Re: [PATCH] selftests/mm: remove obsolete hugetlb vmemmap test

[SJ Park <sj@kernel.org>]

On Fri, 10 Jul 2026 17:24:27 +0800 Muchun Song <songmuchun@bytedance.com> wrote:

> The hugetlb vmemmap selftest was added to check the old HVO
> layout where tail vmemmap pages reused the head page.  That
> assumption no longer matches the current HVO mapping layout.
> 
> HVO now keeps a private backing page for the head vmemmap page
> and remaps redundant tail vmemmap pages to a shared read-only
> backing page.  The old page flag check is therefore testing an
> obsolete implementation detail rather than the current ABI or
> behavior.
> 
> Remove the stale test and its build, run, and ignore entries.

Nice!

> 
> Signed-off-by: Muchun Song <songmuchun@bytedance.com>

Reviewed-by: SJ Park <sj@kernel.org>

> ---
>  tools/testing/selftests/mm/.gitignore        |   2 -
>  tools/testing/selftests/mm/Makefile          |   1 -
>  tools/testing/selftests/mm/hugetlb-vmemmap.c | 132 -------------------
>  tools/testing/selftests/mm/run_vmtests.sh    |   1 -
>  4 files changed, 136 deletions(-)
>  delete mode 100644 tools/testing/selftests/mm/hugetlb-vmemmap.c
> 
> diff --git a/tools/testing/selftests/mm/.gitignore b/tools/testing/selftests/mm/.gitignore
> index 9ccd9e1447e6..227476d8ff1f 100644
> --- a/tools/testing/selftests/mm/.gitignore
> +++ b/tools/testing/selftests/mm/.gitignore
> @@ -3,11 +3,9 @@ cow
>  hugepage-mmap
>  hugepage-mremap
>  hugepage-shm
> -hugepage-vmemmap
>  hugetlb-mmap
>  hugetlb-mremap
>  hugetlb-shm
> -hugetlb-vmemmap
>  hugetlb-madvise
>  hugetlb-read-hwpoison
>  hugetlb-soft-offline

I remember Linus argued [1] we shouldn't remove names from giignore.  I don't
really care, so feel free to add my R-b: tag.  I just wanted to call out,
though.

[1] https://lore.kernel.org/all/CAHk-=whVY4mq1Fd=Vhwy0dK8fR5cOLkiy1OZMG4QyEMaGEd9ZQ@mail.gmail.com/


Thanks,
SJ

[...]