generic/469: Skip the test if fallocate keep_size is not supported
diff mbox series

Message ID 20191016105347.16650-1-misono.tomohiro@jp.fujitsu.com
State New
Headers show
Series
  • generic/469: Skip the test if fallocate keep_size is not supported
Related show

Commit Message

Misono Tomohiro Oct. 16, 2019, 10:53 a.m. UTC
This test requires fallocate keep_size. So, skip the test if
a target FS does not support the function to avoid false negative.

Signed-off-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>
---
 tests/generic/469 | 1 +
 1 file changed, 1 insertion(+)

Comments

Yang Xu Oct. 17, 2019, 2:42 a.m. UTC | #1
on 2019/10/16 18:53, Misono Tomohiro wrote:
> This test requires fallocate keep_size. So, skip the test if
> a target FS does not support the function to avoid false negative.

Can you tell me your kernel version and filesystem. IMO this fallocate 
keep_size feature has been supported since long times ago(xfs 
implemented it in 2008 and ext4 in 2007).
> 
> Signed-off-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>
> ---
>   tests/generic/469 | 1 +
>   1 file changed, 1 insertion(+)
> 
> diff --git a/tests/generic/469 b/tests/generic/469
> index 47fdf0cf..0c6f8ae5 100755
> --- a/tests/generic/469
> +++ b/tests/generic/469
> @@ -40,6 +40,7 @@ rm -f $seqres.full
>   _supported_fs generic
>   _supported_os Linux
>   _require_test
> +_require_xfs_io_command "falloc" "-k"
>   
>   run_fsx()
>   {
>
misono.tomohiro@fujitsu.com Oct. 17, 2019, 2:52 a.m. UTC | #2
> > This test requires fallocate keep_size. So, skip the test if a target
> > FS does not support the function to avoid false negative.
> 
> Can you tell me your kernel version and filesystem. IMO this fallocate keep_size feature has been supported since long times ago(xfs
> implemented it in 2008 and ext4 in 2007).

Hi,

I'm now testing virtiofs and want to avoid noise. I think NFS 4 does not support fallocate keep_size neither.

Thanks.

> >
> > Signed-off-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>
> > ---
> >   tests/generic/469 | 1 +
> >   1 file changed, 1 insertion(+)
> >
> > diff --git a/tests/generic/469 b/tests/generic/469 index
> > 47fdf0cf..0c6f8ae5 100755
> > --- a/tests/generic/469
> > +++ b/tests/generic/469
> > @@ -40,6 +40,7 @@ rm -f $seqres.full
> >   _supported_fs generic
> >   _supported_os Linux
> >   _require_test
> > +_require_xfs_io_command "falloc" "-k"
> >
> >   run_fsx()
> >   {
> >
Yang Xu Oct. 17, 2019, 3:55 a.m. UTC | #3
on 2019/10/17 10:52, Misono, Tomohiro/味曽野 智礼 wrote:
>>> This test requires fallocate keep_size. So, skip the test if a target
>>> FS does not support the function to avoid false negative.
>>
>> Can you tell me your kernel version and filesystem. IMO this fallocate keep_size feature has been supported since long times ago(xfs
>> implemented it in 2008 and ext4 in 2007).
> 
> Hi,
> 
> I'm now testing virtiofs and want to avoid noise. I think NFS 4 does not support fallocate keep_size neither.
> 
Hi

I see.  But in generic/499, it also uses fallocate keep size, so I think 
you can also add _require_xfs_io_command "falloc" "-k" in that case(I 
guess generic/499 doesn't run on your system).

ps: In generic/469, it can also add fzero check.

Thanks.
> Thanks.
> 
>>>
>>> Signed-off-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>
>>> ---
>>>    tests/generic/469 | 1 +
>>>    1 file changed, 1 insertion(+)
>>>
>>> diff --git a/tests/generic/469 b/tests/generic/469 index
>>> 47fdf0cf..0c6f8ae5 100755
>>> --- a/tests/generic/469
>>> +++ b/tests/generic/469
>>> @@ -40,6 +40,7 @@ rm -f $seqres.full
>>>    _supported_fs generic
>>>    _supported_os Linux
>>>    _require_test
>>> +_require_xfs_io_command "falloc" "-k"
>>>
>>>    run_fsx()
>>>    {
>>>
misono.tomohiro@fujitsu.com Oct. 17, 2019, 7:08 a.m. UTC | #4
> >>> This test requires fallocate keep_size. So, skip the test if a
> >>> target FS does not support the function to avoid false negative.
> >>
> >> Can you tell me your kernel version and filesystem. IMO this
> >> fallocate keep_size feature has been supported since long times ago(xfs implemented it in 2008 and ext4 in 2007).
> >
> > Hi,
> >
> > I'm now testing virtiofs and want to avoid noise. I think NFS 4 does not support fallocate keep_size neither.
> >
> Hi
> 
> I see.  But in generic/499, it also uses fallocate keep size, so I think you can also add _require_xfs_io_command "falloc" "-k"
> in that case(I guess generic/499 doesn't run on your system).
> 
> ps: In generic/469, it can also add fzero check.

Thanks, I have sent v2.
Eryu Guan Oct. 18, 2019, 8:39 a.m. UTC | #5
On Wed, Oct 16, 2019 at 07:53:47PM +0900, Misono Tomohiro wrote:
> This test requires fallocate keep_size. So, skip the test if
> a target FS does not support the function to avoid false negative.
> 
> Signed-off-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>

Would you please paste the false failure you've seen? I can't recall I
hit such failures, and I think fsx would just disable unsupported
operations and don't cause failure, it just prints a message to stderr,
and we already redirect stderr to $seqres.full in run_fsx in 469.

Thanks,
Eryu

> ---
>  tests/generic/469 | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/tests/generic/469 b/tests/generic/469
> index 47fdf0cf..0c6f8ae5 100755
> --- a/tests/generic/469
> +++ b/tests/generic/469
> @@ -40,6 +40,7 @@ rm -f $seqres.full
>  _supported_fs generic
>  _supported_os Linux
>  _require_test
> +_require_xfs_io_command "falloc" "-k"
>  
>  run_fsx()
>  {
> -- 
> 2.21.0
>
misono.tomohiro@fujitsu.com Oct. 18, 2019, 9:02 a.m. UTC | #6
> On Wed, Oct 16, 2019 at 07:53:47PM +0900, Misono Tomohiro wrote:
> > This test requires fallocate keep_size. So, skip the test if a target
> > FS does not support the function to avoid false negative.
> >
> > Signed-off-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>
> 
> Would you please paste the false failure you've seen? I can't recall I hit such failures, and I think fsx would just disable unsupported
> operations and don't cause failure, it just prints a message to stderr, and we already redirect stderr to $seqres.full in run_fsx
> in 469.

Hi, I just got output mismatch error:

$ cat results/generic/469.out.bad
QA output created by 469
fsx --replay-ops fsxops.0
Seed set to 1
main: filesystem does not support fallocate mode FALLOC_FL_KEEP_SIZE, disabling!
main: filesystem does not support fallocate mode FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE, disabling!
main: filesystem does not support fallocate mode FALLOC_FL_ZERO_RANGE, disabling!
main: filesystem does not support fallocate mode FALLOC_FL_COLLAPSE_RANGE, disabling!
main: filesystem does not support fallocate mode FALLOC_FL_INSERT_RANGE, disabling!
main: filesystem does not support clone range, disabling!
main: filesystem does not support dedupe range, disabling!
fallocate: 0x0 to 0x1000
do_preallocate: fallocate: Operation not supported
LOG DUMP (1 total operations):
1(  1 mod 256): FALLOC   0x0 thru 0x1000        (0x1000 bytes) PAST_EOF
Log of operations saved to "/test1/469.fsx.fsxops"; replay with --replay-ops
Correct content saved for comparison
(maybe hexdump "/test1/469.fsx" vs "/test1/469.fsx.fsxgood")

Thanks.

> 
> Thanks,
> Eryu
> 
> > ---
> >  tests/generic/469 | 1 +
> >  1 file changed, 1 insertion(+)
> >
> > diff --git a/tests/generic/469 b/tests/generic/469 index
> > 47fdf0cf..0c6f8ae5 100755
> > --- a/tests/generic/469
> > +++ b/tests/generic/469
> > @@ -40,6 +40,7 @@ rm -f $seqres.full
> >  _supported_fs generic
> >  _supported_os Linux
> >  _require_test
> > +_require_xfs_io_command "falloc" "-k"
> >
> >  run_fsx()
> >  {
> > --
> > 2.21.0
> >
Eryu Guan Oct. 18, 2019, 10:24 a.m. UTC | #7
On Fri, Oct 18, 2019 at 09:02:12AM +0000, misono.tomohiro@fujitsu.com wrote:
> > On Wed, Oct 16, 2019 at 07:53:47PM +0900, Misono Tomohiro wrote:
> > > This test requires fallocate keep_size. So, skip the test if a target
> > > FS does not support the function to avoid false negative.
> > >
> > > Signed-off-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>
> > 
> > Would you please paste the false failure you've seen? I can't recall I hit such failures, and I think fsx would just disable unsupported
> > operations and don't cause failure, it just prints a message to stderr, and we already redirect stderr to $seqres.full in run_fsx
> > in 469.
> 
> Hi, I just got output mismatch error:
> 
> $ cat results/generic/469.out.bad
> QA output created by 469
> fsx --replay-ops fsxops.0
> Seed set to 1
> main: filesystem does not support fallocate mode FALLOC_FL_KEEP_SIZE, disabling!
> main: filesystem does not support fallocate mode FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE, disabling!
> main: filesystem does not support fallocate mode FALLOC_FL_ZERO_RANGE, disabling!
> main: filesystem does not support fallocate mode FALLOC_FL_COLLAPSE_RANGE, disabling!
> main: filesystem does not support fallocate mode FALLOC_FL_INSERT_RANGE, disabling!
> main: filesystem does not support clone range, disabling!
> main: filesystem does not support dedupe range, disabling!
> fallocate: 0x0 to 0x1000
> do_preallocate: fallocate: Operation not supported
> LOG DUMP (1 total operations):
> 1(  1 mod 256): FALLOC   0x0 thru 0x1000        (0x1000 bytes) PAST_EOF
> Log of operations saved to "/test1/469.fsx.fsxops"; replay with --replay-ops
> Correct content saved for comparison
> (maybe hexdump "/test1/469.fsx" vs "/test1/469.fsx.fsxgood")

Ah, I recalled that I actually tried to fix this issue before, but Amir
didn't like my patch back then :)

https://spinics.net/lists/fstests/msg07953.html

Because my patch silently ignored 'keep_size' flag when replaying
operations, and look at it again, I don't like it either.

I think we could just skip the keep_size operations instead of ignoring
it.

Would you please try this patch?

diff --git a/ltp/fsx.c b/ltp/fsx.c
index 06d08e4e93f3..ae89bf665db4 100644
--- a/ltp/fsx.c
+++ b/ltp/fsx.c
@@ -2044,6 +2044,9 @@ have_op:
                if (!fallocate_calls) {
                        log4(OP_FALLOCATE, offset, size, FL_SKIPPED);
                        goto out;
+               } else if (keep_size && !keep_size_calls) {
+                       log4(OP_FALLOCATE, offset, size, FL_SKIPPED | FL_KEEP_SIZE);
+                       goto out;
                }
                break;
        case OP_PUNCH_HOLE:
@@ -2056,6 +2059,9 @@ have_op:
                if (!zero_range_calls) {
                        log4(OP_ZERO_RANGE, offset, size, FL_SKIPPED);
                        goto out;
+               } else if (keep_size && !keep_size_calls) {
+                       log4(OP_ZERO_RANGE, offset, size, FL_SKIPPED | FL_KEEP_SIZE);
+                       goto out;
                }
                break;
        case OP_COLLAPSE_RANGE:

Thanks!

Eryu
> 
> Thanks.
> 
> > 
> > Thanks,
> > Eryu
> > 
> > > ---
> > >  tests/generic/469 | 1 +
> > >  1 file changed, 1 insertion(+)
> > >
> > > diff --git a/tests/generic/469 b/tests/generic/469 index
> > > 47fdf0cf..0c6f8ae5 100755
> > > --- a/tests/generic/469
> > > +++ b/tests/generic/469
> > > @@ -40,6 +40,7 @@ rm -f $seqres.full
> > >  _supported_fs generic
> > >  _supported_os Linux
> > >  _require_test
> > > +_require_xfs_io_command "falloc" "-k"
> > >
> > >  run_fsx()
> > >  {
> > > --
> > > 2.21.0
> > >
misono.tomohiro@fujitsu.com Oct. 18, 2019, 10:52 a.m. UTC | #8
> On Fri, Oct 18, 2019 at 09:02:12AM +0000, misono.tomohiro@fujitsu.com wrote:
> > > On Wed, Oct 16, 2019 at 07:53:47PM +0900, Misono Tomohiro wrote:
> > > > This test requires fallocate keep_size. So, skip the test if a
> > > > target FS does not support the function to avoid false negative.
> > > >
> > > > Signed-off-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>
> > >
> > > Would you please paste the false failure you've seen? I can't recall
> > > I hit such failures, and I think fsx would just disable unsupported
> > > operations and don't cause failure, it just prints a message to stderr, and we already redirect stderr to $seqres.full in
> run_fsx in 469.
> >
> > Hi, I just got output mismatch error:
> >
> > $ cat results/generic/469.out.bad
> > QA output created by 469
> > fsx --replay-ops fsxops.0
> > Seed set to 1
> > main: filesystem does not support fallocate mode FALLOC_FL_KEEP_SIZE, disabling!
> > main: filesystem does not support fallocate mode FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE, disabling!
> > main: filesystem does not support fallocate mode FALLOC_FL_ZERO_RANGE, disabling!
> > main: filesystem does not support fallocate mode FALLOC_FL_COLLAPSE_RANGE, disabling!
> > main: filesystem does not support fallocate mode FALLOC_FL_INSERT_RANGE, disabling!
> > main: filesystem does not support clone range, disabling!
> > main: filesystem does not support dedupe range, disabling!
> > fallocate: 0x0 to 0x1000
> > do_preallocate: fallocate: Operation not supported LOG DUMP (1 total
> > operations):
> > 1(  1 mod 256): FALLOC   0x0 thru 0x1000        (0x1000 bytes) PAST_EOF
> > Log of operations saved to "/test1/469.fsx.fsxops"; replay with
> > --replay-ops Correct content saved for comparison (maybe hexdump
> > "/test1/469.fsx" vs "/test1/469.fsx.fsxgood")
> 
> Ah, I recalled that I actually tried to fix this issue before, but Amir didn't like my patch back then :)
> 
> https://spinics.net/lists/fstests/msg07953.html
> 
> Because my patch silently ignored 'keep_size' flag when replaying operations, and look at it again, I don't like it either.
> 
> I think we could just skip the keep_size operations instead of ignoring it.
> 
> Would you please try this patch?
> 
> diff --git a/ltp/fsx.c b/ltp/fsx.c
> index 06d08e4e93f3..ae89bf665db4 100644
> --- a/ltp/fsx.c
> +++ b/ltp/fsx.c
> @@ -2044,6 +2044,9 @@ have_op:
>                 if (!fallocate_calls) {
>                         log4(OP_FALLOCATE, offset, size, FL_SKIPPED);
>                         goto out;
> +               } else if (keep_size && !keep_size_calls) {
> +                       log4(OP_FALLOCATE, offset, size, FL_SKIPPED | FL_KEEP_SIZE);
> +                       goto out;
>                 }
>                 break;
>         case OP_PUNCH_HOLE:
> @@ -2056,6 +2059,9 @@ have_op:
>                 if (!zero_range_calls) {
>                         log4(OP_ZERO_RANGE, offset, size, FL_SKIPPED);
>                         goto out;
> +               } else if (keep_size && !keep_size_calls) {
> +                       log4(OP_ZERO_RANGE, offset, size, FL_SKIPPED | FL_KEEP_SIZE);
> +                       goto out;
>                 }
>                 break;
>         case OP_COLLAPSE_RANGE:
> 

With this patch, 469 just succeeds on my environment (virtiofs).
I'm fine with this approach:
  Reviewed-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>

Thanks!

> Thanks!
> 
> Eryu
> >
> > Thanks.
> >
> > >
> > > Thanks,
> > > Eryu
> > >
> > > > ---
> > > >  tests/generic/469 | 1 +
> > > >  1 file changed, 1 insertion(+)
> > > >
> > > > diff --git a/tests/generic/469 b/tests/generic/469 index
> > > > 47fdf0cf..0c6f8ae5 100755
> > > > --- a/tests/generic/469
> > > > +++ b/tests/generic/469
> > > > @@ -40,6 +40,7 @@ rm -f $seqres.full  _supported_fs generic
> > > > _supported_os Linux  _require_test
> > > > +_require_xfs_io_command "falloc" "-k"
> > > >
> > > >  run_fsx()
> > > >  {
> > > > --
> > > > 2.21.0
> > > >

Patch
diff mbox series

diff --git a/tests/generic/469 b/tests/generic/469
index 47fdf0cf..0c6f8ae5 100755
--- a/tests/generic/469
+++ b/tests/generic/469
@@ -40,6 +40,7 @@  rm -f $seqres.full
 _supported_fs generic
 _supported_os Linux
 _require_test
+_require_xfs_io_command "falloc" "-k"
 
 run_fsx()
 {