Revision a742994aa2e271eb8cd8e043d276515ec858ed73 authored by Filipe Manana on 13 February 2015, 16:56:14 UTC, committed by Chris Mason on 14 February 2015, 16:22:49 UTC
If we are recording in the tree log that an inode has new names (new hard
links were added), we would drop items, belonging to the inode, that we
shouldn't:

1) When the flag BTRFS_INODE_COPY_EVERYTHING is set in the inode's runtime
   flags, we ended up dropping all the extent and xattr items that were
   previously logged. This was done only in memory, since logging a new
   name doesn't imply syncing the log;

2) When the flag BTRFS_INODE_COPY_EVERYTHING is set in the inode's runtime
   flags, we ended up dropping all the xattr items that were previously
   logged. Like the case before, this was done only in memory because
   logging a new name doesn't imply syncing the log.

This led to some surprises in scenarios such as the following:

1) write some extents to an inode;
2) fsync the inode;
3) truncate the inode or delete/modify some of its xattrs
4) add a new hard link for that inode
5) fsync some other file, to force the log tree to be durably persisted
6) power failure happens

The next time the fs is mounted, the fsync log replay code is executed,
and the resulting file doesn't have the content it had when the last fsync
against it was performed, instead if has a content matching what it had
when the last transaction commit happened.

So change the behaviour such that when a new name is logged, only the inode
item and reference items are processed.

This is easy to reproduce with the test I just made for xfstests, whose
main body is:

  _scratch_mkfs >> $seqres.full 2>&1
  _init_flakey
  _mount_flakey

  # Create our test file with some data.
  $XFS_IO_PROG -f -c "pwrite -S 0xaa -b 8K 0 8K" \
      $SCRATCH_MNT/foo | _filter_xfs_io

  # Make sure the file is durably persisted.
  sync

  # Append some data to our file, to increase its size.
  $XFS_IO_PROG -f -c "pwrite -S 0xcc -b 4K 8K 4K" \
      $SCRATCH_MNT/foo | _filter_xfs_io

  # Fsync the file, so from this point on if a crash/power failure happens, our
  # new data is guaranteed to be there next time the fs is mounted.
  $XFS_IO_PROG -c "fsync" $SCRATCH_MNT/foo

  # Now shrink our file to 5000 bytes.
  $XFS_IO_PROG -c "truncate 5000" $SCRATCH_MNT/foo

  # Now do an expanding truncate to a size larger than what we had when we last
  # fsync'ed our file. This is just to verify that after power failure and
  # replaying the fsync log, our file matches what it was when we last fsync'ed
  # it - 12Kb size, first 8Kb of data had a value of 0xaa and the last 4Kb of
  # data had a value of 0xcc.
  $XFS_IO_PROG -c "truncate 32K" $SCRATCH_MNT/foo

  # Add one hard link to our file. This made btrfs drop all of our file's
  # metadata from the fsync log, including the metadata relative to the
  # extent we just wrote and fsync'ed. This change was made only to the fsync
  # log in memory, so adding the hard link alone doesn't change the persisted
  # fsync log. This happened because the previous truncates set the runtime
  # flag BTRFS_INODE_NEEDS_FULL_SYNC in the btrfs inode structure.
  ln $SCRATCH_MNT/foo $SCRATCH_MNT/foo_link

  # Now make sure the in memory fsync log is durably persisted.
  # Creating and fsync'ing another file will do it.
  # After this our persisted fsync log will no longer have metadata for our file
  # foo that points to the extent we wrote and fsync'ed before.
  touch $SCRATCH_MNT/bar
  $XFS_IO_PROG -c "fsync" $SCRATCH_MNT/bar

  # As expected, before the crash/power failure, we should be able to see a file
  # with a size of 32Kb, with its first 5000 bytes having the value 0xaa and all
  # the remaining bytes with value 0x00.
  echo "File content before:"
  od -t x1 $SCRATCH_MNT/foo

  # Simulate a crash/power loss.
  _load_flakey_table $FLAKEY_DROP_WRITES
  _unmount_flakey

  _load_flakey_table $FLAKEY_ALLOW_WRITES
  _mount_flakey

  # After mounting the fs again, the fsync log was replayed.
  # The expected result is to see a file with a size of 12Kb, with its first 8Kb
  # of data having the value 0xaa and its last 4Kb of data having a value of 0xcc.
  # The btrfs bug used to leave the file as it used te be as of the last
  # transaction commit - that is, with a size of 8Kb with all bytes having a
  # value of 0xaa.
  echo "File content after:"
  od -t x1 $SCRATCH_MNT/foo

The test case for xfstests follows soon.

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
1 parent 1a4bcf4
Raw File
config
#!/bin/bash
# Manipulate options in a .config file from the command line

myname=${0##*/}

# If no prefix forced, use the default CONFIG_
CONFIG_="${CONFIG_-CONFIG_}"

usage() {
	cat >&2 <<EOL
Manipulate options in a .config file from the command line.
Usage:
$myname options command ...
commands:
	--enable|-e option   Enable option
	--disable|-d option  Disable option
	--module|-m option   Turn option into a module
	--set-str option string
	                     Set option to "string"
	--set-val option value
	                     Set option to value
	--undefine|-u option Undefine option
	--state|-s option    Print state of option (n,y,m,undef)

	--enable-after|-E beforeopt option
                             Enable option directly after other option
	--disable-after|-D beforeopt option
                             Disable option directly after other option
	--module-after|-M beforeopt option
                             Turn option into module directly after other option

	commands can be repeated multiple times

options:
	--file config-file   .config file to change (default .config)
	--keep-case|-k       Keep next symbols' case (dont' upper-case it)

$myname doesn't check the validity of the .config file. This is done at next
make time.

By default, $myname will upper-case the given symbol. Use --keep-case to keep
the case of all following symbols unchanged.

$myname uses 'CONFIG_' as the default symbol prefix. Set the environment
variable CONFIG_ to the prefix to use. Eg.: CONFIG_="FOO_" $myname ...
EOL
	exit 1
}

checkarg() {
	ARG="$1"
	if [ "$ARG" = "" ] ; then
		usage
	fi
	case "$ARG" in
	${CONFIG_}*)
		ARG="${ARG/${CONFIG_}/}"
		;;
	esac
	if [ "$MUNGE_CASE" = "yes" ] ; then
		ARG="`echo $ARG | tr a-z A-Z`"
	fi
}

txt_append() {
	local anchor="$1"
	local insert="$2"
	local infile="$3"
	local tmpfile="$infile.swp"

	# sed append cmd: 'a\' + newline + text + newline
	cmd="$(printf "a\\%b$insert" "\n")"

	sed -e "/$anchor/$cmd" "$infile" >"$tmpfile"
	# replace original file with the edited one
	mv "$tmpfile" "$infile"
}

txt_subst() {
	local before="$1"
	local after="$2"
	local infile="$3"
	local tmpfile="$infile.swp"

	sed -e "s:$before:$after:" "$infile" >"$tmpfile"
	# replace original file with the edited one
	mv "$tmpfile" "$infile"
}

txt_delete() {
	local text="$1"
	local infile="$2"
	local tmpfile="$infile.swp"

	sed -e "/$text/d" "$infile" >"$tmpfile"
	# replace original file with the edited one
	mv "$tmpfile" "$infile"
}

set_var() {
	local name=$1 new=$2 before=$3

	name_re="^($name=|# $name is not set)"
	before_re="^($before=|# $before is not set)"
	if test -n "$before" && grep -Eq "$before_re" "$FN"; then
		txt_append "^$before=" "$new" "$FN"
		txt_append "^# $before is not set" "$new" "$FN"
	elif grep -Eq "$name_re" "$FN"; then
		txt_subst "^$name=.*" "$new" "$FN"
		txt_subst "^# $name is not set" "$new" "$FN"
	else
		echo "$new" >>"$FN"
	fi
}

undef_var() {
	local name=$1

	txt_delete "^$name=" "$FN"
	txt_delete "^# $name is not set" "$FN"
}

if [ "$1" = "--file" ]; then
	FN="$2"
	if [ "$FN" = "" ] ; then
		usage
	fi
	shift 2
else
	FN=.config
fi

if [ "$1" = "" ] ; then
	usage
fi

MUNGE_CASE=yes
while [ "$1" != "" ] ; do
	CMD="$1"
	shift
	case "$CMD" in
	--keep-case|-k)
		MUNGE_CASE=no
		continue
		;;
	--refresh)
		;;
	--*-after|-E|-D|-M)
		checkarg "$1"
		A=$ARG
		checkarg "$2"
		B=$ARG
		shift 2
		;;
	-*)
		checkarg "$1"
		shift
		;;
	esac
	case "$CMD" in
	--enable|-e)
		set_var "${CONFIG_}$ARG" "${CONFIG_}$ARG=y"
		;;

	--disable|-d)
		set_var "${CONFIG_}$ARG" "# ${CONFIG_}$ARG is not set"
		;;

	--module|-m)
		set_var "${CONFIG_}$ARG" "${CONFIG_}$ARG=m"
		;;

	--set-str)
		# sed swallows one level of escaping, so we need double-escaping
		set_var "${CONFIG_}$ARG" "${CONFIG_}$ARG=\"${1//\"/\\\\\"}\""
		shift
		;;

	--set-val)
		set_var "${CONFIG_}$ARG" "${CONFIG_}$ARG=$1"
		shift
		;;
	--undefine|-u)
		undef_var "${CONFIG_}$ARG"
		;;

	--state|-s)
		if grep -q "# ${CONFIG_}$ARG is not set" $FN ; then
			echo n
		else
			V="$(grep "^${CONFIG_}$ARG=" $FN)"
			if [ $? != 0 ] ; then
				echo undef
			else
				V="${V/#${CONFIG_}$ARG=/}"
				V="${V/#\"/}"
				V="${V/%\"/}"
				V="${V//\\\"/\"}"
				echo "${V}"
			fi
		fi
		;;

	--enable-after|-E)
		set_var "${CONFIG_}$B" "${CONFIG_}$B=y" "${CONFIG_}$A"
		;;

	--disable-after|-D)
		set_var "${CONFIG_}$B" "# ${CONFIG_}$B is not set" "${CONFIG_}$A"
		;;

	--module-after|-M)
		set_var "${CONFIG_}$B" "${CONFIG_}$B=m" "${CONFIG_}$A"
		;;

	# undocumented because it ignores --file (fixme)
	--refresh)
		yes "" | make oldconfig
		;;

	*)
		usage
		;;
	esac
done
back to top