apfs/apfstests
0
0
Fork 0
mirror of https://github.com/linux-apfs/apfstests.git synced 2026-05-01 15:01:44 -07:00
Code Issues Packages Projects Releases Wiki Activity

xfstests: btrfs/011 improvement for compressed filesystems

Browse Source 
Josef noticed that using /dev/zero to generate most of the test
data doesn't work if someone overrides the mount options to
enable compression. The test that performs a cancellation failed
because the replace operation was already finished when the
cancel request was executed.

Since /dev/urandom is too slow to generate multiple GB, the
way how the filesystem data is generated is completely changed
with this patch. Now /dev/urandom is used to generate one 1MB
file and this file is copied up to 2048 times. /dev/zero is no
longer used.

The runtime of the test is about the same as before. Compression
works now, online duplication will again cause issues, but
we don't have online duplication today.

Reported-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Stefan Behrens <sbehrens@giantdisaster.de>
Reviewed-by: Jan Schmidt <list.xfs@jan-o-sch.net>
Signed-off-by: Rich Johnston <rjohnston@sgi.com>
This commit is contained in:
Stefan Behrens
2013-09-13 10:27:21 +00:00
committed by Rich Johnston
parent a4d5b247b5
commit e15d36e67c
1 changed files with 30 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/btrfs/011
+30 -16
View File
@@ -78,6 +78,7 @@ workout()
local quick="$4"
local source_dev="`echo ${SCRATCH_DEV_POOL} | awk '{print $1}'`"
local target_dev="`echo ${SCRATCH_DEV_POOL} | awk '{print $NF}'`"
local fssize
if [ "`echo $SCRATCH_DEV_POOL | wc -w`" -lt `expr $num_devs4raid + 1` ]; then
echo "Skip workout $1 $2 $3 $4" >> $seqres.full
@@ -107,33 +108,46 @@ workout()
_notrun "Different device sizes detected"
fi
if [ `$BTRFS_SHOW_SUPER_PROG $SCRATCH_DEV | grep dev_item.total_bytes | awk '{print $2}'` -lt 2500000000 ]; then
_notrun "device size too small"
fi
_scratch_mount
# Generate 500 times 20K extents in the data chunk and fill up
# metadata with inline extents. Ignore ENOSPC.
# Generate metadata and some minimal user data, generate 500 times
# 20K extents in the data chunk and fill up metadata with inline
# extents.
for i in `seq 1 500`; do
dd if=/dev/urandom of=$SCRATCH_MNT/l$i bs=16385 count=1
dd if=/dev/urandom of=$SCRATCH_MNT/s$i bs=3800 count=1
done > /dev/null 2>&1
# /dev/urandom is slow but has the benefit that the generated
# contents does not shrink during compression.
# Generate a template once and quickly copy it multiple times.
# Obviously with online deduplication this will not work anymore.
dd if=/dev/urandom of=$SCRATCH_MNT/t0 bs=1M count=1 > /dev/null 2>&1
if [ "${quick}Q" = "thoroughQ" ]; then
# The intention of this "thorough" test is to increase
# the probability of random errors, in particular in
# conjunction with the background noise generator and
# a sync call while the replace operation in ongoing.
# Unfortunately it takes quite some time to generate
# the test filesystem, therefore most data consists out
# of zeros although this data is not very useful for
# detecting misplaced read/write requests.
# Ignore ENOSPC, it's not a problem..
dd if=/dev/urandom of=$SCRATCH_MNT/r bs=1M count=200 >> $seqres.full 2>&1 &
dd if=/dev/zero of=$SCRATCH_MNT/0 bs=1M count=2000 >> $seqres.full 2>&1
wait
# a sync call while the replace operation is ongoing.
fssize=2048
elif [ "${with_cancel}Q" = "cancelQ" ]; then
# produce some data to prevent that the replace operation
# finishes before the cancel request is started
dd if=/dev/zero of=$SCRATCH_MNT/0 bs=1M count=1000 >> $seqres.full 2>&1
# The goal is to produce enough data to prevent that the
# replace operation finishes before the cancel request
# is started.
fssize=1024
else
fssize=64
fi
# since the available size was tested before, do not tolerate
# any failures
for i in `seq $fssize`; do
cp $SCRATCH_MNT/t0 $SCRATCH_MNT/t$i || _fail "cp failed"
done > /dev/null 2>> $seqres.full
sync; sync
btrfs_replace_test $source_dev $target_dev "" $with_cancel $quick
@@ -214,7 +228,7 @@ btrfs_replace_test()
# before the status is printed
$BTRFS_UTIL_PROG replace status $SCRATCH_MNT > $tmp.tmp 2>&1
cat $tmp.tmp >> $seqres.full
grep -q canceled $tmp.tmp || _fail "btrfs replace status failed"
grep -q canceled $tmp.tmp || _fail "btrfs replace status (canceled) failed"
else
if [ "${quick}Q" = "thoroughQ" ]; then
# On current hardware, the thorough test runs
@@ -226,7 +240,7 @@ btrfs_replace_test()
$BTRFS_UTIL_PROG replace status $SCRATCH_MNT > $tmp.tmp 2>&1
cat $tmp.tmp >> $seqres.full
grep -q finished $tmp.tmp || _fail "btrfs replace status failed"
grep -q finished $tmp.tmp || _fail "btrfs replace status (finished) failed"
fi
if ps -p $noise_pid | grep -q $noise_pid; then