On Tue, Sep 29, 2015 at 06:13:37PM +0800, Qu Wenruo wrote:
>
>
> 在 2015年09月29日 18:00, Hugo Mills 写道:
> >On Tue, Sep 29, 2015 at 05:34:24PM +0800, Qu Wenruo wrote:
> >>Normally, a bull fallocate call on a fully written and synced file
> >>should not add an extent.
> >
> > What's a "bull" fallocate call? Is it a typo, or simply something
> >I'm not familiar with?
> >
> > Hugo.
>
> Oh, it should be null.
> But null still seems not appropriate here.
>
> I mean a fallocate call which doesn't really allocate any space...
>
> Any good ideas?
Not in a single word. Maybe one of the following:
Normally, an fallocate call which only touches existing extents should
not add any more extents on a fully written and synced file.
Normally, an fallocate call which does not allocate outside existing
extents should not add any more extents on a fully written and synced
file.
I think I slightly prefer the second option.
Hugo
> Thanks,
> Qu
> >
> >>But not all filesystem follows the correct behavior.
> >>
> >>Btrfs has a bug to always truncate the last page if the fallocate start
> >>offset is smaller than inode size.
> >>
> >>So add this test case to detect such malfunction.
> >>
> >>Signed-off-by: Qu Wenruo <quwenruo@xxxxxxxxxxxxxx>
> >>---
> >>v2:
> >> Add author info...
> >> Fix some comment typo
> >>---
> >> tests/generic/110 | 78 +++++++++++++++++++++++++++++++++++++++++++++++++++
> >> tests/generic/110.out | 3 ++
> >> tests/generic/group | 1 +
> >> 3 files changed, 82 insertions(+)
> >> create mode 100755 tests/generic/110
> >> create mode 100644 tests/generic/110.out
> >>
> >>diff --git a/tests/generic/110 b/tests/generic/110
> >>new file mode 100755
> >>index 0000000..b2b140c
> >>--- /dev/null
> >>+++ b/tests/generic/110
> >>@@ -0,0 +1,78 @@
> >>+#! /bin/bash
> >>+# FS QA Test 110
> >>+#
> >>+# Test if fallocate will create uneeded extra tailing extent
> >>+#
> >>+#-----------------------------------------------------------------------
> >>+# Copyright (c) 2015 Fujitsu. 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 /
> >>+ rm -f $tmp.*
> >>+}
> >>+
> >>+# get standard environment, filters and checks
> >>+. ./common/rc
> >>+. ./common/filter
> >>+. ./common/defrag
> >>+
> >>+# remove previous $seqres.full before test
> >>+rm -f $seqres.full
> >>+
> >>+# real QA test starts here
> >>+
> >>+_supported_fs generic
> >>+_supported_os IRIX Linux
> >>+_require_scratch
> >>+_need_to_be_root
> >>+
> >>+# Use 64K file size to match any sectorsize
> >>+# And with a unaligned tailing range to ensure it will be at least 2 pages
> >>+filesize=$(( 64 * 1024 + 1024 ))
> >>+
> >>+_scratch_mkfs > /dev/null 2>&1
> >>+_scratch_mount
> >>+$XFS_IO_PROG -f -c "pwrite 0 $filesize" $SCRATCH_MNT/foo | _filter_xfs_io
> >>+sync
> >>+orig_extent_nr=`_extent_count $SCRATCH_MNT/foo`
> >>+
> >>+# As all space are allocated and even written to disk, this falloc
> >>+# should do nothing with extent modification.
> >>+$XFS_IO_PROG -f -c "falloc 0 $filesize" $SCRATCH_MNT/foo
> >>+sync
> >>+new_extent_nr=`_extent_count $SCRATCH_MNT/foo`
> >>+
> >>+echo "orig: $orig_extent_nr, new: $new_extent_nr" >> $seqres.full
> >>+
> >>+if [ "x$orig_extent_nr" != "x$new_extent_nr" ]; then
> >>+ echo "number of extents mis-match after bull fallocate"
> >>+fi
> >>+
> >>+# success, all done
> >>+status=0
> >>+exit
> >>diff --git a/tests/generic/110.out b/tests/generic/110.out
> >>new file mode 100644
> >>index 0000000..64049da
> >>--- /dev/null
> >>+++ b/tests/generic/110.out
> >>@@ -0,0 +1,3 @@
> >>+QA output created by 110
> >>+wrote 66560/66560 bytes at offset 0
> >>+XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
> >>diff --git a/tests/generic/group b/tests/generic/group
> >>index 4ae256f..428f3e3 100644
> >>--- a/tests/generic/group
> >>+++ b/tests/generic/group
> >>@@ -112,6 +112,7 @@
> >> 107 auto quick metadata
> >> 108 auto quick rw
> >> 109 auto metadata dir
> >>+110 auto quick prealloc
> >> 112 rw aio auto quick
> >> 113 rw aio auto quick
> >> 117 attr auto quick
> >
--
Hugo Mills | For months now, we have been making triumphant
hugo@... carfax.org.uk | retreats before a demoralised enemy who is advancing
http://carfax.org.uk/ | in utter disorder.
PGP: E2AB1DE4 | Eric Frank Russell, Wasp
Attachment:
signature.asc
Description: Digital signature
