Re: [PATCH 4/4] selftests: memcg: Remove protection from top level memcg

[Roman Gushchin <roman.gushchin-fxUVXftIFDnyG1zEObXtfA@public.gmane.org>]

On Tue, May 17, 2022 at 05:24:43PM -0700, Andrew Morton wrote:
> On Fri, 13 May 2022 11:59:56 -0700 Roman Gushchin <roman.gushchin-fxUVXftIFDnyG1zEObXtfA@public.gmane.org> wrote:
> 
> > On Fri, May 13, 2022 at 07:18:11PM +0200, Michal Koutny wrote:
> > > The reclaim is triggered by memory limit in a subtree, therefore the
> > > testcase does not need configured protection against external reclaim.
> > > 
> > > Also, correct/deduplicate respective comments
> > > 
> > > Signed-off-by: Michal Koutný <mkoutny-IBi9RG/b67k@public.gmane.org>
> > > ---
> > >  tools/testing/selftests/cgroup/test_memcontrol.c | 12 ++++--------
> > >  1 file changed, 4 insertions(+), 8 deletions(-)
> > > 
> > > diff --git a/tools/testing/selftests/cgroup/test_memcontrol.c b/tools/testing/selftests/cgroup/test_memcontrol.c
> > > index 9ffacf024bbd..9d370aafd799 100644
> > > --- a/tools/testing/selftests/cgroup/test_memcontrol.c
> > > +++ b/tools/testing/selftests/cgroup/test_memcontrol.c
> > > @@ -247,7 +247,7 @@ static int cg_test_proc_killed(const char *cgroup)
> > >  
> > >  /*
> > >   * First, this test creates the following hierarchy:
> > > - * A       memory.min = 50M,  memory.max = 200M
> > > + * A       memory.min = 0,    memory.max = 200M
> > >   * A/B     memory.min = 50M,  memory.current = 50M
> > >   * A/B/C   memory.min = 75M,  memory.current = 50M
> > >   * A/B/D   memory.min = 25M,  memory.current = 50M
> > > @@ -257,7 +257,7 @@ static int cg_test_proc_killed(const char *cgroup)
> > >   * Usages are pagecache, but the test keeps a running
> > >   * process in every leaf cgroup.
> > >   * Then it creates A/G and creates a significant
> > > - * memory pressure in it.
> > > + * memory pressure in A.
> > >   *
> > >   * A/B    memory.current ~= 50M
> > >   * A/B/C  memory.current ~= 29M
> > > @@ -335,8 +335,6 @@ static int test_memcg_min(const char *root)
> > >  			      (void *)(long)fd);
> > >  	}
> > >  
> > > -	if (cg_write(parent[0], "memory.min", "50M"))
> > > -		goto cleanup;
> > >  	if (cg_write(parent[1], "memory.min", "50M"))
> > >  		goto cleanup;
> > >  	if (cg_write(children[0], "memory.min", "75M"))
> > > @@ -404,8 +402,8 @@ static int test_memcg_min(const char *root)
> > >  
> > >  /*
> > >   * First, this test creates the following hierarchy:
> > > - * A       memory.low = 50M,  memory.max = 200M
> > > - * A/B     memory.low = 50M,  memory.current = 50M
> > > + * A       memory.low = 0,    memory.max = 200M
> > > + * A/B     memory.low = 50M,  memory.current = ...
> > 
> > Can you, please, just remove "memory.current = ...", it's not
> > because obvious what "..." means here.
> > 
> 
> You mean this?
> 
> --- a/tools/testing/selftests/cgroup/test_memcontrol.c~selftests-memcg-remove-protection-from-top-level-memcg-fix
> +++ a/tools/testing/selftests/cgroup/test_memcontrol.c
> @@ -403,15 +403,14 @@ cleanup:
>  /*
>   * First, this test creates the following hierarchy:
>   * A       memory.low = 0,    memory.max = 200M
> - * A/B     memory.low = 50M,  memory.current = ...
> + * A/B     memory.low = 50M
>   * A/B/C   memory.low = 75M,  memory.current = 50M
>   * A/B/D   memory.low = 25M,  memory.current = 50M
>   * A/B/E   memory.low = 0,    memory.current = 50M
>   * A/B/F   memory.low = 500M, memory.current = 0
>   *
>   * Usages are pagecache.
> - * Then it creates A/G an creates a significant
> - * memory pressure in it.
> + * Then it creates A/G and creates significant memory pressure in it.
>   *
>   * Then it checks actual memory usages and expects that:
>   * A/B    memory.current ~= 50M
> _
> 
> (includes gratuitous comment cleanup)

Yes, thank you!

> 
> I assume your comment in
> https://lkml.kernel.org/r/Yn6pBPq+lAXm9NG8@carbon can be addressed in a
> later patch.
> 
> I'm not sure what to amke of https://lkml.kernel.org/r/Yn6pWPodGPlz+D8G@carbon
> 
> Do we feel this series needs more work before merging it up?
> 

Please, go ahead with it. If anything comes up, it can be addressed later.

Thanks!