linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Jeff Xu <jeffxu@google.com>
To: jeffxu@chromium.org
Cc: akpm@linux-foundation.org, keescook@chromium.org,
	 torvalds@linux-foundation.org, usama.anjum@collabora.com,
	corbet@lwn.net,  Liam.Howlett@oracle.com,
	lorenzo.stoakes@oracle.com, jorgelo@chromium.org,
	 groeck@chromium.org, linux-kernel@vger.kernel.org,
	 linux-kselftest@vger.kernel.org, linux-mm@kvack.org,
	jannh@google.com,  sroettger@google.com, pedro.falcato@gmail.com,
	 linux-hardening@vger.kernel.org, willy@infradead.org,
	 gregkh@linuxfoundation.org, deraadt@openbsd.org,
	surenb@google.com,  merimus@google.com, rdunlap@infradead.org
Subject: Re: [PATCH] munmap sealed memory cause memory to split (bug)
Date: Wed, 16 Oct 2024 19:38:08 -0700	[thread overview]
Message-ID: <CALmYWFsnx+QvkqLyYRSBW1ueJonEJeXROq6UM7Bbktn3bu+PMQ@mail.gmail.com> (raw)
In-Reply-To: <20241017022627.3112811-1-jeffxu@chromium.org>

On Wed, Oct 16, 2024 at 7:26 PM <jeffxu@chromium.org> wrote:
>
> From: Jeff Xu <jeffxu@google.com>
>
> It appears there is a regression on the latest mm,
> when munmap seals memory, it can cause an unexpected VMA split.
> E.g. repro use this test.

It appears that this test has some dependency tests that haven't been
merged, so can't be run as is.
This is the repro step:

- Allocate 12 pages (0-11).
- Seal middle 4 pages (4567)
- munmap (2345) - this will fail

Seeing VMA for page (0123) is split as 2 VMAs (01)-(23), those 2 VMA
have the same attribute, and should be merged as one.

> ---
>  tools/testing/selftests/mm/mseal_test.c | 76 +++++++++++++++++++++++++
>  1 file changed, 76 insertions(+)
>
> diff --git a/tools/testing/selftests/mm/mseal_test.c b/tools/testing/selftests/mm/mseal_test.c
> index fa74dbe4a684..0af33e13b606 100644
> --- a/tools/testing/selftests/mm/mseal_test.c
> +++ b/tools/testing/selftests/mm/mseal_test.c
> @@ -1969,6 +1969,79 @@ static void test_madvise_filebacked_was_writable(bool seal)
>         REPORT_TEST_PASS();
>  }
>
> +static void test_munmap_free_multiple_ranges_with_split(bool seal)
> +{
> +       void *ptr;
> +       unsigned long page_size = getpagesize();
> +       unsigned long size = 12 * page_size;
> +       int ret;
> +       int prot;
> +
> +       setup_single_address(size, &ptr);
> +       FAIL_TEST_IF_FALSE(ptr != (void *)-1);
> +
> +       /* seal the middle 4 page */
> +       if (seal) {
> +               ret = sys_mseal(ptr + 4 * page_size, 4 * page_size);
> +               FAIL_TEST_IF_FALSE(!ret);
> +
> +               size = get_vma_size(ptr, &prot);
> +               FAIL_TEST_IF_FALSE(size == 4 * page_size);
> +               FAIL_TEST_IF_FALSE(prot == 4);
> +
> +               size = get_vma_size(ptr +  4 * page_size, &prot);
> +               FAIL_TEST_IF_FALSE(size == 4 * page_size);
> +               FAIL_TEST_IF_FALSE(prot == 4);
> +
> +               size = get_vma_size(ptr +  8 * page_size, &prot);
> +               FAIL_TEST_IF_FALSE(size == 4 * page_size);
> +               FAIL_TEST_IF_FALSE(prot == 4);
> +       }
> +
> +       /* munmap 4  pages from the third page */
> +       ret = sys_munmap(ptr + 2 * page_size, 4 * page_size);
> +       if (seal) {
> +               FAIL_TEST_IF_FALSE(ret);
> +               FAIL_TEST_IF_FALSE(errno == EPERM);
> +
> +               size = get_vma_size(ptr, &prot);
> +               FAIL_TEST_IF_FALSE(size == 4 * page_size);
> +               FAIL_TEST_IF_FALSE(prot == 4);
> +
> +               size = get_vma_size(ptr +  4 * page_size, &prot);
> +               FAIL_TEST_IF_FALSE(size == 4 * page_size);
> +               FAIL_TEST_IF_FALSE(prot == 4);
> +
> +               size = get_vma_size(ptr +  8 * page_size, &prot);
> +               FAIL_TEST_IF_FALSE(size == 4 * page_size);
> +               FAIL_TEST_IF_FALSE(prot == 4);
> +       } else
> +               FAIL_TEST_IF_FALSE(!ret);
> +
> +       /* munmap 4 pages from the sealed page */
> +       ret = sys_munmap(ptr + 6 * page_size, 4 * page_size);
> +       if (seal) {
> +               FAIL_TEST_IF_FALSE(ret);
> +               FAIL_TEST_IF_FALSE(errno == EPERM);
> +
> +               size = get_vma_size(ptr + 4 * page_size, &prot);
> +               FAIL_TEST_IF_FALSE(size == 4 * page_size);
> +               FAIL_TEST_IF_FALSE(prot == 4);
> +
> +               size = get_vma_size(ptr +  4 * page_size, &prot);
> +               FAIL_TEST_IF_FALSE(size == 4 * page_size);
> +               FAIL_TEST_IF_FALSE(prot == 4);
> +
> +               size = get_vma_size(ptr +  8 * page_size, &prot);
> +               FAIL_TEST_IF_FALSE(size == 4 * page_size);
> +               FAIL_TEST_IF_FALSE(prot == 4);
> +       } else
> +               FAIL_TEST_IF_FALSE(!ret);
> +
> +       REPORT_TEST_PASS();
> +}
> +
> +
>  int main(int argc, char **argv)
>  {
>         bool test_seal = seal_support();
> @@ -2099,5 +2172,8 @@ int main(int argc, char **argv)
>         test_madvise_filebacked_was_writable(false);
>         test_madvise_filebacked_was_writable(true);
>
> +       test_munmap_free_multiple_ranges_with_split(false);
> +       test_munmap_free_multiple_ranges_with_split(true);
> +
>         ksft_finished();
>  }
> --
> 2.47.0.rc1.288.g06298d1525-goog
>


  reply	other threads:[~2024-10-17  2:38 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-10-17  2:26 jeffxu
2024-10-17  2:38 ` Jeff Xu [this message]
2024-10-17  6:03 ` Greg KH
2024-10-17 16:17   ` Jeff Xu
2024-10-17  8:18 ` Lorenzo Stoakes
2024-10-17 16:20   ` Jeff Xu
2024-10-17 19:14     ` Pedro Falcato
2024-10-17 19:16       ` Lorenzo Stoakes
2024-10-17 20:12       ` Jeff Xu
2024-10-17  9:46 ` Lorenzo Stoakes
2024-10-17  9:48   ` Lorenzo Stoakes
2024-10-17  9:59   ` Lorenzo Stoakes

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=CALmYWFsnx+QvkqLyYRSBW1ueJonEJeXROq6UM7Bbktn3bu+PMQ@mail.gmail.com \
    --to=jeffxu@google.com \
    --cc=Liam.Howlett@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=corbet@lwn.net \
    --cc=deraadt@openbsd.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=groeck@chromium.org \
    --cc=jannh@google.com \
    --cc=jeffxu@chromium.org \
    --cc=jorgelo@chromium.org \
    --cc=keescook@chromium.org \
    --cc=linux-hardening@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-kselftest@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=lorenzo.stoakes@oracle.com \
    --cc=merimus@google.com \
    --cc=pedro.falcato@gmail.com \
    --cc=rdunlap@infradead.org \
    --cc=sroettger@google.com \
    --cc=surenb@google.com \
    --cc=torvalds@linux-foundation.org \
    --cc=usama.anjum@collabora.com \
    --cc=willy@infradead.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