On 2024/6/26 13:08, Jiaqi Yan wrote:
> Add regression and new tests when hugepage has correctable memory
> errors, and how userspace wants to deal with it:
> * if enable_soft_offline=1, mapped hugepage is soft offlined
> * if enable_soft_offline=0, mapped hugepage is intact
>
> Free hugepages case is not explicitly covered by the tests.
>
> Hugepage having corrected memory errors is emulated with
> MADV_SOFT_OFFLINE.
Thanks for update.
>
> Signed-off-by: Jiaqi Yan <jiaqiyan@xxxxxxxxxx>
> ---
> tools/testing/selftests/mm/.gitignore | 1 +
> tools/testing/selftests/mm/Makefile | 1 +
> .../selftests/mm/hugetlb-soft-offline.c | 228 ++++++++++++++++++
> tools/testing/selftests/mm/run_vmtests.sh | 4 +
> 4 files changed, 234 insertions(+)
> create mode 100644 tools/testing/selftests/mm/hugetlb-soft-offline.c
...
> +static void test_soft_offline_common(int enable_soft_offline)
> +{
> + int fd;
> + int expect_errno = enable_soft_offline ? 0 : EOPNOTSUPP;
> + struct statfs file_stat;
> + unsigned long hugepagesize_kb = 0;
> + unsigned long nr_hugepages_before = 0;
> + unsigned long nr_hugepages_after = 0;
> + int ret;
> +
> + ksft_print_msg("Test soft-offline when enabled_soft_offline=%d\n",
> + enable_soft_offline);
> +
> + fd = create_hugetlbfs_file(&file_stat);
> + if (fd < 0)
> + ksft_exit_fail_msg("Failed to create hugetlbfs file\n");
> +
> + hugepagesize_kb = file_stat.f_bsize / 1024;
> + ksft_print_msg("Hugepagesize is %ldkB\n", hugepagesize_kb);
> +
> + if (set_enable_soft_offline(enable_soft_offline)) {
Nit: should this be written as if (set_enable_soft_offline(enable_soft_offline) != 0) to keep consistent with below code?
> + close(fd);
> + ksft_exit_fail_msg("Failed to set enable_soft_offline\n");
> + }
> +
> + if (read_nr_hugepages(hugepagesize_kb, &nr_hugepages_before) != 0) {
> + close(fd);
> + ksft_exit_fail_msg("Failed to read nr_hugepages\n");
> + }
> +
> + ksft_print_msg("Before MADV_SOFT_OFFLINE nr_hugepages=%ld\n",
> + nr_hugepages_before);
> +
> + ret = do_soft_offline(fd, 2 * file_stat.f_bsize, expect_errno);
> +
> + if (read_nr_hugepages(hugepagesize_kb, &nr_hugepages_after) != 0) {
> + close(fd);
> + ksft_exit_fail_msg("Failed to read nr_hugepages\n");
> + }
> +
...
> diff --git a/tools/testing/selftests/mm/run_vmtests.sh b/tools/testing/selftests/mm/run_vmtests.sh
> index 3157204b9047..781117fac1ba 100755
> --- a/tools/testing/selftests/mm/run_vmtests.sh
> +++ b/tools/testing/selftests/mm/run_vmtests.sh
> @@ -331,6 +331,10 @@ CATEGORY="hugetlb" run_test ./thuge-gen
> CATEGORY="hugetlb" run_test ./charge_reserved_hugetlb.sh -cgroup-v2
> CATEGORY="hugetlb" run_test ./hugetlb_reparenting_test.sh -cgroup-v2
> if $RUN_DESTRUCTIVE; then
> +nr_hugepages_tmp=$(cat /proc/sys/vm/nr_hugepages)
> +echo 8 > /proc/sys/vm/nr_hugepages
> +CATEGORY="hugetlb" run_test ./hugetlb-soft-offline
> +echo "$nr_hugepages_tmp" > /proc/sys/vm/nr_hugepages
Should we save and restore the value of /proc/sys/vm/enable_soft_offline too?
With above fixed, this patch looks good to me.
Acked-by: Miaohe Lin <linmiaohe@xxxxxxxxxx>
Thanks.
.
