`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.
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/9eb0b7e5826fc76cad7b025bc46f267d4a8... [3]: https://github.com/bus1/dbus-broker/pull/366 [4]: https://lore.kernel.org/lkml/20230714114753.170814-1-david@readahead.eu/
Cc: stable@vger.kernel.org Signed-off-by: Barnabás Pőcze pobrn@protonmail.com ---
* v3: https://lore.kernel.org/linux-mm/20240611231409.3899809-1-jeffxu@chromium.or... * 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); }
Hi
On Sun, Jun 30, 2024 at 11:49 AM Barnabás Pőcze pobrn@protonmail.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.
The documentation is in linux main (653c5c75115c), I hope this gives clarity to the usage of MFD_NOEXEC_SEAL flag to application developers, furthermore I'm working on man page for memfd_create.
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.
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.
During V3 patch discussion, I sent my reasoning, but here are summaries:
- As one might have noticed, unlike other flags in memfd_create, MFD_NOEXEC_SEAL is actually a combination of multiple flags. The idea is to make it easier to use memfd in the most common way, which is NOEXEC + F_SEAL_EXEC + MFD_ALLOW_SEALING.
- The new sysctl vm.noexec = 1 helps existing applications move to a more secure way of using memfd. IMO, MFD_ALLOW_SEALING is included by default because most applications would rather have it than not. In any case, an app can set F_SEAL_SEAL to disable the sealing.
- MFD_NOEXEC_SEAL has been added for more than one year, multiple applications and distributions have backported and utilized it. Altering ABI now presents a degree of risk and may lead to disruptions.
Best regards, -Jeff
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].
Cc: stable@vger.kernel.org Signed-off-by: Barnabás Pőcze pobrn@protonmail.com
- v3: https://lore.kernel.org/linux-mm/20240611231409.3899809-1-jeffxu@chromium.or...
- 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
On 2024-06-30, Barnabás Pőcze pobrn@protonmail.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@cyphar.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].
Cc: stable@vger.kernel.org Signed-off-by: Barnabás Pőcze pobrn@protonmail.com
- v3: https://lore.kernel.org/linux-mm/20240611231409.3899809-1-jeffxu@chromium.or...
- 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
Hi
Gentle ping. Is there any chance we could move forward with this? I am not aware of any breakage it would cause; but longer the wait, the higher the likelihood.
Regards, Barnabás Pőcze
2024. június 30., vasárnap 20:49 keltezéssel, Barnabás Pőcze pobrn@protonmail.com írta:
`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.
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].
Cc: stable@vger.kernel.org Signed-off-by: Barnabás Pőcze pobrn@protonmail.com
- v3: https://lore.kernel.org/linux-mm/20240611231409.3899809-1-jeffxu@chromium.or...
- 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
linux-kselftest-mirror@lists.linaro.org