extN can report f_blocks in statfs in 2 different ways, based
on whether or not metadata overhead is counted. This has broken
in the past, so here's a test for it.
It looks at dumpe2fs output to get total blocks and free blocks
right after mkfs. The difference should be, by definition, the
exact amount of metadata overhead.
It then compares this to what's reported via stat -f for f_blocks.
For "minix" df, it should be exactly equal to the total blocks,
and for "bsd" df, it should be total blocks less overhead.
It tests that the latter is accurate to within a 1% tolerance.
Today the journal doesn't count as overhead in the statfs call;
this should be fixed kernelside but for many filesystems it'll be
within the threshold anyway.
Signed-off-by: Eric Sandeen <[email protected]>
---
diff --git a/285 b/285
new file mode 100755
index 0000000..cda8531
--- /dev/null
+++ b/285
@@ -0,0 +1,97 @@
+#! /bin/bash
+# FS QA Test No. 286
+#
+# Test overhead & df output for extN filesystems
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2012 Red Hat, Inc. 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
+#-----------------------------------------------------------------------
+#
+# creator
[email protected]
+
+seq=`basename $0`
+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
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs ext2 ext3 ext4
+_supported_os Linux
+_require_scratch
+
+rm -f $seq.full
+
+_scratch_mkfs >> $seq.full 2>&1
+
+TOTAL_BLOCKS=`dumpe2fs -h $SCRATCH_DEV 2>/dev/null \
+ | awk '/Block count:/{print $3}'`
+
+FREE_BLOCKS=`dumpe2fs -h $SCRATCH_DEV 2>/dev/null \
+ | awk '/Free blocks:/{print $3}'`
+
+# nb: kernels today don't count journal blocks as overhead, but should.
+# For most fileystems this will still be within tolerance.
+OVERHEAD=$(($TOTAL_BLOCKS-$FREE_BLOCKS))
+
+# bsddf|minixdf
+# Set the behaviour for the statfs system call. The minixdf
+# behaviour is to return in the f_blocks field the total number of
+# blocks of the filesystem, while the bsddf behaviour (which is
+# the default) is to subtract the overhead blocks used by the ext2
+# filesystem and not available for file storage.
+
+# stat -f output looks like; we get f_blocks from that, which
+# varies depending on the df mount options used below:
+
+# Filesystem 4K-blocks Used Available Use% Mounted on
+# /dev/sda8 2405312 2208933 196379 92% /
+
+_scratch_mount "-o minixdf"
+MINIX_F_BLOCKS=`stat -f $SCRATCH_MNT | awk '/^Blocks/{print $3}'`
+umount $SCRATCH_MNT
+
+_scratch_mount "-o bsddf"
+BSD_F_BLOCKS=`stat -f $SCRATCH_MNT | awk '/^Blocks/{print $3}'`
+umount $SCRATCH_MNT
+
+echo "Overhead is $OVERHEAD out of $TOTAL_BLOCKS" >> $seq.full
+echo "BSD blocks $BSD_F_BLOCKS" >> $seq.full
+echo "MINIX blocks $MINIX_F_BLOCKS" >> $seq.full
+
+# minix should be exactly equal (hence 0)
+_within_tolerance "minix f_blocks" $MINIX_F_BLOCKS $TOTAL_BLOCKS 0 -v
+# bsd should be within ... we'll say 1%
+_within_tolerance "bsd f_blocks" $BSD_F_BLOCKS $(($TOTAL_BLOCKS-$OVERHEAD)) 1% -v
+
+# success, all done
+status=0
+exit
diff --git a/285.out b/285.out
new file mode 100644
index 0000000..2075e21
--- /dev/null
+++ b/285.out
@@ -0,0 +1,3 @@
+QA output created by 285
+minix f_blocks is in range
+bsd f_blocks is in range
diff --git a/group b/group
index 104ed35..0b33178 100644
--- a/group
+++ b/group
@@ -403,3 +403,4 @@ deprecated
282 dump ioctl auto quick
283 dump ioctl auto quick
284 auto
+285 auto
On Thu, Sep 6, 2012 at 6:20 AM, Eric Sandeen <[email protected]> wrote:
> extN can report f_blocks in statfs in 2 different ways, based
> on whether or not metadata overhead is counted. This has broken
> in the past, so here's a test for it.
>
> It looks at dumpe2fs output to get total blocks and free blocks
> right after mkfs. The difference should be, by definition, the
> exact amount of metadata overhead.
>
> It then compares this to what's reported via stat -f for f_blocks.
> For "minix" df, it should be exactly equal to the total blocks,
> and for "bsd" df, it should be total blocks less overhead.
>
> It tests that the latter is accurate to within a 1% tolerance.
>
> Today the journal doesn't count as overhead in the statfs call;
> this should be fixed kernelside but for many filesystems it'll be
> within the threshold anyway.
>
> Signed-off-by: Eric Sandeen <[email protected]>
> ---
>
> diff --git a/285 b/285
> new file mode 100755
> index 0000000..cda8531
> --- /dev/null
> +++ b/285
> @@ -0,0 +1,97 @@
> +#! /bin/bash
> +# FS QA Test No. 286
The Test No. should be 285 here?
> +#
> +# Test overhead & df output for extN filesystems
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2012 Red Hat, Inc. 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
> +#-----------------------------------------------------------------------
> +#
> +# creator
> [email protected]
> +
> +seq=`basename $0`
> +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
> +
> +# real QA test starts here
> +
> +# Modify as appropriate.
> +_supported_fs ext2 ext3 ext4
> +_supported_os Linux
> +_require_scratch
> +
> +rm -f $seq.full
> +
> +_scratch_mkfs >> $seq.full 2>&1
> +
> +TOTAL_BLOCKS=`dumpe2fs -h $SCRATCH_DEV 2>/dev/null \
> + | awk '/Block count:/{print $3}'`
> +
> +FREE_BLOCKS=`dumpe2fs -h $SCRATCH_DEV 2>/dev/null \
> + | awk '/Free blocks:/{print $3}'`
I guess TOTAL_BLOCKS and FREE_BLOCKS will be wrong when testing ext4
on RHEL5 which has a separate dumpe4fs for ext4.
Thanks,
Eryu Guan
> +
> +# nb: kernels today don't count journal blocks as overhead, but should.
> +# For most fileystems this will still be within tolerance.
> +OVERHEAD=$(($TOTAL_BLOCKS-$FREE_BLOCKS))
> +
> +# bsddf|minixdf
> +# Set the behaviour for the statfs system call. The minixdf
> +# behaviour is to return in the f_blocks field the total number of
> +# blocks of the filesystem, while the bsddf behaviour (which is
> +# the default) is to subtract the overhead blocks used by the ext2
> +# filesystem and not available for file storage.
> +
> +# stat -f output looks like; we get f_blocks from that, which
> +# varies depending on the df mount options used below:
> +
> +# Filesystem 4K-blocks Used Available Use% Mounted on
> +# /dev/sda8 2405312 2208933 196379 92% /
> +
> +_scratch_mount "-o minixdf"
> +MINIX_F_BLOCKS=`stat -f $SCRATCH_MNT | awk '/^Blocks/{print $3}'`
> +umount $SCRATCH_MNT
> +
> +_scratch_mount "-o bsddf"
> +BSD_F_BLOCKS=`stat -f $SCRATCH_MNT | awk '/^Blocks/{print $3}'`
> +umount $SCRATCH_MNT
> +
> +echo "Overhead is $OVERHEAD out of $TOTAL_BLOCKS" >> $seq.full
> +echo "BSD blocks $BSD_F_BLOCKS" >> $seq.full
> +echo "MINIX blocks $MINIX_F_BLOCKS" >> $seq.full
> +
> +# minix should be exactly equal (hence 0)
> +_within_tolerance "minix f_blocks" $MINIX_F_BLOCKS $TOTAL_BLOCKS 0 -v
> +# bsd should be within ... we'll say 1%
> +_within_tolerance "bsd f_blocks" $BSD_F_BLOCKS $(($TOTAL_BLOCKS-$OVERHEAD)) 1% -v
> +
> +# success, all done
> +status=0
> +exit
> diff --git a/285.out b/285.out
> new file mode 100644
> index 0000000..2075e21
> --- /dev/null
> +++ b/285.out
> @@ -0,0 +1,3 @@
> +QA output created by 285
> +minix f_blocks is in range
> +bsd f_blocks is in range
> diff --git a/group b/group
> index 104ed35..0b33178 100644
> --- a/group
> +++ b/group
> @@ -403,3 +403,4 @@ deprecated
> 282 dump ioctl auto quick
> 283 dump ioctl auto quick
> 284 auto
> +285 auto
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
On 9/6/12 3:49 AM, Eryu Guan wrote:
> On Thu, Sep 6, 2012 at 6:20 AM, Eric Sandeen <[email protected]> wrote:
>> extN can report f_blocks in statfs in 2 different ways, based
>> on whether or not metadata overhead is counted. This has broken
>> in the past, so here's a test for it.
>>
>> It looks at dumpe2fs output to get total blocks and free blocks
>> right after mkfs. The difference should be, by definition, the
>> exact amount of metadata overhead.
>>
>> It then compares this to what's reported via stat -f for f_blocks.
>> For "minix" df, it should be exactly equal to the total blocks,
>> and for "bsd" df, it should be total blocks less overhead.
>>
>> It tests that the latter is accurate to within a 1% tolerance.
>>
>> Today the journal doesn't count as overhead in the statfs call;
>> this should be fixed kernelside but for many filesystems it'll be
>> within the threshold anyway.
>>
>> Signed-off-by: Eric Sandeen <[email protected]>
>> ---
>>
>> diff --git a/285 b/285
>> new file mode 100755
>> index 0000000..cda8531
>> --- /dev/null
>> +++ b/285
>> @@ -0,0 +1,97 @@
>> +#! /bin/bash
>> +# FS QA Test No. 286
>
> The Test No. should be 285 here?
Oh, yes. I figure I'll have to renumber before commit anyway ;)
>> +#
>> +# Test overhead & df output for extN filesystems
>> +#
>> +#-----------------------------------------------------------------------
>> +# Copyright (c) 2012 Red Hat, Inc. 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
>> +#-----------------------------------------------------------------------
>> +#
>> +# creator
>> [email protected]
>> +
>> +seq=`basename $0`
>> +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
>> +
>> +# real QA test starts here
>> +
>> +# Modify as appropriate.
>> +_supported_fs ext2 ext3 ext4
>> +_supported_os Linux
>> +_require_scratch
>> +
>> +rm -f $seq.full
>> +
>> +_scratch_mkfs >> $seq.full 2>&1
>> +
>> +TOTAL_BLOCKS=`dumpe2fs -h $SCRATCH_DEV 2>/dev/null \
>> + | awk '/Block count:/{print $3}'`
>> +
>> +FREE_BLOCKS=`dumpe2fs -h $SCRATCH_DEV 2>/dev/null \
>> + | awk '/Free blocks:/{print $3}'`
>
> I guess TOTAL_BLOCKS and FREE_BLOCKS will be wrong when testing ext4
> on RHEL5 which has a separate dumpe4fs for ext4.
dumpe2fs would fail, yes ... I don't know if we really want to special case rhel5's
weirdness in xfstests or not ...
-Eric
> Thanks,
> Eryu Guan
>> +
>> +# nb: kernels today don't count journal blocks as overhead, but should.
>> +# For most fileystems this will still be within tolerance.
>> +OVERHEAD=$(($TOTAL_BLOCKS-$FREE_BLOCKS))
>> +
>> +# bsddf|minixdf
>> +# Set the behaviour for the statfs system call. The minixdf
>> +# behaviour is to return in the f_blocks field the total number of
>> +# blocks of the filesystem, while the bsddf behaviour (which is
>> +# the default) is to subtract the overhead blocks used by the ext2
>> +# filesystem and not available for file storage.
>> +
>> +# stat -f output looks like; we get f_blocks from that, which
>> +# varies depending on the df mount options used below:
>> +
>> +# Filesystem 4K-blocks Used Available Use% Mounted on
>> +# /dev/sda8 2405312 2208933 196379 92% /
>> +
>> +_scratch_mount "-o minixdf"
>> +MINIX_F_BLOCKS=`stat -f $SCRATCH_MNT | awk '/^Blocks/{print $3}'`
>> +umount $SCRATCH_MNT
>> +
>> +_scratch_mount "-o bsddf"
>> +BSD_F_BLOCKS=`stat -f $SCRATCH_MNT | awk '/^Blocks/{print $3}'`
>> +umount $SCRATCH_MNT
>> +
>> +echo "Overhead is $OVERHEAD out of $TOTAL_BLOCKS" >> $seq.full
>> +echo "BSD blocks $BSD_F_BLOCKS" >> $seq.full
>> +echo "MINIX blocks $MINIX_F_BLOCKS" >> $seq.full
>> +
>> +# minix should be exactly equal (hence 0)
>> +_within_tolerance "minix f_blocks" $MINIX_F_BLOCKS $TOTAL_BLOCKS 0 -v
>> +# bsd should be within ... we'll say 1%
>> +_within_tolerance "bsd f_blocks" $BSD_F_BLOCKS $(($TOTAL_BLOCKS-$OVERHEAD)) 1% -v
>> +
>> +# success, all done
>> +status=0
>> +exit
>> diff --git a/285.out b/285.out
>> new file mode 100644
>> index 0000000..2075e21
>> --- /dev/null
>> +++ b/285.out
>> @@ -0,0 +1,3 @@
>> +QA output created by 285
>> +minix f_blocks is in range
>> +bsd f_blocks is in range
>> diff --git a/group b/group
>> index 104ed35..0b33178 100644
>> --- a/group
>> +++ b/group
>> @@ -403,3 +403,4 @@ deprecated
>> 282 dump ioctl auto quick
>> 283 dump ioctl auto quick
>> 284 auto
>> +285 auto
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>> the body of a message to [email protected]
>> More majordomo info at http://vger.kernel.org/majordomo-info.html