Date: Wed, 1 May 2024 10:15:54 -0700
From: Yosry Ahmed <yosryahmed@...gle.com>
To: Usama Arif <usamaarif642@...il.com>
Cc: akpm@...ux-foundation.org, hannes@...xchg.org, nphamcs@...il.com, 
	chengming.zhou@...ux.dev, linux-mm@...ck.org, linux-kernel@...r.kernel.org, 
	kernel-team@...a.com
Subject: Re: [PATCH] selftests: cgroup: add tests to verify the zswap
 writeback path

Hi Usama,

On Wed, May 1, 2024 at 3:04 AM Usama Arif <usamaarif642@...il.com> wrote:
>
> The condition for writeback can be triggered by allocating random
> memory more than memory.high to push memory into zswap, more than
> zswap.max to trigger writeback if enabled, but less than memory.max
> so that OOM is not triggered. Both values of memory.zswap.writeback
> are tested.

Thanks for working on this :)

>
> Signed-off-by: Usama Arif <usamaarif642@...il.com>
> ---
>  tools/testing/selftests/cgroup/test_zswap.c | 83 +++++++++++++++++++++
>  1 file changed, 83 insertions(+)
>
> diff --git a/tools/testing/selftests/cgroup/test_zswap.c b/tools/testing/selftests/cgroup/test_zswap.c
> index f0e488ed90d8..fe0e7221525c 100644
> --- a/tools/testing/selftests/cgroup/test_zswap.c
> +++ b/tools/testing/selftests/cgroup/test_zswap.c
> @@ -94,6 +94,19 @@ static int allocate_bytes(const char *cgroup, void *arg)
>         return 0;
>  }
>
> +static int allocate_random_bytes(const char *cgroup, void *arg)
> +{
> +       size_t size = (size_t)arg;
> +       char *mem = (char *)malloc(size);
> +
> +       if (!mem)
> +               return -1;
> +       for (int i = 0; i < size; i++)
> +               mem[i] = rand() % 128;
> +       free(mem);
> +       return 0;
> +}
> +
>  static char *setup_test_group_1M(const char *root, const char *name)
>  {
>         char *group_name = cg_name(root, name);
> @@ -248,6 +261,74 @@ static int test_zswapin(const char *root)
>         return ret;
>  }
>
> +/* Test to verify the zswap writeback path */
> +static int test_zswap_writeback(const char *root, bool wb)
> +{
> +       int ret = KSFT_FAIL;
> +       char *test_group;
> +       long zswpwb_before, zswpwb_after;
> +
> +       test_group = cg_name(root,
> +               wb ? "zswap_writeback_enabled_test" : "zswap_writeback_disabled_test");
> +       if (!test_group)
> +               goto out;
> +       if (cg_create(test_group))
> +               goto out;
> +       if (cg_write(test_group, "memory.max", "8M"))
> +               goto out;
> +       if (cg_write(test_group, "memory.high", "2M"))
> +               goto out;
> +       if (cg_write(test_group, "memory.zswap.max", "2M"))
> +               goto out;
> +       if (cg_write(test_group, "memory.zswap.writeback", wb ? "1" : "0"))
> +               goto out;
> +
> +       zswpwb_before = cg_read_key_long(test_group, "memory.stat", "zswpwb ");
> +       if (zswpwb_before < 0) {
> +               ksft_print_msg("failed to get zswpwb_before\n");
> +               goto out;
> +       }
> +
> +       /*
> +        * Allocate more than memory.high to push memory into zswap,
> +        * more than zswap.max to trigger writeback if enabled,
> +        * but less than memory.max so that OOM is not triggered
> +        */
> +       if (cg_run(test_group, allocate_random_bytes, (void *)MB(3)))
> +               goto out;

We set the zswap limit to 2M. So for this to work properly we need to
guarantee that the 3M of random data will compress into more than 2M.
Is this true for all possible zpool implementations and compression
algorithms? How likely for this to break and start producing false
negatives if zswap magically becomes more efficient?

One alternative approach that I used before, although more complex, is
to start by compressing the memory (i.e. through reclaim) without a
zswap limit, and check the zswap usage. Then, fault the memory back
in, set the zswap limit lower than the observed usage, and repeat.
This should guarantee writeback AFAICT.

Also, using memory.reclaim may be easier than memory.high if you
follow this approach, as you would need to raise memory.high again to
be able to decompress the memory.

> +
> +       /* Verify that zswap writeback occurred only if writeback was enabled */
> +       zswpwb_after = cg_read_key_long(test_group, "memory.stat", "zswpwb ");
> +       if (wb) {
> +               if (zswpwb_after <= zswpwb_before) {
> +                       ksft_print_msg("writeback enabled and zswpwb_after <= zswpwb_before\n");
> +                       goto out;
> +               }
> +       } else {
> +               if (zswpwb_after != zswpwb_before) {
> +                       ksft_print_msg("writeback disabled and zswpwb_after != zswpwb_before\n");
> +                       goto out;
> +               }
> +       }
> +
> +       ret = KSFT_PASS;
> +
> +out:
> +       cg_destroy(test_group);
> +       free(test_group);
> +       return ret;
> +}
> +
> +static int test_zswap_writeback_enabled(const char *root)
> +{
> +       return test_zswap_writeback(root, true);
> +}
> +
> +static int test_zswap_writeback_disabled(const char *root)
> +{
> +       return test_zswap_writeback(root, false);
> +}
> +
>  /*
>   * When trying to store a memcg page in zswap, if the memcg hits its memory
>   * limit in zswap, writeback should affect only the zswapped pages of that
> @@ -425,6 +506,8 @@ struct zswap_test {
>         T(test_zswap_usage),
>         T(test_swapin_nozswap),
>         T(test_zswapin),
> +       T(test_zswap_writeback_enabled),
> +       T(test_zswap_writeback_disabled),
>         T(test_no_kmem_bypass),
>         T(test_no_invasive_cgroup_shrink),
>  };
> --
> 2.43.0
>

Powered by blists - more mailing lists