[RFC][PATCH 5/7][v4] Protect cinit from blocked fatal signals

From: Sukadev Bhattiprolu
Date: Wed Dec 24 2008 - 06:54:18 EST



From: Sukadev Bhattiprolu <sukadev@xxxxxxxxxxxxxxxxxx>
Subject: [RFC][PATCH 5/7][v4] Protect cinit from blocked fatal signals

Normally SIG_DFL signals to global and container-init are dropped early.
But if a signal is blocked when it is posted, we cannot drop the signal
since the receiver may install a handler before unblocking the signal.
Once this signal is queued however, the receiver container-init has
no way of knowing if the signal was sent from an ancestor or descendant
namespace. This patch ensures that contianer-init drops all SIG_DFL
signals in get_signal_to_deliver() except SIGKILL/SIGSTOP.

If SIGSTOP/SIGKILL originate from a descendant of container-init they
are never queued (i.e dropped in sig_ignored() in an earler patch).

If SIGSTOP/SIGKILL originate from parent namespace, the signal is queued
and container-init processes the signal.

See comments in patch below for details.

Changelog[v2]:
- Rename sig_unkillable() to unkillable_by_sig()
- Remove SIGNAL_UNKILLABLE_FROM_NS flag and simplify (Oleg Nesterov)
- Set SIGNAL_UNKILLABLE for container-init in this patch.

Signed-off-by: Sukadev Bhattiprolu <sukadev@xxxxxxxxxxxxxxxxxx>
---
kernel/fork.c | 2 ++
kernel/signal.c | 41 +++++++++++++++++++++++++++++++++++++++--
2 files changed, 41 insertions(+), 2 deletions(-)

diff --git a/kernel/fork.c b/kernel/fork.c
index dba2d3f..d3e93ef 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -812,6 +812,8 @@ static int copy_signal(unsigned long clone_flags, struct task_struct *tsk)
atomic_set(&sig->live, 1);
init_waitqueue_head(&sig->wait_chldexit);
sig->flags = 0;
+ if (clone_flags & CLONE_NEWPID)
+ sig->flags |= SIGNAL_UNKILLABLE;
sig->group_exit_code = 0;
sig->group_exit_task = NULL;
sig->group_stop_count = 0;
diff --git a/kernel/signal.c b/kernel/signal.c
index 5c4374f..660fadd 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -1818,6 +1818,41 @@ static int ptrace_signal(int signr, siginfo_t *info,
return signr;
}

+/*
+ * Return 1 if the process owning @signal should NOT terminate as a result of
+ * the signal @signr. Return 0 otherwise.
+ *
+ * Specifically if process owning @signal is
+ * - neither global nor a container-init, return 0
+ * - the global-init, return 1.
+ * - container-init, return 0 if signal is SIGKILL or SIGSTOP. Return
+ * 1 otherwise.
+ *
+ * sig_ignored() drops any unblocked fatal signals to global/container-init
+ * from within the same namespace. This of course includes SIGKILL/SIGSTOP
+ * which can never be blocked. sig_ignored() does not drop the SIGKILL/
+ * SIGSTOP if the are from an ancestor namespace.
+ *
+ * So, @signal is for a container-init and if @signr is either SIGKILL or
+ * SIGSTOP, it must have come from an ancestor namespace. So container-init
+ * should be killable (return 0).
+ *
+ * If @signal refers to a container-init and @signr is neither SIGKILL nor
+ * SIGSTOP, it was queued because it was blocked when it was posted. The
+ * signal may have come from same container - hence it should not be
+ * killable (return 1).
+ *
+ * Note:
+ * This means that SIGKILL is the only sure way to terminate a
+ * container-init even from ancestor namespace.
+ */
+static int unkillable_by_sig(struct signal_struct *signal, int signr)
+{
+ if ((signal->flags & SIGNAL_UNKILLABLE) && !sig_kernel_only(signr))
+ return 1;
+ return 0;
+}
+
int get_signal_to_deliver(siginfo_t *info, struct k_sigaction *return_ka,
struct pt_regs *regs, void *cookie)
{
@@ -1909,9 +1944,11 @@ relock:

/*
* Global init gets no signals it doesn't want.
+ * Container-init gets no signals it doesn't want from same
+ * container.
*/
- if (unlikely(signal->flags & SIGNAL_UNKILLABLE) &&
- !signal_group_exit(signal))
+ if (unkillable_by_sig(signal, signr) &&
+ !signal_group_exit(signal))
continue;

if (sig_kernel_stop(signr)) {
--
1.5.2.5

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/