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