[PATCH] xfs/263: increase data section size to 1024M

[PATCH] xfs/263: increase data section size to 1024M

[Yang Xu]

On machine with using raid, this case will trigger
the following error:
==== NO CRC ====
+mkfs.xfs: small data volume, ignoring data volume stripe unit 512 and stripe width 512
== Options: rw ==
== Options: usrquota,rw ==

mkfs.xfs generates this error since xfsprogs commit 42371fb36
("mkfs: ignore data blockdev stripe geometry for small filesystems").
It disables automatic detection of stripe unit and width if the
data device is less than 1GB.

To slove false poistive, just increase data section size to 1G.

Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
---
 tests/xfs/263 | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tests/xfs/263 b/tests/xfs/263
index fadd6280..48581bac 100755
--- a/tests/xfs/263
+++ b/tests/xfs/263
@@ -73,11 +73,11 @@ function test_all_state()
 
 echo "==== NO CRC ===="
 # Control size to control inode numbers
-_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=512m" >> $seqres.full
+_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=1024m" >> $seqres.full
 test_all_state
 
 echo "==== CRC ===="
-_scratch_mkfs_xfs "-m crc=1 -d size=512m" >>$seqres.full
+_scratch_mkfs_xfs "-m crc=1 -d size=1024m" >>$seqres.full
 test_all_state
 
 status=0
-- 
2.39.1

Re: [PATCH] xfs/263: increase data section size to 1024M

[Darrick J. Wong]

On Wed, Nov 15, 2023 at 12:56:59AM -0500, Yang Xu wrote:
> On machine with using raid, this case will trigger
> the following error:
> ==== NO CRC ====
> +mkfs.xfs: small data volume, ignoring data volume stripe unit 512 and stripe width 512
> == Options: rw ==
> == Options: usrquota,rw ==
> 
> mkfs.xfs generates this error since xfsprogs commit 42371fb36
> ("mkfs: ignore data blockdev stripe geometry for small filesystems").
> It disables automatic detection of stripe unit and width if the
> data device is less than 1GB.
> 
> To slove false poistive, just increase data section size to 1G.

Is there a particular reason why this test needs -d size= at all?

There's a single comment about "Control size to control inode numbers"
but then filter_quota_state() seds the inode numbers to #XXX.  So
perhaps that part of the mkfs argument isn't necessary anymore?

--D

> Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
> ---
>  tests/xfs/263 | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/tests/xfs/263 b/tests/xfs/263
> index fadd6280..48581bac 100755
> --- a/tests/xfs/263
> +++ b/tests/xfs/263
> @@ -73,11 +73,11 @@ function test_all_state()
>  
>  echo "==== NO CRC ===="
>  # Control size to control inode numbers
> -_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=512m" >> $seqres.full
> +_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=1024m" >> $seqres.full
>  test_all_state
>  
>  echo "==== CRC ===="
> -_scratch_mkfs_xfs "-m crc=1 -d size=512m" >>$seqres.full
> +_scratch_mkfs_xfs "-m crc=1 -d size=1024m" >>$seqres.full
>  test_all_state
>  
>  status=0
> -- 
> 2.39.1
> 
> 

Re: [PATCH] xfs/263: increase data section size to 1024M

[Bill O'Donnell]

On Wed, Nov 15, 2023 at 12:56:59AM -0500, Yang Xu wrote:
> On machine with using raid, this case will trigger
> the following error:
> ==== NO CRC ====
> +mkfs.xfs: small data volume, ignoring data volume stripe unit 512 and stripe width 512
> == Options: rw ==
> == Options: usrquota,rw ==
> 
> mkfs.xfs generates this error since xfsprogs commit 42371fb36
> ("mkfs: ignore data blockdev stripe geometry for small filesystems").
> It disables automatic detection of stripe unit and width if the
> data device is less than 1GB.
> 
> To slove false poistive, just increase data section size to 1G.
     ^^^^^^^^^^^^^^^^^^^^
     solve false positive
> 
> Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>

Reviewed-by: Bill O'Donnell <bodonnel@redhat.com>

> ---
>  tests/xfs/263 | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/tests/xfs/263 b/tests/xfs/263
> index fadd6280..48581bac 100755
> --- a/tests/xfs/263
> +++ b/tests/xfs/263
> @@ -73,11 +73,11 @@ function test_all_state()
>  
>  echo "==== NO CRC ===="
>  # Control size to control inode numbers
> -_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=512m" >> $seqres.full
> +_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=1024m" >> $seqres.full
>  test_all_state
>  
>  echo "==== CRC ===="
> -_scratch_mkfs_xfs "-m crc=1 -d size=512m" >>$seqres.full
> +_scratch_mkfs_xfs "-m crc=1 -d size=1024m" >>$seqres.full
>  test_all_state
>  
>  status=0
> -- 
> 2.39.1
> 
> 

RE: [PATCH] xfs/263: increase data section size to 1024M

[Yang Xu (Fujitsu)]

>On Wed, Nov 15, 2023 at 12:56:59AM -0500, Yang Xu wrote:
>> On machine with using raid, this case will trigger the following 
>> error:
>> ==== NO CRC ====
>> +mkfs.xfs: small data volume, ignoring data volume stripe unit 512 and 
>> +stripe width 512
>> == Options: rw ==
>> == Options: usrquota,rw ==
>> 
>> mkfs.xfs generates this error since xfsprogs commit 42371fb36
>> ("mkfs: ignore data blockdev stripe geometry for small filesystems").
>> It disables automatic detection of stripe unit and width if the data 
>> device is less than 1GB.
>> 
>> To slove false poistive, just increase data section size to 1G.

>Is there a particular reason why this test needs -d size= at all?

>There's a single comment about "Control size to control inode numbers"
>but then filter_quota_state() seds the inode numbers to #XXX.  So perhaps that part of the mkfs argument isn't necessary anymore?

Yes, you sed the inode number to xxx in 2017. I am fine with removing this mkfs argument but let's cc original author  Eric Sandeen.
@ Eric Sandeen What do you think removing the data section size argument?

Best Regards
Yang Xu

>--D

>> Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
>> ---
>>  tests/xfs/263 | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
>> 
>> diff --git a/tests/xfs/263 b/tests/xfs/263 index fadd6280..48581bac 
>> 100755
>> --- a/tests/xfs/263
>> +++ b/tests/xfs/263
>> @@ -73,11 +73,11 @@ function test_all_state()
> >  
> >  echo "==== NO CRC ===="
> >  # Control size to control inode numbers -_scratch_mkfs_xfs "-m crc=0 
> > -n ftype=0 -d size=512m" >> $seqres.full
> > +_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=1024m" >> $seqres.full
> >  test_all_state
> >  
> >  echo "==== CRC ===="
> > -_scratch_mkfs_xfs "-m crc=1 -d size=512m" >>$seqres.full
> > +_scratch_mkfs_xfs "-m crc=1 -d size=1024m" >>$seqres.full
> >  test_all_state
> >  
> >  status=0
> > --
> > 2.39.1
> > 
> > 

Re: [PATCH] xfs/263: increase data section size to 1024M

[Zorro Lang]

On Wed, Nov 15, 2023 at 08:05:43AM -0800, Darrick J. Wong wrote:
> On Wed, Nov 15, 2023 at 12:56:59AM -0500, Yang Xu wrote:
> > On machine with using raid, this case will trigger
> > the following error:
> > ==== NO CRC ====
> > +mkfs.xfs: small data volume, ignoring data volume stripe unit 512 and stripe width 512
> > == Options: rw ==
> > == Options: usrquota,rw ==
> > 
> > mkfs.xfs generates this error since xfsprogs commit 42371fb36
> > ("mkfs: ignore data blockdev stripe geometry for small filesystems").
> > It disables automatic detection of stripe unit and width if the
> > data device is less than 1GB.
> > 
> > To slove false poistive, just increase data section size to 1G.
> 
> Is there a particular reason why this test needs -d size= at all?
> 
> There's a single comment about "Control size to control inode numbers"
> but then filter_quota_state() seds the inode numbers to #XXX.  So
> perhaps that part of the mkfs argument isn't necessary anymore?

As we've filter out the inode number, so the "inode numbers control" doesn't
make sense anymore. It's fine for me to remove the "-d size=xxx" part (if it's
still test passed).

Thanks,
Zorro

> 
> --D
> 
> > Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
> > ---
> >  tests/xfs/263 | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/tests/xfs/263 b/tests/xfs/263
> > index fadd6280..48581bac 100755
> > --- a/tests/xfs/263
> > +++ b/tests/xfs/263
> > @@ -73,11 +73,11 @@ function test_all_state()
> >  
> >  echo "==== NO CRC ===="
> >  # Control size to control inode numbers
> > -_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=512m" >> $seqres.full
> > +_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=1024m" >> $seqres.full
> >  test_all_state
> >  
> >  echo "==== CRC ===="
> > -_scratch_mkfs_xfs "-m crc=1 -d size=512m" >>$seqres.full
> > +_scratch_mkfs_xfs "-m crc=1 -d size=1024m" >>$seqres.full
> >  test_all_state
> >  
> >  status=0
> > -- 
> > 2.39.1
> > 
> > 
> 

RE: [PATCH] xfs/263: increase data section size to 1024M

[Yang Xu (Fujitsu)]

Hi Zorro

> As we've filter out the inode number, so the "inode numbers control" doesn't make sense anymore. It's fine for me to remove the "-d size=xxx" part (if it's still test passed).
I have tested and pass. Will send a v2.

Best Regards
Yang Xu

-----Original Message-----
From: Zorro Lang <zlang@redhat.com> 
Sent: Thursday, November 16, 2023 10:34 AM
To: Xu, Yang/徐 杨 <xuyang2018.jy@fujitsu.com>
Cc: Darrick J. Wong <djwong@kernel.org>; fstests@vger.kernel.org
Subject: Re: [PATCH] xfs/263: increase data section size to 1024M

On Wed, Nov 15, 2023 at 08:05:43AM -0800, Darrick J. Wong wrote:
> On Wed, Nov 15, 2023 at 12:56:59AM -0500, Yang Xu wrote:
> > On machine with using raid, this case will trigger the following 
> > error:
> > ==== NO CRC ====
> > +mkfs.xfs: small data volume, ignoring data volume stripe unit 512 
> > +and stripe width 512
> > == Options: rw ==
> > == Options: usrquota,rw ==
> > 
> > mkfs.xfs generates this error since xfsprogs commit 42371fb36
> > ("mkfs: ignore data blockdev stripe geometry for small filesystems").
> > It disables automatic detection of stripe unit and width if the data 
> > device is less than 1GB.
> > 
> > To slove false poistive, just increase data section size to 1G.
> 
> Is there a particular reason why this test needs -d size= at all?
> 
> There's a single comment about "Control size to control inode numbers"
> but then filter_quota_state() seds the inode numbers to #XXX.  So 
> perhaps that part of the mkfs argument isn't necessary anymore?

As we've filter out the inode number, so the "inode numbers control" doesn't make sense anymore. It's fine for me to remove the "-d size=xxx" part (if it's still test passed).

Thanks,
Zorro

> 
> --D
> 
> > Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
> > ---
> >  tests/xfs/263 | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/tests/xfs/263 b/tests/xfs/263 index fadd6280..48581bac 
> > 100755
> > --- a/tests/xfs/263
> > +++ b/tests/xfs/263
> > @@ -73,11 +73,11 @@ function test_all_state()
> >  
> >  echo "==== NO CRC ===="
> >  # Control size to control inode numbers -_scratch_mkfs_xfs "-m 
> > crc=0 -n ftype=0 -d size=512m" >> $seqres.full
> > +_scratch_mkfs_xfs "-m crc=0 -n ftype=0 -d size=1024m" >> 
> > +$seqres.full
> >  test_all_state
> >  
> >  echo "==== CRC ===="
> > -_scratch_mkfs_xfs "-m crc=1 -d size=512m" >>$seqres.full
> > +_scratch_mkfs_xfs "-m crc=1 -d size=1024m" >>$seqres.full
> >  test_all_state
> >  
> >  status=0
> > --
> > 2.39.1
> > 
> > 
> 

Re: [PATCH] xfs/263: increase data section size to 1024M

[Eric Sandeen]

On 11/16/23 1:05 AM, Yang Xu (Fujitsu) wrote:
> Hi Zorro
> 
>> As we've filter out the inode number, so the "inode numbers control" doesn't make sense anymore. It's fine for me to remove the "-d size=xxx" part (if it's still test passed).
> I have tested and pass. Will send a v2.
> 
> Best Regards
> Yang Xu
> 
> -----Original Message-----
> From: Zorro Lang <zlang@redhat.com> 
> Sent: Thursday, November 16, 2023 10:34 AM
> To: Xu, Yang/徐 杨 <xuyang2018.jy@fujitsu.com>
> Cc: Darrick J. Wong <djwong@kernel.org>; fstests@vger.kernel.org
> Subject: Re: [PATCH] xfs/263: increase data section size to 1024M
> 
> On Wed, Nov 15, 2023 at 08:05:43AM -0800, Darrick J. Wong wrote:
>> On Wed, Nov 15, 2023 at 12:56:59AM -0500, Yang Xu wrote:
>>> On machine with using raid, this case will trigger the following 
>>> error:
>>> ==== NO CRC ====
>>> +mkfs.xfs: small data volume, ignoring data volume stripe unit 512 
>>> +and stripe width 512
>>> == Options: rw ==
>>> == Options: usrquota,rw ==
>>>
>>> mkfs.xfs generates this error since xfsprogs commit 42371fb36
>>> ("mkfs: ignore data blockdev stripe geometry for small filesystems").
>>> It disables automatic detection of stripe unit and width if the data 
>>> device is less than 1GB.
>>>
>>> To slove false poistive, just increase data section size to 1G.
>>
>> Is there a particular reason why this test needs -d size= at all?
>>
>> There's a single comment about "Control size to control inode numbers"
>> but then filter_quota_state() seds the inode numbers to #XXX.  So 
>> perhaps that part of the mkfs argument isn't necessary anymore?
> 
> As we've filter out the inode number, so the "inode numbers control" doesn't make sense anymore. It's fine for me to remove the "-d size=xxx" part (if it's still test passed).

Sorry for the late reply, but I agree that it's not needed now.

(I wonder if we should just filter the "small data volume, ignoring data volume
stripe" message in a common filter, but we don't need to tie up this fix, I
think it's fine to change the size for this test.)

Thanks,
-Eric