From: "David Hildenbrand (Arm)" <david@kernel.org>
To: Donggeun Yoo <donggeunyoo.kernel@gmail.com>,
akpm@linux-foundation.org, rppt@kernel.org
Cc: peterx@redhat.com, surenb@google.com, aarcange@redhat.com,
ljs@kernel.org, liam@infradead.org, vbabka@kernel.org,
mhocko@suse.com, shuah@kernel.org, kirill@shutemov.name,
linux-mm@kvack.org, linux-kselftest@vger.kernel.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3 2/2] selftests/mm: add tests for UFFDIO_MOVE of a uffd-protected swap entry
Date: Mon, 28 Sep 2026 11:11:50 +0200 [thread overview]
Message-ID: <cc123233-5b48-4069-8f84-f95301a83dcf@kernel.org> (raw)
In-Reply-To: <20260926124145.2878520-3-donggeunyoo.kernel@gmail.com>
On 9/26/26 14:41, Donggeun Yoo wrote:
> Move a swapped-out page out of a write-protected or RWP-protected area into
> a destination registered for missing faults only, and read pagemap bit 57
> at the destination. The destination was never protected, so the bit must
> be clear.
>
> Assisted-by: LLM
> Signed-off-by: Donggeun Yoo <donggeunyoo.kernel@gmail.com>
> ---
> tools/testing/selftests/mm/uffd-unit-tests.c | 84 ++++++++++++++++++++
> 1 file changed, 84 insertions(+)
>
> diff --git a/tools/testing/selftests/mm/uffd-unit-tests.c b/tools/testing/selftests/mm/uffd-unit-tests.c
> index ef9b3956bdcfd..580178630eded 100644
> --- a/tools/testing/selftests/mm/uffd-unit-tests.c
> +++ b/tools/testing/selftests/mm/uffd-unit-tests.c
> @@ -2037,6 +2037,75 @@ static void uffd_move_pmd_split_test(uffd_global_test_opts_t *gopts, uffd_test_a
> uffd_move_pmd_handle_fault);
> }
>
> +/*
> + * Moving a swapped-out page out of a write-protected or RWP-protected area
> + * must not carry the uffd bit into a destination registered for missing
> + * faults only: such a bit is never cleared afterwards, so pagemap keeps
> + * reporting the page as uffd-tracked.
> + *
> + * Needs a swap device; skipped if MADV_PAGEOUT cannot evict the page.
> + */
> +static void uffd_move_swap_test_common(uffd_global_test_opts_t *gopts,
> + bool rwp)
> +{
> + unsigned long page_size = gopts->page_size;
> + struct uffdio_move move = { };
> + int pagemap_fd;
> +
> + if (rwp) {
> + if (uffd_register_rwp(gopts->uffd, gopts->area_src, page_size))
> + err("register src failure");
> + } else if (uffd_register(gopts->uffd, gopts->area_src, page_size,
> + false, true, false)) {
> + err("register src failure");
> + }
Why not combine it with the protection?
if (rwp) {
if (uffd_register_rwp(gopts->uffd, gopts->area_src, page_size))
err("register src failure");
rwprotect_range(gopts->uffd, (unsigned long)gopts->area_src,
page_size, true);
} else {
if (uffd_register(gopts->uffd, gopts->area_src, page_size,
false, true, false))
err("register src failure");
wp_range(gopts->uffd, (unsigned long)gopts->area_src,
page_size, true);
}
> + if (uffd_register(gopts->uffd, gopts->area_dst, page_size,
> + true, false, false))
> + err("register dst failure");
> +
> + if (rwp)
> + rwprotect_range(gopts->uffd, (unsigned long)gopts->area_src,
> + page_size, true);
> + else
> + wp_range(gopts->uffd, (unsigned long)gopts->area_src,
> + page_size, true);
> +
> + pagemap_fd = pagemap_open();
Why not do that at the very top?
int pagemap_fd = pagemap_open();
> + if (madvise(gopts->area_src, page_size, MADV_PAGEOUT))
> + err("MADV_PAGEOUT");
> + if (!pagemap_is_swapped(pagemap_fd, gopts->area_src)) {
> + uffd_test_skip("MADV_PAGEOUT did not swap the page; is swap enabled?");
> + goto out;
> + }
> +
> + move.dst = (unsigned long)gopts->area_dst;
> + move.src = (unsigned long)gopts->area_src;
> + move.len = page_size;
You can set that all already when initializing aboce
struct uffdio_move move = {
.dst = (unsigned long)gopts->area_dst,
[...]
};
> + if (ioctl(gopts->uffd, UFFDIO_MOVE, &move))
> + err("UFFDIO_MOVE");
BTW, why do we fail the entire test here instead of only this test case?
--
Cheers,
David
prev parent reply other threads:[~2026-09-28 9:11 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-26 12:41 [PATCH v3 0/2] userfaultfd: clear the inherited uffd bit in move_swap_pte() Donggeun Yoo
2026-09-26 12:41 ` [PATCH v3 1/2] " Donggeun Yoo
2026-09-27 20:54 ` Andrew Morton
2026-09-28 7:33 ` David Hildenbrand (Arm)
2026-09-26 12:41 ` [PATCH v3 2/2] selftests/mm: add tests for UFFDIO_MOVE of a uffd-protected swap entry Donggeun Yoo
2026-09-28 8:53 ` David Hildenbrand (Arm)
2026-09-28 8:53 ` David Hildenbrand (Arm)
2026-09-28 9:11 ` David Hildenbrand (Arm) [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=cc123233-5b48-4069-8f84-f95301a83dcf@kernel.org \
--to=david@kernel.org \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=donggeunyoo.kernel@gmail.com \
--cc=kirill@shutemov.name \
--cc=liam@infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=ljs@kernel.org \
--cc=mhocko@suse.com \
--cc=peterx@redhat.com \
--cc=rppt@kernel.org \
--cc=shuah@kernel.org \
--cc=surenb@google.com \
--cc=vbabka@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®