[v3,0/1] seal system mappings

Message

Jeff Xu Nov. 13, 2024, 7:16 p.m. UTC

From: Jeff Xu <[e-mail]>

Seal vdso, vvar, sigpage, uprobes and vsyscall.

Those mappings are readonly or executable only, sealing can protect
them from ever changing or unmapped during the life time of the process.
For complete descriptions of memory sealing, please see mseal.rst [1].

System mappings such as vdso, vvar, and sigpage (for arm) are
generated by the kernel during program initialization, and are
sealed after creation.

Unlike the aforementioned mappings, the uprobe mapping is not
established during program startup. However, its lifetime is the same
as the process's lifetime [1]. It is sealed from creation.

The vdso, vvar, sigpage, and uprobe mappings all invoke the
_install_special_mapping() function. As no other mappings utilize this
function, it is logical to incorporate sealing logic within
_install_special_mapping(). This approach avoids the necessity of
modifying code across various architecture-specific implementations.

The vsyscall mapping, which has its own initialization function, is
sealed in the XONLY case, it seems to be the most common and secure
case of using vsyscall.

It is important to note that the CHECKPOINT_RESTORE feature (CRIU) may
alter the mapping of vdso, vvar, and sigpage during restore
operations. Consequently, this feature cannot be universally enabled
across all systems. To address this, a kernel configuration option has
been introduced to enable or disable this functionality.

[1] Documentation/userspace-api/mseal.rst
[2] https://lore.kernel.org/all/[e-mail]/

History:
V3:
  Revert uprobe to v1 logic (Oleg Nesterov)
  use CONFIG_SEAL_SYSTEM_MAPPINGS instead of _ALWAYS/_NEVER (Kees Cook)
  Move kernel cmd line from fs/exec.c to mm/mseal.c and misc. refactor (Liam R. Howlett)

V2:
  https://lore.kernel.org/all/[e-mail]/
  Seal uprobe always (Oleg Nesterov)
  Update comments and description (Randy Dunlap, Liam R.Howlett, Oleg Nesterov)
  Rebase to linux_main

V1:
https://lore.kernel.org/all/[e-mail]/

Jeff Xu (1):
  exec: seal system mappings

 .../admin-guide/kernel-parameters.txt         | 10 +++++
 arch/x86/entry/vsyscall/vsyscall_64.c         |  9 ++++-
 include/linux/mm.h                            | 12 ++++++
 mm/mmap.c                                     | 10 +++++
 mm/mseal.c                                    | 39 +++++++++++++++++++
 security/Kconfig                              | 11 ++++++
 6 files changed, 89 insertions(+), 2 deletions(-)

Comments

Lorenzo Stoakes Nov. 13, 2024, 8:36 p.m. UTC | #1

You left kernel test bots and review unanswered on v2, which makes it
difficult to know whether you actually addressed everything.

Please respond to all outstanding comments in the v2 thread so we know,
thanks, even if it is to say 'this is no longer an issue as v3 addresses'.

On Wed, Nov 13, 2024 at 07:16:01PM +0000, [e-mail] wrote:
> From: Jeff Xu <[e-mail]>
>
> Seal vdso, vvar, sigpage, uprobes and vsyscall.
>
> Those mappings are readonly or executable only, sealing can protect
> them from ever changing or unmapped during the life time of the process.
> For complete descriptions of memory sealing, please see mseal.rst [1].
>
> System mappings such as vdso, vvar, and sigpage (for arm) are
> generated by the kernel during program initialization, and are
> sealed after creation.
>
> Unlike the aforementioned mappings, the uprobe mapping is not
> established during program startup. However, its lifetime is the same
> as the process's lifetime [1]. It is sealed from creation.
>
> The vdso, vvar, sigpage, and uprobe mappings all invoke the
> _install_special_mapping() function. As no other mappings utilize this
> function, it is logical to incorporate sealing logic within
> _install_special_mapping(). This approach avoids the necessity of
> modifying code across various architecture-specific implementations.

Some arches unmap VDSO's which mseal prevents, so are those broken now? Did
you test this?

>
> The vsyscall mapping, which has its own initialization function, is
> sealed in the XONLY case, it seems to be the most common and secure
> case of using vsyscall.
>
> It is important to note that the CHECKPOINT_RESTORE feature (CRIU) may
> alter the mapping of vdso, vvar, and sigpage during restore
> operations. Consequently, this feature cannot be universally enabled
> across all systems. To address this, a kernel configuration option has
> been introduced to enable or disable this functionality.
>
> [1] Documentation/userspace-api/mseal.rst
> [2] https://lore.kernel.org/all/[e-mail]/

I don't see any mention to testing, and this affects multiple different
architectures.

Please describe what testing you performed and on what architectures.

I suggest we allow this only for architectures you have explicitly tested,
especially as this is 'hidden' from arch maintainers who might find this
change surprising.

>
> History:
> V3:
>   Revert uprobe to v1 logic (Oleg Nesterov)
>   use CONFIG_SEAL_SYSTEM_MAPPINGS instead of _ALWAYS/_NEVER (Kees Cook)
>   Move kernel cmd line from fs/exec.c to mm/mseal.c and misc. refactor (Liam R. Howlett)
>
> V2:
>   https://lore.kernel.org/all/[e-mail]/
>   Seal uprobe always (Oleg Nesterov)
>   Update comments and description (Randy Dunlap, Liam R.Howlett, Oleg Nesterov)
>   Rebase to linux_main
>
> V1:
> https://lore.kernel.org/all/[e-mail]/
>
> Jeff Xu (1):
>   exec: seal system mappings
>
>  .../admin-guide/kernel-parameters.txt         | 10 +++++
>  arch/x86/entry/vsyscall/vsyscall_64.c         |  9 ++++-
>  include/linux/mm.h                            | 12 ++++++
>  mm/mmap.c                                     | 10 +++++
>  mm/mseal.c                                    | 39 +++++++++++++++++++
>  security/Kconfig                              | 11 ++++++
>  6 files changed, 89 insertions(+), 2 deletions(-)
>
> --
> 2.47.0.277.g8800431eea-goog
>