Andrea Righi andrea.righi@canonical.com writes:
On Fri, Oct 29, 2021 at 10:09:04AM -0500, Eric W. Biederman wrote:
As Andy pointed out that there are races between force_sig_info_to_task and sigaction[1] when force_sig_info_task. As Kees discovered[2] ptrace is also able to change these signals.
In the case of seeccomp killing a process with a signal it is a security violation to allow the signal to be caught or manipulated.
Solve this problem by introducing a new flag SA_IMMUTABLE that prevents sigaction and ptrace from modifying these forced signals. This flag is carefully made kernel internal so that no new ABI is introduced.
Longer term I think this can be solved by guaranteeing short circuit delivery of signals in this case. Unfortunately reliable and guaranteed short circuit delivery of these signals is still a ways off from being implemented, tested, and merged. So I have implemented a much simpler alternative for now.
[1] https://lkml.kernel.org/r/b5d52d25-7bde-4030-a7b1-7c6f8ab90660@www.fastmail.... [2] https://lkml.kernel.org/r/202110281136.5CE65399A7@keescook Cc: stable@vger.kernel.org Fixes: 307d522f5eb8 ("signal/seccomp: Refactor seccomp signal and coredump generation") Signed-off-by: "Eric W. Biederman" ebiederm@xmission.com
FWIW I've tested this patch and I confirm that it fixes the failure that I reported with the seccomp_bpf selftest.
Tested-by: Andrea Righi andrea.righi@canonical.com
Sigh. Except for the extra 0 in the definition of SA_IMMUTABLE that caused it to conflict with the x86 specific signal numbers.
Eric