packs
 186         rm -f gfi-packs
 187         return 0
 188 }
 189
 190 # combine small packs into larger pack(s)
 191 # we avoid any keep, bndl or bitmap packs
 192 # if the optional argument is non-empty even a single small pack will be redeltad
 193 combine_small_packs() {
 194         _didprogress=
 195         _minsmallpacks=2
 196         if [ -n "$1" ] && [ -n "$noreusedeltaopt" ]; then
 197                 _minsmallpacks=1
 198         fi
 199         _lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl --quiet"
 200         _lpo="$_lpo --object-limit $var_redelta_threshold objects/pack"
 201         while
 202                 _cnt="$(list_packs --count $_lpo || :)"
 203                 test "${_cnt:-0}" -ge $_minsmallpacks
 204         do
 205                 [ -n "$_didprogress" ] || {
 206                         progress "~ [$proj] combining small packs into a single larger pack"
 207                         _didprogress=1
 208                 }
 209                 _newp="$(list_packs $_lpo | combine_packs --names $noreusedeltaopt)"
 210                 _newc="$(echo $(echo "$_newp" | LC_ALL=C wc -w))"
 211                 # be paranoid and exit the loop if we haven't reduced the number of packs
 212                 [ $_newc -lt $_cnt ] || break
 213                 _minsmallpacks=2
 214         done
 215         return 0
 216 }
 217
 218 # Unfortunately git-svn lacks the ability to store newly fetched revisions as a pack.
 219 # However, the fetch code conveniently sets .svnpack just before it runs git-svn fetch
 220 # so that it's easy to find all the objects that have been fetched by git-svn and
 221 # combine them into a pack.  The --no-reuse-delta option is meaningless here since
 222 # everything to be packed is a loose object and therefore not a delta so deltification
 223 # will always take place.
 224 make_svn_pack() {
 225         [ -f .svnpack ] && [ -n "$svn_mirror" ] || return 0
 226         rm -f .svnpackgc
 227         mv -f .svnpack .svnpackgc
 228         progress "~ [$proj] combining loose git-svn objects into a pack"
 229         _newp="$(find objects/$octet -maxdepth 1 -type f -newer .svnpackgc -name "$octet19" -print 2>/dev/null |
 230         LC_ALL=C awk -F / '{print $2 $3}' |
 231         run_combine_packs --objects --names $packopts --incremental --all-progress-implied $quiet --non-empty)" || {
 232                 mv -f .svnpackgc .svnpack
 233                 >.gc_failed
 234                 return 1
 235         }
 236         if [ -n "$_newp" ]; then
 237                 # remove the now-redundant loose objects -- this is always safe
 238                 # even during a concurrent push because a reprepare_packed_git
 239                 # will be triggered if an object that should be there is not
 240                 # found thereby finding it in the new pack instead
 241                 git prune-packed $quiet
 242         fi
 243         rm -f .svnpackgc
 244 }
 245
 246 # HEADSHA="$(pack_is_complete /full/path/to/some.pack /full/path/to/packed-refs "$(cat HEAD)")"
 247 pack_is_complete() {
 248         # Must have a matching .idx file and a non-empty packed-refs file
 249         [ -s "${1%.pack}.idx" ] || return 1
 250         [ -s "$2" ] || return 1
 251         _headsha=
 252         case "$3" in
 253                 $octet20)
 254                         _headsha="$3"
 255                         ;;
 256                 "ref: refs/"?*|"ref:refs/"?*|"refs/"?*)
 257                         _headmatch="${3#ref:}"
 258                         _headmatch="${_headmatch# }"
 259                         _headmatchpat="$(echo "$_headmatch" | LC_ALL=C sed -e 's/\([.$]\)/\\\1/g')"
 260                         _headsha="$(LC_ALL=C grep -e "^$octet20 $_headmatchpat\$" < "$2" | \
 261                                     LC_ALL=C cut -d ' ' -f 1)"
 262                         case "$_headsha" in $octet20) :;; *)
 263                                 return 1
 264                         esac
 265                         ;;
 266                 *)
 267                         # bad HEAD
 268                         return 1
 269         esac
 270         rm -rf pack_is_complete_test
 271         mkdir pack_is_complete_test
 272         mkdir pack_is_complete_test/refs
 273         mkdir pack_is_complete_test/objects
 274         mkdir pack_is_complete_test/objects/pack
 275         echo "$_headsha" > pack_is_complete_test/HEAD
 276         ln -s "$1" pack_is_complete_test/objects/pack/
 277         ln -s "${1%.pack}.idx" pack_is_complete_test/objects/pack/
 278         ln -s "$2" pack_is_complete_test/packed-refs
 279         _count="$(git --git-dir=pack_is_complete_test rev-list --count --all 2>/dev/null || :)"
 280         rm -rf pack_is_complete_test
 281         [ -n "$_count" ] || return 1
 282         [ "$_count" -gt 0 ] 2>/dev/null || return 1
 283         echo "$_headsha"
 284 }
 285
 286 # On return a "$lockf" will have been created that must be removed when gc is done
 287 lock_gc() {
 288         # be compatibile with gc.pid file from newer Git releases
 289         lockf=gc.pid
 290         hn="$(hostname)"
 291         active=
 292         if [ "$(createlock "$lockf")" ]; then
 293                 # If $lockf is:
 294                 #   1) less than 12 hours old
 295                 #   2) contains two fields (pid hostname) NO trailing NL
 296                 #   3) the hostname is different OR the pid is still alive
 297                 # then we exit as another active process is holding the lock
 298                 if [ "$(find "$lockf" -maxdepth 1 -mmin -720 -print 2>/dev/null)" ]; then
 299                         apid=
 300                         ahost=
 301                         read -r apid ahost ajunk < "$lockf" || :
 302                         if [ "$apid" ] && [ "$ahost" ]; then
 303                                 if [ "$ahost" != "$hn" ] || pidactive "$apid"; then
 304                                         active=1
 305                                 fi
 306                         fi
 307                 fi
 308         else
 309                 echo >&2 "[$proj] unable to create gc.pid.lock file"
 310                 exit 1
 311         fi
 312         if [ -n "$active" ]; then
 313                 rm -f "$lockf.lock"
 314                 echo >&2 "[$proj] gc already running on machine '$ahost' pid '$apid'"
 315                 exit 1
 316         fi
 317         printf "%s %s" "$$" "$hn" > "$lockf.lock"
 318         chmod 0664 "$lockf.lock"
 319         mv -f "$lockf.lock" "$lockf"
 320 }
 321
 322 # Create a repack subdirectory such that running repack in it will pack the
 323 # same things that a pack in the normal directory would except that the pack
 324 # is guaranteed to be generated in an optimized order by adding a suitable
 325 # synthesized ref in the refs/tags namespace (yes, pack-objects.c really does
 326 # behave differently depending on the contents of the refs/tags namespace).
 327 # Before calling this, pack-refs --all MUST be performed or the wrong pack
 328 # will end up being made.
 329 # If a ref deletion is pushed after making the repack subdir but before the
 330 # the actual repack, the discarded objects will be packed -- no big deal,
 331 # they'll get discarded the next time gc runs.
 332 # If a fast-forward ref update is pushed after making the repack subdir but
 333 # before the actual repack, it will be picked up and the new objects packed
 334 # (subject to the normal git repack race about picking such updates up).
 335 # If a non-fast-forward ref update is pushed after making the repack subdir but
 336 # before the actual repack, it will be picked up like a fast-forward update but
 337 # the discarded objects will be included like a ref deletion (until the next
 338 # scheduled gc takes place).
 339 make_repack_dir() {
 340         ! [ -d repack ] || rm -rf repack
 341         ! [ -d repack ] || { echo >&2 "[$proj] cannot remove repack subdirectory"; exit 1; }
 342         mkdir repack
 343         [ -d info ] || mkdir info
 344         ln -s ../config  repack/config
 345         ln -s ../info    repack/info
 346         ln -s ../objects repack/objects
 347         ln -s ../refs    repack/refs
 348         sed 's, refs/, refs/!/,' < packed-refs > repack/packed-refs
 349         optref="$(git rev-list -n 1 --all 2>/dev/null || :)"
 350         if [ -n "$optref" ]; then
 351                 echo "$optref refs/tags/!" >> repack/packed-refs
 352                 echo "$optref" > repack/HEAD
 353         else
 354                 cat HEAD > repack/HEAD
 355         fi
 356 }
 357
 358 # Remove any crud that's been left behind by interrupted operations
 359 # that did not clean up after themselves
 360 remove_crud() {
 361         # Remove any existing FETCH_HEAD
 362         # There can only be a FETCH_HEAD if we've been fetching, not if we've been
 363         # receiving pushes (those never create a FETCH_HEAD).
 364         # And if we're fetching because we're a mirror, we know we're not fetching right
 365         # now since jobd.pl never runs a project's fetch simultaneously with its gc.
 366         # Therefore any existing FETCH_HEAD is junk.  And it may be many megabytes if
 367         # there were a lot of refs.
 368         rm -f FETCH_HEAD
 369
 370         # remove any existing pack_is_complete_test or repack subdirectories
 371         # If either exists when this function is called it's crud
 372         rm -rf pack_is_complete_test repack
 373
 374         # Remove any stale pack remnants that are more than an hour old.
 375         # Stale pack fragments are defined as any pack-<sha1>.ext where .ext is NOT
 376         # .pack AND the corresponding .pack DOES NOT exist.  A bunch of stale
 377         # pack-<sha1>.idx files without their corresponding .pack files are worthless
 378         # and just waste space.  Normally there shouldn't be any remnants but actually
 379         # this can happen when things are interrupted at just the wrong time.
 380         # Note that the objects/pack directory is created by git init and should
 381         # always exist.
 382         find objects/pack -maxdepth 1 -type f -mmin +60 -name "pack-$octet20.?*" -print | \
 383         LC_ALL=C sed -e 's/^objects\/pack\/pack-//; s/\..*$//' | LC_ALL=C sort -u | \
 384         while read packsha; do
 385                 [ ! -e "objects/pack/pack-$packsha.pack" ] || continue
 386                 rm -f "objects/pack/pack-$packsha".?*
 387         done
 388
 389         # Remove any stale tmp reflogs files that are more than one hour old.
 390         # Since they are created only while the pre-receive hook is running and
 391         # all it does is process a bunch of refs passed to it on standard input
 392         # it's inconceivable that it would ever take as much as an hour to run.
 393         if [ -d reflogs ]; then
 394                 find reflogs -maxdepth 1 -type f -mmin +60 -name "tmp_*" -print0 | xargs -0 rm -f
 395         fi
 396
 397         # Remove any stale pack .keep files that are more than 12 hours old.
 398         # We don't do anything to create any permanent pack .keep files, so they must
 399         # be remnants from some failed push or something.  Removing the .keep will
 400         # allow the pack to be properly repacked.
 401         find objects/pack -maxdepth 1 -type f -mmin +720 -name "pack-$octet20.keep" -print0 | xargs -0 rm -f
 402
 403         # Remove any stale tmp_pack_*, tmp_idx_*, tmp_bitmap_*, packtmp-* or .tmp-*-pack files
 404         # that are more than 12 hours old.
 405         find objects/pack -maxdepth 1 -type f -mmin +720 \( \
 406                 -name "tmp_pack_?*" -o -name "tmp_idx_?*" -o -name "tmp_bitmap_?*" -o \
 407                 -name "packtmp-?*" -o -name ".tmp-?*-pack" \
 408                 \) -print0 | xargs -0 rm -f
 409
 410         # Remove any stale shallow_* files that are more than 12 hours old.
 411         # These can be left behind by Git >= 1.8.4.2 and < 2.0.0 when a client
 412         # requests a shallow clone.
 413         find . -maxdepth 1 -type f -mmin +720 -name "shallow_?*" -print0 | xargs -0 rm -f
 414
 415         # Remove any stale *.temp files in the objects area that are more than 12 hours old.
 416         # This can be stale sha1.temp, or stale *.pack.temp so we kill all stale *.temp.
 417         find objects -type f -mmin +720 -name "*.temp" -print0 | xargs -0 rm -f
 418
 419         # Remove any stale *.lock files in the htmlcache area that might have been left
 420         # behind after an abnormal exit during an attempt to update a cached file and
 421         # are more than 1 hour old.
 422         ! [ -d htmlcache ] || find htmlcache -type f -mmin +60 -name "*.lock" -print0 | xargs -0 rm -f
 423
 424         # Remove any stale git-svn temp files that are more than 12 hours old.
 425         # The git-svn process creates temp files with random 10 character names
 426         # in the root of $GIT_DIR.  Unfortunately they do not have a recognizable
 427         # prefix, so we just have to kill any files with a 10-character name.  We
 428         # do this only for git-svn mirrors.  All characters are chosen from
 429         # [A-Za-z0-9_] so we can at least check that and fortunately the only
 430         # collision is 'FETCH_HEAD' but that shouldn't matter.
 431         # There may also be temp files with a Git_ prefix as well.
 432         if [ -n "$svn_mirror" ]; then
 433                 _randchar='[A-Za-z0-9_]'
 434                 _randchar2="$_randchar$_randchar"
 435                 _randchar4="$_randchar2$_randchar2"
 436                 _randchar10="$_randchar4$_randchar4$_randchar2"
 437                 find . -maxdepth 1 -type f -mmin +720 -name "$_randchar10" -print0 | xargs -0 rm -f
 438                 find . -maxdepth 1 -type f -mmin +720 -name "Git_*" -print0 | xargs -0 rm -f
 439         fi
 440
 441         # Remove any stale fast_import_crash_<pid> files that are more than 3 days old.
 442         if [ -n "$gfi_mirror" ]; then
 443                 find . -maxdepth 1 -type f -mmin +4320 -name "fast_import_crash_?*" -print0 | xargs -0 rm -f
 444         fi
 445 }
 446
 447 #
 448 ## Garbage Collection Types
 449 ##
 450 ## There are two kinds of possible garbage collection (gc) operations:
 451 ##
 452 ## 1.  A normal, full gc
 453 ## 2.  A "mini" gc
 454 ##
 455 ## If the full garbage collection interval has expired (or gc has never been
 456 ## run), then a normal, full gc will take place.  Otherwise, a "mini" gc will
 457 ## take place if the file .needsgc exists.
 458 ##
 459 ## A "mini" gc is similar to "git gc --auto" in that it may not end up actually
 460 ## doing anything unless the right conditions are present so it's not a burden
 461 ## to run it often.  If the file .needsgc exists, a "mini" gc will occur at
 462 ## the next opportunity.
 463 ##
 464 ## Note, however, that the .nogc file suppresses ALL gc activity (normal or mini).
 465 #
 466
 467 proj="${1%.git}"
 468 shift
 469 cd "$cfg_reporoot/$proj.git"
 470 [ -d objects/pack ] || { rm -f gfi-packs; mkdir -p objects/pack; }
 471 mirror_url="$(get_mirror_url)"
 472 svn_mirror=
 473 ! is_svn_mirror_url "$mirror_url" || svn_mirror=1
 474 gfi_mirror=
 475 if [ -f gfi-packs -a -s gfi-packs ] && is_gfi_mirror_url "$mirror_url"; then
 476         gfi_mirror=1
 477 fi
 478
 479 # If git config --bool --get girocco.redelta is explicitly false then automatic
 480 # redelta when there are less than $var_redelta_threshold objects will be suppressed.
 481 # On the other hand, if git config --get girocco.redelta is "always" then, on a full
 482 # gc only, for the final repack, deltas will always be recomputed.
 483 # This can be set on a per-project basis to avoid unusual pathological gc behavior.
 484 # Setting this will hurt efficiency of the affected repository.
 485 # Note that fast-import packs ALWAYS get new deltas regardless of this setting.
 486 noreusedeltaopt="--no-reuse-delta"
 487 [ "$(git config --bool --get girocco.redelta 2>/dev/null || :)" != "false" ] || noreusedeltaopt=
 488 alwaysredelta=
 489 [ "$(git config --get girocco.redelta 2>/dev/null || :)" != "always" ] || alwaysredelta=1
 490
 491 trap 'e=$?; rm -f .gc_in_progress; if [ $e != 0 ]; then echo "gc failed dir: $PWD" >&2; fi' EXIT
 492 trap 'exit 130' INT
 493 trap 'exit 143' TERM
 494
 495 # date -R is linux-only, POSIX equivalent is '+%a, %d %b %Y %T %z'
 496 datefmt='+%a, %d %b %Y %T %z'
 497
 498 isminigc=
 499 if check_interval lastgc $cfg_min_gc_interval; then
 500         if [ -e .needsgc ]; then
 501                 isminigc=1
 502         else
 503                 progress "= [$proj] garbage check skip (last at $(config_get lastgc))"
 504                 exit 0
 505         fi
 506 fi
 507 if [ -e .nogc ]; then
 508         progress "x [$proj] garbage check disabled"
 509         exit 0
 510 fi
 511
 512 if [ -n "$isminigc" ]; then
 513         # Perform a "mini" gc
 514         # Note that .delaygc is ignored here as that's only intended for full gc
 515         lock_gc
 516         rm -f .allowgc .needsgc
 517         remove_crud
 518         coalesce_reflogs
 519         prune_reflogs
 520         compact_reflogs
 521         miniactive=
 522         if [ -f .svnpack ] && [ -n "$svn_mirror" ]; then
 523                 miniactive=1
 524                 progress "+ [$proj] mini garbage check (`date`)"
 525                 make_svn_pack
 526         fi
 527         if [ -z "$cfg_delay_gfi_redelta" ] && [ -n "$gfi_mirror" ]; then
 528                 # $Girocco::Config::delay_gfi_redelta is false, force redeltification now
 529                 if [ -z "$miniactive" ]; then
 530                         miniactive=1
 531                         progress "+ [$proj] mini garbage check (`date`)"
 532                 fi
 533                 repack_gfi_packs
 534         fi
 535         # If there aren't at least 10 non-keep, non-bitmap, non-bndl packs then
 536         # don't actually process them yet
 537         lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl --quiet"
 538         packcnt="$(list_packs --count $lpo objects/pack || :)"
 539         if [ "${packcnt:-0}" -ge 10 ]; then
 540                 if [ -z "$miniactive" ]; then
 541                         miniactive=1
 542                         progress "+ [$proj] mini garbage check (`date`)"
 543                 fi
 544                 if [ -n "$gfi_mirror" ]; then
 545                         repack_gfi_packs
 546                         packcnt="$(list_packs --count $lpo objects/pack || :)"
 547                 fi
 548                 # if repack_gfi_packs dropped the pack count to < 10 don't combine
 549                 if [ "${packcnt:-0}" -ge 10 ]; then
 550                         combine_small_packs
 551                         packcnt="$(list_packs --count $lpo objects/pack || :)"
 552                 fi
 553                 # if we still have more than 10 packs trigger a full gc
 554                 if [ "${packcnt:-0}" -ge 10 ]; then
 555                         # We shouldn't be in a .delaygc state at this point, but if
 556                         # we are then nuke it because we really need a full gc now
 557                         rm -f .delaygc
 558                         git config --unset gitweb.lastgc
 559                         rm -f "$lockf"
 560                         progress "- [$proj] mini garbage check triggering full gc too many packs (`date`)"
 561                         exit 0
 562                 fi
 563         fi
 564         rm -f "$lockf"
 565         if [ -n "$miniactive" ]; then
 566                 git update-server-info
 567                 progress "- [$proj] mini garbage check (`date`)"
 568         else
 569                 progress "= [$proj] mini garbage check nothing but crud removal to do (`date`)"
 570         fi
 571         exit 0
 572 fi
 573
 574 # Avoid unnecessary garbage collections:
 575 #   1. If lastreceive is set and is older than lastgc
 576 #   -AND-
 577 #   2. We are not a fork (! -s alternates) -OR- lastparentgc is older than lastgc
 578
 579 # If lastgc is NOT set or lastreceive is NOT set we MUST run gc
 580 # If we are a fork and lastparentgc is NOT set we MUST run gc
 581
 582 # If the repo is dirty after removing any crud we MUST run gc
 583
 584 gcstart="$(date "$datefmt")"
 585 skipgc=
 586 isfork=
 587 [ -s objects/info/alternates ] && isfork=1
 588 lastparentgcsecs=
 589 [ -n "$isfork" ] && lastparentgcsecs="$(config_get_date_seconds lastparentgc || :)"
 590 lastreceivesecs=
 591 if lastreceivesecs="$(config_get_date_seconds lastreceive)" && \
 592    lastgcsecs="$(config_get_date_seconds lastgc)" && \
 593    [ $lastreceivesecs -lt $lastgcsecs ]; then
 594         # We've run gc since we last received, so maybe we can skip,
 595         # check if not fork or fork and lastparentgc < lastgc
 596         if [ -n "$isfork" ]; then
 597                 if [ -n "$lastparentgcsecs" ] && \
 598                    [ $lastparentgcsecs -lt $lastgcsecs ]; then
 599                         # We've run gc since our parent ran gc so we can skip
 600                         skipgc=1
 601                 fi
 602         else
 603                 # We don't have any alternates (we're not a forK) so we can skip
 604                 skipgc=1
 605         fi
 606 fi
 607
 608 # Prevent any other simultaneous gc operations
 609 lock_gc
 610
 611 # At this point, if .allowgc exists, it's now crud to be removed
 612 rm -f .allowgc
 613
 614 # Ideally we would do this in post-receive, but that would mean duplicating the
 615 # logic so it's available in the chroot jail and that's highly undesirable
 616 # Instead, since the first gc will be triggered immediately following the first
 617 # push, we do the check here as it's quick and harmless if HEAD is already valid
 618 check_and_set_head || :
 619
 620 # Always get rid of crud
 621 remove_crud
 622
 623 # Always perform reflogs maintenance
 624 coalesce_reflogs
 625 prune_reflogs
 626 compact_reflogs
 627
 628 # Run 'git svn gc' now for svn mirrors
 629 if [ -n "$svn_mirror" ]; then
 630         git svn gc || :
 631 fi
 632
 633 # Skip the actual gc if .delaygc is set
 634 if [ -e .delaygc ]; then
 635         progress "x [$proj] garbage check delayed (except for crud removal)"
 636         rm -f "$lockf"
 637         exit 0
 638 fi
 639
 640 # Do not skip gc if the repo is dirty
 641 if [ -n "$skipgc" ] && ! is_dirty; then
 642         progress "= [$proj] garbage check nothing but crud removal to do (`date`)"
 643         config_set lastgc "$gcstart"
 644         rm -f "$lockf"
 645         exit 0
 646 fi
 647
 648 bumptime=
 649 if [ -n "$isfork" ] && [ -z "$lastparentgcsecs" ]; then
 650         # set lastparentgc and then update gcstart to be at least 1 second later
 651         config_set lastparentgc "$gcstart"
 652         bumptime=1
 653 fi
 654 if [ -z "$lastreceivesecs" ]; then
 655         # set lastreceive and then update gcstart to be at least 1 second later
 656         config_set lastreceive "$gcstart"
 657         bumptime=1
 658 fi
 659 if [ -n "$bumptime" ]; then
 660         sleep 1
 661         gcstart="$(date "$datefmt")"
 662 fi
 663
 664 progress "+ [$proj] garbage check (`date`)"
 665
 666 newdeltas=
 667 [ -z "$alwaysredelta" ] || newdeltas=-f
 668 if [ -z "$newdeltas" ] && [ -n "$gfi_mirror" ]; then
 669         if [ $(list_packs --exclude-no-idx --count objects/pack) -le \
 670              $(list_packs --exclude-no-idx --count --quiet --only gfi-packs) ]; then
 671                 # Don't bother with repack_gfi_packs since everything's being repacked
 672                 newdeltas=-f
 673         fi
 674 fi
 675 if [ -z "$newdeltas" ] && [ -n "$noreusedeltaopt" ] && \
 676    [ $(list_packs --exclude-no-idx --count-objects objects/pack) -le $var_redelta_threshold ]; then
 677         # There aren't enough objects to worry about so just redelta to get the best pack
 678         newdeltas=-f
 679 fi
 680 if [ -z "$newdeltas" ]; then
 681         # Since we're not going to recompute deltas overall, we need to do the
 682         # "mini" maintenance so that we can get more optimal deltas
 683         [ -z "$noreusedeltaopt" ] || make_svn_pack
 684         repack_gfi_packs
 685         force_single_pack_redelta=
 686         [ -n "$gfi_mirror" ] || [ -n "$svn_mirror" ] || force_single_pack_redelta=1
 687         [ -z "$noreusedeltaopt" ] || combine_small_packs $force_single_pack_redelta
 688 fi
 689
 690 #
 691 ## Safe Pruning In Forks
 692 ##
 693 ## We are about to perform garbage collection.  We do NOT use the "git gc"
 694 ## command directly as it does not provide enough control over the fine details
 695 ## that we require.  However, we DO maintain a "gc.pid" file during our garbage
 696 ## collection so that a simultaneous "git gc" by an administrator will be
 697 ## blocked (and similarly we refuse to start garbage collection if we cannot
 698 ## create the "gc.pid" file).  When we say "gc" in the below description we are
 699 ## referring to our "gc.sh" script, NOT the "git gc" command.
 700 ##
 701 ## If the project we are running garbage collection (gc) on has any forks we
 702 ## must be careful not to remove any objects that while no longer referenced by
 703 ## this project (the parent) are still referenced by one or more forks (the
 704 ## children) otherwise the children will become corrupt and we can't abide
 705 ## corrupt children.
 706 ##
 707 ## One way to accomplish this is to simply hard-link all currently existing
 708 ## loose objects and packs in the parent into all the children that refer to the
 709 ## parent (via a line in their objects/info/alternates file) before beginning
 710 ## the gc operation and then relying on a subsequent gc in the child to clean up
 711 ## any excess objects/packs.  We used to use this strategy but it's very
 712 ## inefficient because:
 713 ##
 714 ##   1. The disk space used by the old pack(s)/object(s) will not be reclaimed
 715 ##      until all children (and their children, if any) run gc by which time
 716 ##      it's quite possible the topmost parent will have run gc again and
 717 ##      hard-linked yet another old pack down to its children (not to mention
 718 ##      loose objects).
 719 ##
 720 ##   2. As we are now using the "-A" option with "git repack", any new objects
 721 ##      in the parent that are not referenced by children will continually get
 722 ##      exploded out of the hard-linked pack in the children whenever the
 723 ##      children run gc.
 724 ##
 725 ##   3. To avoid suboptimal and/or unnecessarily many packs being hard-linked
 726 ##      into child forks, we must run the "mini" gc maintenance before we
 727 ##      perform the hard-linking into the children which provides yet another
 728 ##      source of inefficiency.
 729 ##
 730 ## Since we are using the "-A" option to "git repack" (that was not always the
 731 ## case) to guarantee we can access old ref values for long enough to send out
 732 ## a meaningful mail.sh notification, we now have another, more efficient,
 733 ## option available to prevent corruption of child forks that continue to refer
 734 ## to objects that are no longer reachable from any ref in the parent.
 735 ##
 736 ## The only things that need be copied (or hard-linked) into the child fork(s)
 737 ## are those objects that have become unreachable from any ref in the parent.
 738 ## They are the only things that could ever be removed by "git prune" and
 739 ## therefore the only things we need to prevent the loss of in order to avoid
 740 ## corruption of the child fork(s).
 741 ##
 742 ## Therefore we now use the following strategy instead to avoid excessive disk
 743 ## use and lots of unnecessary loose objects in child forks:
 744 ##
 745 ##   1. Run "git repack -A -d -l" in the parent BEFORE doing anything about
 746 ##      child forks.
 747 ##
 748 ##   2. Hard-link all remaining existing loose objects in the parent into the
 749 ##      immediate child forks.
 750 ##
 751 ##   3. Now run "git prune" in the parent.
 752 ##
 753 ## With this new strategy we avoid the need to run any "mini" gc maintenance
 754 ## before copying (or hard-linking) anything down to the child forks.
 755 ## Furthermore, only when the parent performs a non-fast-forward update will
 756 ## anything ever be transferred to the children leaving them unperturbed in the
 757 ## vast majority of cases.  Finally, even if the parent references objects the
 758 ## children do not, those objects will no longer continually end up in the
 759 ## children as unreachable loose objects after the children run gc.
 760 #
 761
 762 git pack-refs --all
 763 make_repack_dir
 764 touch .gc_in_progress
 765 rm -f .gc_failed bundles/*
 766 rm -f objects/pack/pack-*.bndl
 767 # We use the -A option with git repack so that unreachable objects can live
 768 # on for a time as loose objects.  This is particularly helpful if we just
 769 # happen to be in the process of sending out a ref update for a ref that was
 770 # force updated and the old ref value would have otherwise been removed by
 771 # repack because it was now unreachable.  Admittedly the window for gc to run
 772 # and do that before we manage to send out the ref update is not large, but
 773 # it would not be difficult to create such a situation.  Unfortunately, when
 774 # Git unpacks these unreachable objects it will give them the modification
 775 # time of the *.pack file they came out of.  This could be very, very old.
 776 # If that happens, the subsequent git prune --expire some_time_ago will still
 777 # remove the object(s) and our pending ref update will still lose out.
 778 # To prevent this from happening and to get the behavior we want, we now
 779 # touch the modification time of all pack-<sha>.pack files so that any
 780 # loosened objects get a current time.  Git does not provide any other
 781 # mechanism to do this.  We do not want to just touch all loose objects
 782 # left after the repack because that would cause objects that were loosened
 783 # previously to live on which we definitely do not want.
 784 list_packs --exclude-no-idx objects/pack | xargs touch -c 2>/dev/null || :
 785 # We wish to keep deltas from our last full pack so if we're not redeltaing
 786 # then make sure the .pack associated with the .bitmap has a newer mod time
 787 # (If there is no .bitmap then touch the pack with the most objects instead.)
 788 if [ -z "$newdeltas" ]; then
 789         bmpack="$(list_packs --exclude-no-bitmap --exclude-no-idx --max-matches 1 objects/pack)"
 790         [ -n "$bmpack" ] || bmpack="$(list_packs --exclude-no-idx --max-matches 1 --object-limit -1 --include-boundary objects/pack)"
 791         if [ -n "$bmpack" ] && [ -f "$bmpack" -a -s "$bmpack" ]; then
 792                 sleep 1
 793                 touch -c "$bmpack" 2>/dev/null || :
 794         fi
 795 fi
 796 # The git repack command only supports bitmaps if all objects are being packed.
 797 # While it is theoretically possible that a project with a non-empty alternates
 798 # file ends up packing all objects (because it does not actually use any of the
 799 # objects found in the alternates), it's very unlikely.  And, in the unlikely
 800 # event that did occur, clients would see a message about only using one bitmap
 801 # because Git can only use one bitmap at a time and at least one of the
 802 # alternates is bound to have a bitmap.  Therefore if we see a non-empty
 803 # alternates file, we disable writing bitmaps which avoids the warning and any
 804 # possibility of a client warning as well.
 805 nobm=
 806 [ -z "$var_have_git_172" ] || ! [ -s objects/info/alternates ] || \
 807         nobm='-c repack.writebitmaps=false -c pack.writebitmaps=false'
 808 progress "~ [$proj] running full gc repack${nobm:+ (bitmaps disabled)}"
 809 cd repack
 810 # We run git repack from the repack subdirectory so we can force optimized packs
 811 # to be generated even for repositories that do not have any tagged commits
 812 git $nobm repack $packopts -A -d -l $quiet $newdeltas $@
 813 cd ..
 814 rm -rf repack
 815 [ ! -e .gc_failed ] || exit 1
 816 # These, if they exist, are now meaningless and need to be removed
 817 rm -f gfi-packs .needsgc .svnpack .svnpackgc
 818 allpacks="$(echo objects/pack/pack-$octet20.pack)"
 819 curhead="$(cat HEAD)"
 820 pkrf=
 821 [ ! -e packed-refs ] || pkrf=packed-refs
 822 eval "reposizek=$(( $(echo 0 $(du -k $pkrf $allpacks 2>/dev/null | LC_ALL=C awk '{print $1}') | \
 823         LC_ALL=C sed -e 's/ / + /g') ))"
 824 git update-server-info
 825 # The -A option to `git repack` may have caused some loose objects to pop
 826 # out of their packs.  We must make these objects group writable so that they
 827 # can be freshened by other pushers.  Technically we need only do this for
 828 # push projects but to enable mirror projects to be more easily converted to
 829 # push projects, we go ahead and do it for all projects.
 830 { find objects/$octet -type f -name "$octet19" -print0 | xargs -0 chmod ug+w || :; } 2>/dev/null
 831
 832 if has_forks "$proj"; then
 833         progress "~ [$proj] hard-linking loose objects into immediate child forks"
 834         # We have to update the lastparentgc time in the child forks even if they do not get any
 835         # new "loose objects" because they need to run gc just in case the parent now has some
 836         # objects that used to only be in the child so they can be removed from the child.
 837         # For example, a "patch" might be developed first in a fork and then later accepted into
 838         # the parent in which case the objects making up the patch in the child fork are now
 839         # redundant (since they're now in the parent as well) and need to be removed from the
 840         # child fork which can only happen if the child fork runs gc.
 841         forkdir="$proj"
 842         # It is enough to copy objects just one level down and get_repo_list
 843         # takes a regular expression (which is automatically prefixed with '^')
 844         # so we can easily match forks exactly one level down from this project
 845         get_repo_list "$forkdir/[^/]*:" |
 846                 while read fork; do
 847                         # Ignore forks that do not exist or are symbolic links
 848                         [ ! -L "$cfg_reporoot/$fork.git" -a -d "$cfg_reporoot/$fork.git" ] || \
 849                                 continue
 850                         # Or do not have a non-zero length alternates file
 851                         [ -s "$cfg_reporoot/$fork.git/objects/info/alternates" ] || \
 852                                 continue
 853                         # Match objects in parent project
 854                         for d in objects/$octet; do
 855                                 [ "$d" != "objects/$octet" ] || continue
 856                                 mkdir -p "$cfg_reporoot/$fork.git/$d"
 857                                 find "$d" -maxdepth 1 -type f -name "$octet19" -print0 |
 858                                         xargs -0 "$var_sh_bin" -c 'ln -f "$@" '"'$cfg_reporoot/$fork.git/$d/'" sh || :
 859                         done
 860                         # Update the fork's lastparentgc date (must be current, not $gcstart)
 861                         git --git-dir="$cfg_reporoot/$fork.git" config \
 862                                 gitweb.lastparentgc "$(date "$datefmt")"
 863                 done
 864 fi
 865
 866 # The git prune command does not take a -q or --quiet but started outputting
 867 # 'Checking connectivity' progress messages in v1.7.9.  However, we can
 868 # suppress those by piping through cat as it only activates the progress
 869 # messages when stderr is a tty.  We only expire loose objects older than one
 870 # day just in case there's some pending action (such as sending out a ref
 871 # update) in progress that might want to examine them.  This may leave us with
 872 # loose objects.  That's okay because at the next gc interval, we will always
 873 # run gc if we see any loose objects regardless of whether or not we've seen
 874 # any updates or we've received new linked objects from our parent.  Note that
 875 # in order to keep loose objects that just recently became unreferenced but
 876 # have a very old modification date around we rely on some help from both the
 877 # update.sh and hooks/pre-receive scripts.  Furthermore, since Git v2.2.0
 878 # (d3038d22 prune: keep objects reachable from recent objects) an unreachable
 879 # object that would otherwise be pruned (because it's too old) will be kept
 880 # alive by an unreachable object that refers to it that's not old enough to
 881 # be pruned yet.
 882 prunecmd='git prune --expire 1_day_ago'
 883 [ -n "$show_progress" ] || \
 884 prunecmd="{ $prunecmd 2>&1 || touch .gc_failed; } | cat"
 885 progress "~ [$proj] pruning expired unreachable loose objects"
 886 eval "$prunecmd"
 887 [ ! -e .gc_failed ] || exit 1
 888
 889 # darcs:// mirrors have a xxx.log file that will grow endlessly
 890 # if this is a mirror and the file exists, shorten it to 10000 lines
 891 # also take this opportunity to optimize the darcs repo
 892 if [ ! -e .nofetch ] && [ -n "$cfg_mirror" ]; then
 893         url="$(config_get baseurl || :)"
 894         case "$url" in darcs://*)
 895                 if [ -n "$cfg_mirror_darcs" ]; then
 896                         url="${url%/}"
 897                         basedarcs="$(basename "${url#darcs:/}")"
 898                         if [ -f "$basedarcs.log" ]; then
 899                                 tail -n 10000 "$basedarcs.log" > "$basedarcs.log.$$"
 900                                 mv -f "$basedarcs.log.$$" "$basedarcs.log"
 901                         fi
 902                         if [ -d "$basedarcs.darcs" ]; then
 903                                 (
 904                                         cd "$basedarcs.darcs"
 905                                         # without show_progress suppress non-error output
 906                                         [ -n "$show_progress" ] || exec >/dev/null
 907                                         # Note that this does not optimize _darcs/inventories/ :(
 908                                         darcs optimize || :
 909                                 )
 910                         fi
 911                 fi
 912         esac
 913 fi
 914
 915 # Create a matching .bndl header file for the all-in-one pack we just created
 916 # but only if we're not a fork (otherwise the bundle would not be complete)
 917 # and we are running at least Git version 1.7.2 (pack_is_complete always fails otherwise)
 918 if [ ! -s objects/info/alternates ] && [ -n "$var_have_git_172" ]; then
 919         # There should only be one pack in $allpacks but if there was a
 920         # simultaneous push...
 921         # The one we just created will have a .idx and will NOT have a .keep
 922         progress "~ [$proj] creating downloadble bundle header"
 923         pkfound=
 924         pkhead=
 925         for pk in $allpacks; do
 926                 [ -s "$pk" ] || continue
 927                 pkbase="${pk%.pack}"
 928                 [ -s "$pkbase.idx" ] || continue
 929                 [ ! -e "$pkbase.keep" ] || continue
 930                 if pkhead="$(pack_is_complete "$PWD/$pk" "$PWD/packed-refs" "$curhead")"; then
 931                         pkfound="$pkbase"
 932                         break;
 933                 fi
 934         done
 935         if [ -n "$pkfound" -a -n "$pkhead" ]; then
 936                 {
 937                         echo "# v2 git bundle"
 938                         LC_ALL=C sed -ne "/^$octet20 refs\/[^ $tab]*\$/ p" < packed-refs
 939                         echo "$pkhead HEAD"
 940                         echo ""
 941                 } > "$pkbase.bndl"
 942                 bndletag="$("$cfg_basedir/bin/rangecgi" --etag -m 1 "$pkbase.bndl" "$pkbase.pack" || :)"
 943                 bndlsha="$(printf '%s' "$bndletag" | git hash-object --stdin || :)"
 944                 if [ -n "$bndletag" ]; then
 945                         case "$bndlsha" in $octet20)
 946                                 bndlshatrailer="${bndlsha#????????}"
 947                                 bndlshaprefix="${bndlsha%$bndlshatrailer}"
 948                                 bndlname="$(TZ=UTC date +%Y%m%d_%H%M%S)-${bndlshaprefix:-0}"
 949                                 [ -d bundles ] || mkdir bundles
 950                                 echo "${pkbase#objects/pack/}.bndl" > "bundles/$bndlname"
 951                                 echo "${pkbase#objects/pack/}.pack" >> "bundles/$bndlname"
 952                                 ln -s -f -n "$bndlname" bundles/latest
 953                         esac
 954                 fi
 955         fi
 956 fi
 957
 958 # Record the size of this repo as the sum of its *.pack sizes as 1024-byte blocks
 959 config_set_raw girocco.reposizek "${reposizek:-0}"
 960
 961 # We use $gcstart here to avoid a race where a push occurs during the gc itself
 962 # and the next future gc could be incorrectly skipped if we used the current
 963 # timestamp here instead
 964 config_set lastgc "$gcstart"
 965 rm -f "$lockf"
 966
 967 progress "- [$proj] garbage check (`date`)"