signal: requeuing undeliverable signals

Kyle Huey recently reported[1] that rr gets confused if SIGKILL prevents
ptrace_signal from delivering a signal, as the kernel setups up a signal
frame for a signal that rr did not have a chance to observe with ptrace.

In looking into it I found a couple of bugs and a quality of
implementation issue.

- The test for signal_group_exit should be inside the for loop in get_signal.
- Signals should be requeued on the same queue they were dequeued from.
- When a fatal signal is pending ptrace_signal should not return another
signal for delivery.

Kyle Huey has verified[2] an earlier version of this change.

I have reworked things one more time to completely fix the issues
raised, and to keep the code maintainable long term.

I have smoke tested this code and combined with a careful review I
expect this code to work fine. Kyle if you can double check that
my last round of changes still works for rr I would appreciate it.

Eric W. Biederman (3):
signal: In get_signal test for signal_group_exit every time through the loop
signal: Requeue signals in the appropriate queue
signal: Requeue ptrace signals

fs/signalfd.c | 5 +++--
include/linux/sched/signal.h | 7 ++++---
kernel/signal.c | 44 ++++++++++++++++++++++++++------------------
3 files changed, 33 insertions(+), 23 deletions(-)

[1] https://lkml.kernel.org/r/[e-mail]
[2] https://lkml.kernel.org/r/CAP045ApAX725ZfujaK-jJNkfCo5s+oVFpBvNfPJk+DKY8K7d=[e-mail]
Tested-by: Kyle Huey <[e-mail]>
Link: https://lkml.kernel.org/r/[e-mail]
Signed-off-by: "Eric W. Biederman" <[e-mail]>

fs/signalfd.c

···
 static ssize_t signalfd_dequeue(struct signalfd_ctx *ctx, kernel_siginfo_t *info,
 				int nonblock)
 {
+	enum pid_type type;
 	ssize_t ret;
 	DECLARE_WAITQUEUE(wait, current);
 
 	spin_lock_irq(&current->sighand->siglock);
-	ret = dequeue_signal(current, &ctx->sigmask, info);
+	ret = dequeue_signal(current, &ctx->sigmask, info, &type);
 	switch (ret) {
 	case 0:
 		if (!nonblock)
···
 	add_wait_queue(&current->sighand->signalfd_wqh, &wait);
 	for (;;) {
 		set_current_state(TASK_INTERRUPTIBLE);
-		ret = dequeue_signal(current, &ctx->sigmask, info);
+		ret = dequeue_signal(current, &ctx->sigmask, info, &type);
 		if (ret != 0)
 			break;
 		if (signal_pending(current)) {

include/linux/sched/signal.h

···
 extern void flush_signals(struct task_struct *);
 extern void ignore_signals(struct task_struct *);
 extern void flush_signal_handlers(struct task_struct *, int force_default);
-extern int dequeue_signal(struct task_struct *task,
-			  sigset_t *mask, kernel_siginfo_t *info);
+extern int dequeue_signal(struct task_struct *task, sigset_t *mask,
+			  kernel_siginfo_t *info, enum pid_type *type);
 
 static inline int kernel_dequeue_signal(void)
 {
 	struct task_struct *task = current;
 	kernel_siginfo_t __info;
+	enum pid_type __type;
 	int ret;
 
 	spin_lock_irq(&task->sighand->siglock);
-	ret = dequeue_signal(task, &task->blocked, &__info);
+	ret = dequeue_signal(task, &task->blocked, &__info, &__type);
 	spin_unlock_irq(&task->sighand->siglock);
 
 	return ret;

kernel/signal.c

···
  *
  * All callers have to hold the siglock.
  */
-int dequeue_signal(struct task_struct *tsk, sigset_t *mask, kernel_siginfo_t *info)
+int dequeue_signal(struct task_struct *tsk, sigset_t *mask,
+		   kernel_siginfo_t *info, enum pid_type *type)
 {
 	bool resched_timer = false;
 	int signr;
···
 	/* We only dequeue private signals from ourselves, we don't let
 	 * signalfd steal them
 	 */
+	*type = PIDTYPE_PID;
 	signr = __dequeue_signal(&tsk->pending, mask, info, &resched_timer);
 	if (!signr) {
+		*type = PIDTYPE_TGID;
 		signr = __dequeue_signal(&tsk->signal->shared_pending,
 					 mask, info, &resched_timer);
 #ifdef CONFIG_POSIX_TIMERS
···
 	freezable_schedule();
 }
 
-static int ptrace_signal(int signr, kernel_siginfo_t *info)
+static int ptrace_signal(int signr, kernel_siginfo_t *info, enum pid_type type)
 {
 	/*
 	 * We do not check sig_kernel_stop(signr) but set this marker
···
 	}
 
 	/* If the (new) signal is now blocked, requeue it.  */
-	if (sigismember(&current->blocked, signr)) {
-		send_signal(signr, info, current, PIDTYPE_PID);
+	if (sigismember(&current->blocked, signr) ||
+	    fatal_signal_pending(current)) {
+		send_signal(signr, info, current, type);
 		signr = 0;
 	}
 
···
 		goto relock;
 	}
 
-	/* Has this task already been marked for death? */
-	if (signal_group_exit(signal)) {
-		ksig->info.si_signo = signr = SIGKILL;
-		sigdelset(&current->pending.signal, SIGKILL);
-		trace_signal_deliver(SIGKILL, SEND_SIG_NOINFO,
-				&sighand->action[SIGKILL - 1]);
-		recalc_sigpending();
-		goto fatal;
-	}
-
 	for (;;) {
 		struct k_sigaction *ka;
+		enum pid_type type;
+
+		/* Has this task already been marked for death? */
+		if (signal_group_exit(signal)) {
+			ksig->info.si_signo = signr = SIGKILL;
+			sigdelset(&current->pending.signal, SIGKILL);
+			trace_signal_deliver(SIGKILL, SEND_SIG_NOINFO,
+				&sighand->action[SIGKILL - 1]);
+			recalc_sigpending();
+			goto fatal;
+		}
 
 		if (unlikely(current->jobctl & JOBCTL_STOP_PENDING) &&
 		    do_signal_stop(0))
···
 		 * so that the instruction pointer in the signal stack
 		 * frame points to the faulting instruction.
 		 */
+		type = PIDTYPE_PID;
 		signr = dequeue_synchronous_signal(&ksig->info);
 		if (!signr)
-			signr = dequeue_signal(current, &current->blocked, &ksig->info);
+			signr = dequeue_signal(current, &current->blocked,
+					       &ksig->info, &type);
 
 		if (!signr)
 			break; /* will return 0 */
 
 		if (unlikely(current->ptrace) && (signr != SIGKILL) &&
 		    !(sighand->action[signr -1].sa.sa_flags & SA_IMMUTABLE)) {
-			signr = ptrace_signal(signr, &ksig->info);
+			signr = ptrace_signal(signr, &ksig->info, type);
 			if (!signr)
 				continue;
 		}
···
 	ktime_t *to = NULL, timeout = KTIME_MAX;
 	struct task_struct *tsk = current;
 	sigset_t mask = *which;
+	enum pid_type type;
 	int sig, ret = 0;
 
 	if (ts) {
···
 	signotset(&mask);
 
 	spin_lock_irq(&tsk->sighand->siglock);
-	sig = dequeue_signal(tsk, &mask, info);
+	sig = dequeue_signal(tsk, &mask, info, &type);
 	if (!sig && timeout) {
 		/*
 		 * None ready, temporarily unblock those we're interested
···
 		spin_lock_irq(&tsk->sighand->siglock);
 		__set_task_blocked(tsk, &tsk->real_blocked);
 		sigemptyset(&tsk->real_blocked);
-		sig = dequeue_signal(tsk, &mask, info);
+		sig = dequeue_signal(tsk, &mask, info, &type);
 	}
 	spin_unlock_irq(&tsk->sighand->siglock);