Message-ID: <20240702.061524-ancient.sardines.voting.pucker-nTYAKsgH2OOw@cyphar.com>
Date: Mon, 1 Jul 2024 23:24:25 -0700
From: Aleksa Sarai <cyphar@...har.com>
To: Barnabás Pőcze <pobrn@...tonmail.com>
Cc: linux-mm@...ck.org, linux-kernel@...r.kernel.org, 
	linux-kselftest@...r.kernel.org, akpm@...ux-foundation.org, david@...dahead.eu, 
	dmitry.torokhov@...il.com, dverkamp@...omium.org, hughd@...gle.com, jeffxu@...gle.com, 
	jorgelo@...omium.org, keescook@...omium.org, skhan@...uxfoundation.org, 
	stable@...r.kernel.org
Subject: Re: [PATCH v4] memfd: `MFD_NOEXEC_SEAL` should not imply
 `MFD_ALLOW_SEALING`

On 2024-06-30, Barnabás Pőcze <pobrn@...tonmail.com> wrote:
> `MFD_NOEXEC_SEAL` should remove the executable bits and set `F_SEAL_EXEC`
> to prevent further modifications to the executable bits as per the comment
> in the uapi header file:
> 
>   not executable and sealed to prevent changing to executable
> 
> However, commit 105ff5339f498a ("mm/memfd: add MFD_NOEXEC_SEAL and MFD_EXEC")
> that introduced this feature made it so that `MFD_NOEXEC_SEAL` unsets
> `F_SEAL_SEAL`, essentially acting as a superset of `MFD_ALLOW_SEALING`.
> 
> Nothing implies that it should be so, and indeed up until the second version
> of the of the patchset[0] that introduced `MFD_EXEC` and `MFD_NOEXEC_SEAL`,
> `F_SEAL_SEAL` was not removed, however, it was changed in the third revision
> of the patchset[1] without a clear explanation.
> 
> This behaviour is surprising for application developers, there is no
> documentation that would reveal that `MFD_NOEXEC_SEAL` has the additional
> effect of `MFD_ALLOW_SEALING`. Additionally, combined with `vm.memfd_noexec=2`
> it has the effect of making all memfds initially sealable.
> 
> So do not remove `F_SEAL_SEAL` when `MFD_NOEXEC_SEAL` is requested,
> thereby returning to the pre-Linux 6.3 behaviour of only allowing
> sealing when `MFD_ALLOW_SEALING` is specified.

This behaviour makes sense, I'm a little sad I didn't catch it when I
was fixing vm.memfd_noexec. There is a possibility for breakage, but we
should give it a shot, given how new the API is (and the API itself was
also broken until Linux 6.6 anyway)...

Feel free to take my

Reviewed-by: Aleksa Sarai <cyphar@...har.com>

Thanks.

> Now, this is technically a uapi break. However, the damage is expected
> to be minimal. To trigger user visible change, a program has to do the
> following steps:
> 
>  - create memfd:
>    - with `MFD_NOEXEC_SEAL`,
>    - without `MFD_ALLOW_SEALING`;
>  - try to add seals / check the seals.
> 
> But that seems unlikely to happen intentionally since this change
> essentially reverts the kernel's behaviour to that of Linux <6.3,
> so if a program worked correctly on those older kernels, it will
> likely work correctly after this change.
> 
> I have used Debian Code Search and GitHub to try to find potential
> breakages, and I could only find a single one. dbus-broker's
> memfd_create() wrapper is aware of this implicit `MFD_ALLOW_SEALING`
> behaviour, and tries to work around it[2]. This workaround will
> break. Luckily, this only affects the test suite, it does not affect
> the normal operations of dbus-broker. There is a PR with a fix[3].
> 
> I also carried out a smoke test by building a kernel with this change
> and booting an Arch Linux system into GNOME and Plasma sessions.
> 
> There was also a previous attempt to address this peculiarity by
> introducing a new flag[4].
> 
> [0]: https://lore.kernel.org/lkml/20220805222126.142525-3-jeffxu@google.com/
> [1]: https://lore.kernel.org/lkml/20221202013404.163143-3-jeffxu@google.com/
> [2]: https://github.com/bus1/dbus-broker/blob/9eb0b7e5826fc76cad7b025bc46f267d4a8784cb/src/util/misc.c#L114
> [3]: https://github.com/bus1/dbus-broker/pull/366
> [4]: https://lore.kernel.org/lkml/20230714114753.170814-1-david@readahead.eu/
> 
> Cc: stable@...r.kernel.org
> Signed-off-by: Barnabás Pőcze <pobrn@...tonmail.com>
> ---
> 
> * v3: https://lore.kernel.org/linux-mm/20240611231409.3899809-1-jeffxu@chromium.org/
> * v2: https://lore.kernel.org/linux-mm/20240524033933.135049-1-jeffxu@google.com/
> * v1: https://lore.kernel.org/linux-mm/20240513191544.94754-1-pobrn@protonmail.com/
> 
> This fourth version returns to removing the inconsistency as opposed to documenting
> its existence, with the same code change as v1 but with a somewhat extended commit
> message. This is sent because I believe it is worth at least a try; it can be easily
> reverted if bigger application breakages are discovered than initially imagined.
> 
> ---
>  mm/memfd.c                                 | 9 ++++-----
>  tools/testing/selftests/memfd/memfd_test.c | 2 +-
>  2 files changed, 5 insertions(+), 6 deletions(-)
> 
> diff --git a/mm/memfd.c b/mm/memfd.c
> index 7d8d3ab3fa37..8b7f6afee21d 100644
> --- a/mm/memfd.c
> +++ b/mm/memfd.c
> @@ -356,12 +356,11 @@ SYSCALL_DEFINE2(memfd_create,
>  
>  		inode->i_mode &= ~0111;
>  		file_seals = memfd_file_seals_ptr(file);
> -		if (file_seals) {
> -			*file_seals &= ~F_SEAL_SEAL;
> +		if (file_seals)
>  			*file_seals |= F_SEAL_EXEC;
> -		}
> -	} else if (flags & MFD_ALLOW_SEALING) {
> -		/* MFD_EXEC and MFD_ALLOW_SEALING are set */
> +	}
> +
> +	if (flags & MFD_ALLOW_SEALING) {
>  		file_seals = memfd_file_seals_ptr(file);
>  		if (file_seals)
>  			*file_seals &= ~F_SEAL_SEAL;
> diff --git a/tools/testing/selftests/memfd/memfd_test.c b/tools/testing/selftests/memfd/memfd_test.c
> index 95af2d78fd31..7b78329f65b6 100644
> --- a/tools/testing/selftests/memfd/memfd_test.c
> +++ b/tools/testing/selftests/memfd/memfd_test.c
> @@ -1151,7 +1151,7 @@ static void test_noexec_seal(void)
>  			    mfd_def_size,
>  			    MFD_CLOEXEC | MFD_NOEXEC_SEAL);
>  	mfd_assert_mode(fd, 0666);
> -	mfd_assert_has_seals(fd, F_SEAL_EXEC);
> +	mfd_assert_has_seals(fd, F_SEAL_SEAL | F_SEAL_EXEC);
>  	mfd_fail_chmod(fd, 0777);
>  	close(fd);
>  }
> -- 
> 2.45.2
> 
> 

-- 
Aleksa Sarai
Senior Software Engineer (Containers)
SUSE Linux GmbH
<https://www.cyphar.com/>

Download attachment "signature.asc" of type "application/pgp-signature" (229 bytes)

Powered by blists - more mailing lists