tests/generic/038

#! /bin/bash
# SPDX-License-Identifier: GPL-2.0
# Copyright (C) 2014 SUSE Linux Products GmbH. All Rights Reserved.
#
# FSQA Test No. 038
#
# This test was motivated by btrfs issues, but it's generic enough as it
# doesn't use any btrfs specific features.
#
# Stress btrfs' block group allocation and deallocation while running fstrim in
# parallel. Part of the goal is also to get data block groups deallocated so
# that new metadata block groups, using the same physical device space ranges,
# get allocated while fstrim is running. This caused several issues ranging
# from invalid memory accesses, kernel crashes, metadata or data corruption,
# free space cache inconsistencies, free space leaks and memory leaks.
#
# These issues were fixed by the following btrfs linux kernel patches:
#
#   Btrfs: fix invalid block group rbtree access after bg is removed
#   Btrfs: fix crash caused by block group removal
#   Btrfs: fix freeing used extents after removing empty block group
#   Btrfs: fix race between fs trimming and block group remove/allocation
#   Btrfs: fix race between writing free space cache and trimming
#   Btrfs: make btrfs_abort_transaction consider existence of new block groups
#   Btrfs: fix memory leak after block remove + trimming
#   Btrfs: fix fs mapping extent map leak
#   Btrfs: fix unprotected deletion from pending_chunks list
#
# The issues were found on a qemu/kvm guest with 4 virtual CPUs, 4Gb of ram and
# scsi-hd devices with discard support enabled (that means hole punching in the
# disk's image file is performed by the host).
#
. ./common/preamble
_begin_fstest auto stress trim prealloc

# Override the default cleanup function.
_cleanup()
{
	[ -n "${create_pids}" ] && kill ${create_pids[@]}
	[ -n "${fallocate_pids}" ] && kill ${fallocate_pids[@]}
	[ -n "${trim_pids}" ] && kill ${trim_pids[@]}
	wait
	rm -fr $tmp
}

# Import common functions.
. ./common/filter

_require_scratch
_require_xfs_io_command "falloc"

echo "Silence is golden"

# Keep allocating and deallocating 1G of data space with the goal of creating
# and deleting 1 block group constantly. The intention is to race with the
# fstrim loop below.
fallocate_loop()
{
	# Wait for running subcommand before exitting so that
	# mountpoint is not busy when we try to unmount it
	trap "wait; exit" SIGTERM

	local name=$1
	while true; do
		$XFS_IO_PROG -f -c "falloc -k 0 1G" \
			$SCRATCH_MNT/$name &> /dev/null
		sleep 3
		$XFS_IO_PROG -c "truncate 0" \
			$SCRATCH_MNT/$name &> /dev/null
		sleep 3
	done
}

trim_loop()
{
	# Wait for running subcommand before exitting so that
	# mountpoint is not busy when we try to unmount it
	trap "wait; exit" SIGTERM

	while true; do
		$FSTRIM_PROG $SCRATCH_MNT
	done
}

# Create a bunch of small files that get their single extent inlined in the
# btree, so that we consume a lot of metadata space and get a chance of a
# data block group getting deleted and reused for metadata later. Sometimes
# the creation of all these files succeeds other times we get ENOSPC failures
# at some point - this depends on how fast the btrfs' cleaner kthread is
# notified about empty block groups, how fast it deletes them and how fast
# the fallocate calls happen. So we don't really care if they all succeed or
# not, the goal is just to keep metadata space usage growing while data block
# groups are deleted.
#
# Creating 200,000 files sequentially is really slow, so speed it up a bit
# by doing it concurrently with 4 threads in 4 separate directories.
nr_files=$((50000 * LOAD_FACTOR))
create_files()
{
	local prefix=$1
	local now=$(date '+%s')
	local end_time=$(( now + (TIME_FACTOR * 30) ))

	for ((n = 0; n < 4; n++)); do
		mkdir $SCRATCH_MNT/$n
		(
		trap "wait; exit" SIGTERM

		for ((i = 1; i <= $nr_files; i++)); do
			$XFS_IO_PROG -f -c "pwrite -S 0xaa 0 3900" \
				$SCRATCH_MNT/$n/"${prefix}_$i" &> /dev/null
			if [ $? -ne 0 ]; then
				echo "Failed creating file $n/${prefix}_$i" >>$seqres.full
				break
			fi
			if [ "$(date '+%s')" -ge $end_time ]; then
				echo "runtime exceeded @ $i files" >> $seqres.full
				break
			fi
		done
		) &
		create_pids[$n]=$!
	done

	wait ${create_pids[@]}
	unset create_pids
}

_scratch_mkfs >>$seqres.full 2>&1
_scratch_mount
_require_fs_space $SCRATCH_MNT $((10 * 1024 * 1024))
_require_batched_discard $SCRATCH_MNT

for ((i = 0; i < $((4 * $LOAD_FACTOR)); i++)); do
	trim_loop &
	trim_pids[$i]=$!
done

for ((i = 0; i < $((1 * $LOAD_FACTOR)); i++)); do
	fallocate_loop "falloc_file_$i" &
	fallocate_pids[$i]=$!
done

create_files "foobar"

kill ${fallocate_pids[@]}
kill ${trim_pids[@]}
wait
unset fallocate_pids
unset trim_pids

status=0
exit