xfs: map multiple bmbt records to a single rmapbt record

Make sure that we can handle multiple bmbt records mapping to a
single rmapbt record.  This can happen if you fallocate more than
2^21 contiguous blocks to a file.

(Also add some helpers that can create huge devices with some
dm-zero and dm-snapshot fakery.)

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Eryu Guan <eguan@redhat.com>

common/dmhugedisk

@@ -0,0 +1,60 @@
+##/bin/bash
+# Routines for creating huge (fake) disks
+#-----------------------------------------------------------------------
+#  Copyright (c) 2016 Oracle.  All Rights Reserved.
+#  This program is free software; you can redistribute it and/or modify
+#  it under the terms of the GNU General Public License as published by
+#  the Free Software Foundation; either version 2 of the License, or
+#  (at your option) any later version.
+#
+#  This program is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+#  GNU General Public License for more details.
+#
+#  You should have received a copy of the GNU General Public License
+#  along with this program; if not, write to the Free Software
+#  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307
+#  USA
+#
+#  Contact information: Oracle Corporation, 500 Oracle Parkway,
+#  Redwood Shores, CA 94065, USA, or: http://www.oracle.com/
+#-----------------------------------------------------------------------
+
+_require_dmhugedisk()
+{
+	_require_dm_target zero
+	_require_dm_target snapshot
+}
+
+_dmhugedisk_init()
+{
+	test -z "$1" && _fatal "must specify sector count to _dmhugedisk_init"
+	local dm_backing_dev=$SCRATCH_DEV
+
+	$DMSETUP_PROG remove huge-test > /dev/null 2>&1
+	$DMSETUP_PROG remove huge-test-zero > /dev/null 2>&1
+
+	local blk_dev_size=$1
+
+	DMHUGEDISK_ZERO='/dev/mapper/huge-test-zero'
+	DMHUGEDISK_DEV='/dev/mapper/huge-test'
+
+	DMHUGEDISK_ZERO_TABLE="0 $blk_dev_size zero"
+	DMHUGEDISK_DEV_TABLE="0 $blk_dev_size snapshot $DMHUGEDISK_ZERO $SCRATCH_DEV N 512"
+
+	$DMSETUP_PROG create huge-test-zero --table "$DMHUGEDISK_ZERO_TABLE" || \
+		_fatal "failed to create dm huge zero device"
+	$DMSETUP_PROG create huge-test --table "$DMHUGEDISK_DEV_TABLE" || \
+		_fatal "failed to create dm huge device"
+}
+
+_dmhugedisk_cleanup()
+{
+	$UMOUNT_PROG $SCRATCH_MNT > /dev/null 2>&1
+	# wait for device to be fully settled so that 'dmsetup remove' doesn't
+	# fail due to EBUSY
+	$UDEV_SETTLE_PROG >/dev/null 2>&1
+	$DMSETUP_PROG remove huge-test > /dev/null 2>&1
+	$DMSETUP_PROG remove huge-test-zero > /dev/null 2>&1
+}

tests/xfs/310

@@ -0,0 +1,119 @@
+#! /bin/bash
+# FS QA Test No. 310
+#
+# Create a file with more than 2^21 extents (the max length of a bmbt record).
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2016, Oracle and/or its affiliates.  All Rights Reserved.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation.
+#
+# This program is distributed in the hope that it would be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write the Free Software Foundation,
+# Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+#-----------------------------------------------------------------------
+
+seq=`basename $0`
+seqres=$RESULT_DIR/$seq
+echo "QA output created by $seq"
+
+here=`pwd`
+tmp=/tmp/$$
+status=1    # failure is the default!
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+_cleanup()
+{
+	cd /
+	umount $SCRATCH_MNT > /dev/null 2>&1
+	_dmhugedisk_cleanup
+	rm -rf $tmp.*
+	_scratch_mkfs >/dev/null 2>&1
+}
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/filter
+. ./common/dmhugedisk
+
+# real QA test starts here
+_supported_os Linux
+_supported_fs xfs
+_require_scratch
+_require_xfs_io_command "falloc"
+
+rm -f $seqres.full
+
+# Figure out block size
+echo "Figure out block size"
+_scratch_mkfs >/dev/null 2>&1
+_scratch_mount >> $seqres.full
+
+is_rmap=$(xfs_info $SCRATCH_MNT | grep -c "rmapbt=1")
+test $is_rmap -gt 0 || _notrun "rmap not supported on scratch fs"
+
+testdir=$SCRATCH_MNT/test-$seq
+blksz="$(stat -f $SCRATCH_MNT -c '%S')"
+
+umount $SCRATCH_MNT
+
+echo "Format huge device"
+_dmhugedisk_init $((blksz * 2 * 4400))	# a little over 2^22 blocks
+_mkfs_dev -d agcount=2 $DMHUGEDISK_DEV
+_mount $DMHUGEDISK_DEV $SCRATCH_MNT
+xfs_info $SCRATCH_MNT >> $seqres.full
+
+echo "Create the original file blocks"
+mkdir $testdir
+blksz="$(stat -f $testdir -c '%S')"
+nr_blks=2100000	# 2^21 plus a little more
+$XFS_IO_PROG -f -c "falloc 0 $((nr_blks * blksz))" $testdir/file1 >> $seqres.full
+
+echo "Check extent count"
+xfs_bmap -l -p -v $testdir/file1 | grep '^[[:space:]]*1:' -q && xfs_bmap -l -p -v $testdir/file1
+inum=$(stat -c '%i' $testdir/file1)
+umount $SCRATCH_MNT
+
+echo "Check bmap count"
+nr_bmaps=$(xfs_db -c "inode $inum" -c "bmap" $DMHUGEDISK_DEV | grep 'data offset' | wc -l)
+test $nr_bmaps -gt 1 || xfs_db -c "inode $inum" -c "bmap" $DMHUGEDISK_DEV
+#xfs_db -c "agf 0" -c p -c "inode $inum" -c "bmap" $DMHUGEDISK_DEV
+
+echo "Check rmap count"
+nr_rmaps=$(xfs_db -c 'agf 0' -c 'addr rmaproot' -c 'p' $DMHUGEDISK_DEV | grep ",$inum,[0-9]*,1,0,0" | wc -l)
+test $nr_rmaps -eq 1 || xfs_db -c 'agf 0' -c 'addr rmaproot' -c 'p' $DMHUGEDISK_DEV | grep ",$inum,[0-9]*,1,0,0"
+
+echo "Check and fake-repair huge filesystem" | tee -a $seqres.full
+$XFS_DB_PROG -c 'check' $DMHUGEDISK_DEV
+$XFS_REPAIR_PROG -n $DMHUGEDISK_DEV >> $seqres.full 2>&1
+test $? -eq 0 || echo "xfs_repair -n failed, see $seqres.full"
+
+echo "Real repair huge filesystem" | tee -a $seqres.full
+$XFS_REPAIR_PROG $DMHUGEDISK_DEV >> $seqres.full 2>&1
+test $? -eq 0 || echo "xfs_repair failed, see $seqres.full"
+
+echo "Check bmap count again"
+nr_bmaps=$(xfs_db -c "inode $inum" -c "bmap" $DMHUGEDISK_DEV | grep 'data offset' | wc -l)
+test $nr_bmaps -gt 1 || xfs_db -c "inode $inum" -c "bmap" $DMHUGEDISK_DEV
+
+echo "Check rmap count again"
+nr_rmaps=$(xfs_db -c 'agf 0' -c 'addr rmaproot' -c 'p' $DMHUGEDISK_DEV | grep ",$inum,[0-9]*,1,0,0" | wc -l)
+test $nr_rmaps -eq 1 || xfs_db -c 'agf 0' -c 'addr rmaproot' -c 'p' $DMHUGEDISK_DEV | grep ",$inum,[0-9]*,1,0,0"
+
+echo "Check and fake-repair huge filesystem again" | tee -a $seqres.full
+$XFS_DB_PROG -c 'check' $DMHUGEDISK_DEV
+$XFS_REPAIR_PROG -n $DMHUGEDISK_DEV >> $seqres.full 2>&1
+_dmhugedisk_cleanup
+
+echo "Done"
+
+# success, all done
+status=0
+exit

tests/xfs/310.out

@@ -0,0 +1,13 @@
+QA output created by 310
+Figure out block size
+Format huge device
+Create the original file blocks
+Check extent count
+Check bmap count
+Check rmap count
+Check and fake-repair huge filesystem
+Real repair huge filesystem
+Check bmap count again
+Check rmap count again
+Check and fake-repair huge filesystem again
+Done

tests/xfs/group

@@ -289,3 +289,4 @@
 307 auto quick clone
 308 auto quick clone
 309 auto clone
+310 auto clone rmap