[LTP] [PATCH 1/1] runtest/ltp-aiodio.part4: Remove duplicate entries

[LTP] [PATCH 1/1] runtest/ltp-aiodio.part4: Remove duplicate entries

[Petr Vorel]

Duplicate content was added in 67ce2884aa (2006).

It probably did not make sense to run it more times even at the time,
but it certainly does not make sense now (runltp and kirk run tests
sequentially one by one).

While at it, remove bogus comments.

Signed-off-by: Petr Vorel <pvorel@suse.cz>
---
 runtest/ltp-aiodio.part4 | 55 ++--------------------------------------
 1 file changed, 2 insertions(+), 53 deletions(-)

diff --git a/runtest/ltp-aiodio.part4 b/runtest/ltp-aiodio.part4
index de00b8a786..8e7354296d 100644
--- a/runtest/ltp-aiodio.part4
+++ b/runtest/ltp-aiodio.part4
@@ -1,65 +1,14 @@
 aio01 aio01
 aio02 aio02
 
-#Running dio_sparse & dirty tests
 DI000 dirty
 DS000 dio_sparse
-DI001 dirty
-DS001 dio_sparse
-DI002 dirty
-DS002 dio_sparse
-DI003 dirty
-DS003 dio_sparse
-DI004 dirty
-DS004 dio_sparse
-DI005 dirty
-DS005 dio_sparse
-DI006 dirty
-DS006 dio_sparse
-DI007 dirty
-DS007 dio_sparse
-DI008 dirty
-DS008 dio_sparse
-DI009 dirty
-DS009 dio_sparse
-#iteration on dio_sparse
-DIO00 dio_sparse
-DIO01 dio_sparse
-DIO02 dio_sparse
-DIO03 dio_sparse
-DIO04 dio_sparse
-DIO05 dio_sparse
-DIO06 dio_sparse
-DIO07 dio_sparse
-DIO08 dio_sparse
-DIO09 dio_sparse
-#Running aiodio_append
 AD000 aiodio_append
-AD001 aiodio_append
-AD002 aiodio_append
-AD003 aiodio_append
-AD004 aiodio_append
-AD005 aiodio_append
-AD006 aiodio_append
-AD007 aiodio_append
-AD008 aiodio_append
-AD009 aiodio_append
-#Running dio_append
 ADI000 dio_append
-ADI001 dio_append
-ADI002 dio_append
-ADI003 dio_append
-ADI004 dio_append
-ADI005 dio_append
-ADI006 dio_append
-ADI007 dio_append
-ADI008 dio_append
-ADI009 dio_append
-#Running dio_truncate
+
 DIT000 dio_truncate -n 1
 DIT001 dio_truncate
-DIT002 dio_truncate
-#Running dio_read
+
 DOR000 dio_read -n 1 -i 100 -r 512k -w 512k -s 32M
 DOR001 dio_read -n 10 -i 30 -r 512k -w 512k -s 32M
 DOR002 dio_read -n 20 -i 15 -r 512k -w 512k -s 32M
-- 
2.47.1


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH 1/1] runtest/ltp-aiodio.part4: Remove duplicate entries

[Andrea Cervesato via ltp]

Hi Petr,

On 1/6/25 15:04, Petr Vorel wrote:
> Duplicate content was added in 67ce2884aa (2006).
>
> It probably did not make sense to run it more times even at the time,
> but it certainly does not make sense now (runltp and kirk run tests
> sequentially one by one).
>
> While at it, remove bogus comments.
>
> Signed-off-by: Petr Vorel <pvorel@suse.cz>
> ---
>   runtest/ltp-aiodio.part4 | 55 ++--------------------------------------
>   1 file changed, 2 insertions(+), 53 deletions(-)
>
> diff --git a/runtest/ltp-aiodio.part4 b/runtest/ltp-aiodio.part4
> index de00b8a786..8e7354296d 100644
> --- a/runtest/ltp-aiodio.part4
> +++ b/runtest/ltp-aiodio.part4
> @@ -1,65 +1,14 @@
>   aio01 aio01
>   aio02 aio02
>   
> -#Running dio_sparse & dirty tests
>   DI000 dirty
>   DS000 dio_sparse
Uhm, I remember that we discussed with Cyril about it in the past. The 
dirty test was made in order to stress IO before dio_sparse if I'm not 
wrong.
> -DI001 dirty
> -DS001 dio_sparse
> -DI002 dirty
> -DS002 dio_sparse
> -DI003 dirty
> -DS003 dio_sparse
> -DI004 dirty
> -DS004 dio_sparse
> -DI005 dirty
> -DS005 dio_sparse
> -DI006 dirty
> -DS006 dio_sparse
> -DI007 dirty
> -DS007 dio_sparse
> -DI008 dirty
> -DS008 dio_sparse
> -DI009 dirty
> -DS009 dio_sparse
> -#iteration on dio_sparse
> -DIO00 dio_sparse
> -DIO01 dio_sparse
> -DIO02 dio_sparse
> -DIO03 dio_sparse
> -DIO04 dio_sparse
> -DIO05 dio_sparse
> -DIO06 dio_sparse
> -DIO07 dio_sparse
> -DIO08 dio_sparse
> -DIO09 dio_sparse
> -#Running aiodio_append
>   AD000 aiodio_append
> -AD001 aiodio_append
> -AD002 aiodio_append
> -AD003 aiodio_append
> -AD004 aiodio_append
> -AD005 aiodio_append
> -AD006 aiodio_append
> -AD007 aiodio_append
> -AD008 aiodio_append
> -AD009 aiodio_append
> -#Running dio_append
>   ADI000 dio_append
> -ADI001 dio_append
> -ADI002 dio_append
> -ADI003 dio_append
> -ADI004 dio_append
> -ADI005 dio_append
> -ADI006 dio_append
> -ADI007 dio_append
> -ADI008 dio_append
> -ADI009 dio_append
> -#Running dio_truncate
> +
>   DIT000 dio_truncate -n 1
>   DIT001 dio_truncate
> -DIT002 dio_truncate
> -#Running dio_read
> +
>   DOR000 dio_read -n 1 -i 100 -r 512k -w 512k -s 32M
>   DOR001 dio_read -n 10 -i 30 -r 512k -w 512k -s 32M
>   DOR002 dio_read -n 20 -i 15 -r 512k -w 512k -s 32M
Andrea

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH 1/1] runtest/ltp-aiodio.part4: Remove duplicate entries

[Petr Vorel]

> Hi Petr,

> On 1/6/25 15:04, Petr Vorel wrote:
> > Duplicate content was added in 67ce2884aa (2006).

> > It probably did not make sense to run it more times even at the time,
> > but it certainly does not make sense now (runltp and kirk run tests
> > sequentially one by one).

> > While at it, remove bogus comments.

> > Signed-off-by: Petr Vorel <pvorel@suse.cz>
> > ---
> >   runtest/ltp-aiodio.part4 | 55 ++--------------------------------------
> >   1 file changed, 2 insertions(+), 53 deletions(-)

> > diff --git a/runtest/ltp-aiodio.part4 b/runtest/ltp-aiodio.part4
> > index de00b8a786..8e7354296d 100644
> > --- a/runtest/ltp-aiodio.part4
> > +++ b/runtest/ltp-aiodio.part4
> > @@ -1,65 +1,14 @@
> >   aio01 aio01
> >   aio02 aio02
> > -#Running dio_sparse & dirty tests
> >   DI000 dirty
> >   DS000 dio_sparse
> Uhm, I remember that we discussed with Cyril about it in the past. The dirty
> test was made in order to stress IO before dio_sparse if I'm not wrong.

Andrea, thanks for info.

@Cyril? If relevant maybe we should just document it. Does it apply to all
duplicates? I wonder if it makes sense to run whole runset for 45 min when we
are testing something mainline does not care about it's performance.

https://lore.kernel.org/ltp/20240722090012.mlvkaenuxar2x3vr@quack3/

Kind regards,
Petr


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH 1/1] runtest/ltp-aiodio.part4: Remove duplicate entries

[Cyril Hrubis]

Hi!
> @Cyril? If relevant maybe we should just document it. Does it apply to all
> duplicates? I wonder if it makes sense to run whole runset for 45 min when we
> are testing something mainline does not care about it's performance.

Well the ltp-aiodio.part4 is a stress test that iterates between
different I/O stress tests. Stress tests by definition must run for some
time to be useful, the question is for how long these should be run and
for that there is no clear answer I'm afraid.

The dirty and dio_sparse are interleaved because dirty tries to fill in
the page cache with dirty pages before each dio_sparse iteration. The
dio_sparse then creates a sparse file and checks that we correctly read
zeroes all the time while other processes write zeroes to the sparse
file. It may make sense to merge the dirty.c into the dio_sparse.c so
that it's executed at the start of the test and then possibly eliminate
one block of the dio_sparse tests.

We may as well get rid of the multiple lines in the runtest file by
passing -i option since these tests were converted into the new test
library. But over all I wouldn't vote for reducing the runtime for these
tests as drastically as you proposed.

-- 
Cyril Hrubis
chrubis@suse.cz

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH 1/1] runtest/ltp-aiodio.part4: Remove duplicate entries

[Andrea Cervesato via ltp]

Hi Cyril,

On 1/7/25 11:43, Cyril Hrubis wrote:
> Hi!
>> @Cyril? If relevant maybe we should just document it. Does it apply to all
>> duplicates? I wonder if it makes sense to run whole runset for 45 min when we
>> are testing something mainline does not care about it's performance.
> Well the ltp-aiodio.part4 is a stress test that iterates between
> different I/O stress tests. Stress tests by definition must run for some
> time to be useful, the question is for how long these should be run and
> for that there is no clear answer I'm afraid.
>
> The dirty and dio_sparse are interleaved because dirty tries to fill in
> the page cache with dirty pages before each dio_sparse iteration. The
> dio_sparse then creates a sparse file and checks that we correctly read
> zeroes all the time while other processes write zeroes to the sparse
> file. It may make sense to merge the dirty.c into the dio_sparse.c so
> that it's executed at the start of the test and then possibly eliminate
> one block of the dio_sparse tests.
>
> We may as well get rid of the multiple lines in the runtest file by
> passing -i option since these tests were converted into the new test
> library. But over all I wouldn't vote for reducing the runtime for these
> tests as drastically as you proposed.
>
I agree with everything. The usage of runtest file for something 
strictly related to ltp-aiodio.part4 doesn't sound right.
We need to get rid of "dirty" binary and include it inside the test.

Andrea


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp