Apply a seccomp-BPF syscall allowlist by default

Derived from Podman's default profile, stripped of capability-conditional
rules (we never grant capabilities), argument filters, and the explicit
EPERM block. Dangerous syscalls (mount, unshare, ptrace, bpf,
perf_event_open, io_uring_*, keyctl, kexec_*, ...) fall through to the
default ENOSYS action, which also keeps glibc's clone3 -> clone fallback
working. x86_64 and aarch64 are supported; other archs error out.

Toggle with --seccomp / --no-seccomp or seccomp = <bool> in config.

src/lib.rs

@@ -5,6 +5,7 @@ pub mod config;
 mod errors;
 mod preflight;
 mod sandbox;
+mod seccomp;
 
 pub use errors::SandboxError;
 
@@ -23,6 +24,7 @@ pub struct SandboxConfig {
     pub mode: SandboxMode,
     pub hardened: bool,
     pub unshare_net: bool,
+    pub seccomp: bool,
     pub extra_rw: Vec<PathBuf>,
     pub extra_ro: Vec<PathBuf>,
     pub mask: Vec<PathBuf>,