sighook: New utilities for hooking signals

This allows multiple hooks to be installed for a single signal.

3 files changed, 674 insertions, 0 deletions

diff --git a/src/main.c b/src/main.c
index 9d8b206..5dd88e4 100644
--- a/src/main.c
+++ b/src/main.c
@@ -36,6 +36,7 @@
  *     - mtab.[ch]     (parses the system's mount table)
  *     - pwcache.[ch]  (a cache for the user/group tables)
  *     - sanity.h      (sanitizer interfaces)
+ *     - sighook.[ch]  (signal hooks)
  *     - stat.[ch]     (wraps stat(), or statx() on Linux)
  *     - thread.h      (multi-threading)
  *     - trie.[ch]     (a trie set/map implementation)
diff --git a/src/sighook.c b/src/sighook.c
new file mode 100644
index 0000000..745ccb9
--- /dev/null
+++ b/src/sighook.c
@@ -0,0 +1,600 @@
+// Copyright © Tavian Barnes <tavianator@tavianator.com>
+// SPDX-License-Identifier: 0BSD
+
+/**
+ * Dynamic (un)registration of signal handlers.
+ *
+ * Because signal handlers can interrupt any thread at an arbitrary point, they
+ * must be lock-free or risk deadlock.  Therefore, we implement the global table
+ * of signal "hooks" with a simple read-copy-update (RCU) scheme.  Readers get a
+ * reference-counted pointer (struct arc) to the table in a lock-free way, and
+ * release the reference count when finished.
+ *
+ * Updates are managed by struct rcu, which has two slots: one active and one
+ * inactive.  Readers acquire a reference to the active slot.  A single writer
+ * can safely update it by initializing the inactive slot, atomically swapping
+ * the slots, and waiting for the reference count of the newly inactive slot to
+ * drop to zero.  Once it does, the old pointer can be safely freed.
+ */
+
+#include "prelude.h"
+#include "sighook.h"
+#include "alloc.h"
+#include "atomic.h"
+#include "bfstd.h"
+#include "diag.h"
+#include "thread.h"
+#include <errno.h>
+#include <signal.h>
+#include <stdlib.h>
+#include <unistd.h>
+
+#if _POSIX_SEMAPHORES > 0
+#  include <semaphore.h>
+#endif
+
+/**
+ * An atomically reference-counted pointer.
+ */
+struct arc {
+	/** The current reference count (0 means empty). */
+	atomic size_t refs;
+	/** The reference itself. */
+	void *ptr;
+
+#if _POSIX_SEMAPHORES > 0
+	/** A semaphore for arc_wake(). */
+	sem_t sem;
+	/** sem_init() result. */
+	int sem_status;
+#endif
+};
+
+/** Initialize an arc. */
+static void arc_init(struct arc *arc) {
+	atomic_init(&arc->refs, 0);
+	arc->ptr = NULL;
+
+#if _POSIX_SEMAPHORES > 0
+	arc->sem_status = sem_init(&arc->sem, false, 0);
+#endif
+}
+
+/** Get the current refcount. */
+static size_t arc_refs(const struct arc *arc) {
+	return load(&arc->refs, relaxed);
+}
+
+/** Set the pointer in an empty arc. */
+static void arc_set(struct arc *arc, void *ptr) {
+	bfs_assert(arc_refs(arc) == 0);
+	bfs_assert(ptr);
+
+	arc->ptr = ptr;
+	store(&arc->refs, 1, release);
+}
+
+/** Acquire a reference. */
+static void *arc_get(struct arc *arc) {
+	size_t refs = arc_refs(arc);
+	do {
+		if (refs < 1) {
+			return NULL;
+		}
+	} while (!compare_exchange_weak(&arc->refs, &refs, refs + 1, acquire, relaxed));
+
+	return arc->ptr;
+}
+
+/** Release a reference. */
+static void arc_put(struct arc *arc) {
+	size_t refs = fetch_sub(&arc->refs, 1, release);
+
+	if (refs == 1) {
+#if _POSIX_SEMAPHORES > 0
+		if (arc->sem_status == 0 && sem_post(&arc->sem) != 0) {
+			abort();
+		}
+#endif
+	}
+}
+
+/** Wait on the semaphore. */
+static int arc_sem_wait(struct arc *arc) {
+#if _POSIX_SEMAPHORES > 0
+	if (arc->sem_status == 0) {
+		while (sem_wait(&arc->sem) != 0) {
+			bfs_verify(errno == EINTR, "sem_wait(): %s", xstrerror(errno));
+		}
+		return 0;
+	}
+#endif
+
+	return -1;
+}
+
+/** Wait for all references to be released. */
+static void *arc_wait(struct arc *arc) {
+	size_t refs = fetch_sub(&arc->refs, 1, relaxed);
+	bfs_assert(refs > 0);
+
+	--refs;
+	while (refs > 0) {
+		if (arc_sem_wait(arc) == 0) {
+			bfs_assert(arc_refs(arc) == 0);
+			// sem_wait() provides enough ordering, so we can skip the fence
+			goto done;
+		}
+
+		// Some platforms (like macOS) don't support unnamed semaphores,
+		// but we can always busy-wait
+		spin_loop();
+		refs = arc_refs(arc);
+	}
+
+	thread_fence(&arc->refs, acquire);
+
+done:;
+	void *ptr = arc->ptr;
+	arc->ptr = NULL;
+	return ptr;
+}
+
+/** Destroy an arc. */
+static void arc_destroy(struct arc *arc) {
+	bfs_assert(arc_refs(arc) <= 1);
+
+#if _POSIX_SEMAPHORES > 0
+	if (arc->sem_status == 0) {
+		bfs_verify(sem_destroy(&arc->sem) == 0, "sem_destroy(): %s", xstrerror(errno));
+	}
+#endif
+}
+
+/**
+ * A simple read-copy-update memory reclamation scheme.
+ */
+struct rcu {
+	/** The currently active slot. */
+	atomic size_t active;
+	/** The two slots. */
+	struct arc slots[2];
+};
+
+/** Sentinel value for RCU, since arc uses NULL already. */
+static void *RCU_NULL = &RCU_NULL;
+
+/** Initialize an RCU block. */
+static void rcu_init(struct rcu *rcu) {
+	atomic_init(&rcu->active, 0);
+	arc_init(&rcu->slots[0]);
+	arc_init(&rcu->slots[1]);
+	arc_set(&rcu->slots[0], RCU_NULL);
+}
+
+/** Get the active slot. */
+static struct arc *rcu_active(struct rcu *rcu) {
+	size_t i = load(&rcu->active, relaxed);
+	return &rcu->slots[i];
+}
+
+/** Read an RCU-protected pointer. */
+static void *rcu_read(struct rcu *rcu, struct arc **slot) {
+	while (true) {
+		*slot = rcu_active(rcu);
+		void *ptr = arc_get(*slot);
+		if (ptr == RCU_NULL) {
+			return NULL;
+		} else if (ptr) {
+			return ptr;
+		}
+		// Otherwise, the other slot became active; retry
+	}
+}
+
+/** Get the RCU-protected pointer without acquiring a reference. */
+static void *rcu_peek(struct rcu *rcu) {
+	struct arc *arc = rcu_active(rcu);
+	void *ptr = arc->ptr;
+	if (ptr == RCU_NULL) {
+		return NULL;
+	} else {
+		return ptr;
+	}
+}
+
+/** Update an RCU-protected pointer, and return the old one. */
+static void *rcu_update(struct rcu *rcu, void *ptr) {
+	size_t i = load(&rcu->active, relaxed);
+	struct arc *prev = &rcu->slots[i];
+
+	size_t j = i ^ 1;
+	struct arc *next = &rcu->slots[j];
+
+	arc_set(next, ptr ? ptr : RCU_NULL);
+	store(&rcu->active, j, relaxed);
+	return arc_wait(prev);
+}
+
+struct sighook {
+	int sig;
+	sighook_fn *fn;
+	void *arg;
+	enum sigflags flags;
+};
+
+/**
+ * A table of signal hooks.
+ */
+struct sigtable {
+	/** The number of filled slots. */
+	size_t filled;
+	/** The length of the array. */
+	size_t size;
+	/** An array of signal hooks. */
+	struct arc hooks[];
+};
+
+/** Add a hook to a table. */
+static int sigtable_add(struct sigtable *table, struct sighook *hook) {
+	if (!table || table->filled == table->size) {
+		return -1;
+	}
+
+	for (size_t i = 0; i < table->size; ++i) {
+		struct arc *arc = &table->hooks[i];
+		if (arc_refs(arc) == 0) {
+			arc_set(arc, hook);
+			++table->filled;
+			return 0;
+		}
+	}
+
+	return -1;
+}
+
+/** Delete a hook from a table. */
+static int sigtable_del(struct sigtable *table, struct sighook *hook) {
+	for (size_t i = 0; i < table->size; ++i) {
+		struct arc *arc = &table->hooks[i];
+		if (arc->ptr == hook) {
+			arc_wait(arc);
+			--table->filled;
+			return 0;
+		}
+	}
+
+	return -1;
+}
+
+/** Create a bigger copy of a signal table. */
+static struct sigtable *sigtable_grow(struct sigtable *prev) {
+	size_t old_size = prev ? prev->size : 0;
+	size_t new_size = old_size ? 2 * old_size : 1;
+	struct sigtable *table = ALLOC_FLEX(struct sigtable, hooks, new_size);
+	if (!table) {
+		return NULL;
+	}
+
+	table->filled = 0;
+	table->size = new_size;
+	for (size_t i = 0; i < new_size; ++i) {
+		arc_init(&table->hooks[i]);
+	}
+
+	for (size_t i = 0; i < old_size; ++i) {
+		struct sighook *hook = prev->hooks[i].ptr;
+		if (hook) {
+			bfs_verify(sigtable_add(table, hook) == 0);
+		}
+	}
+
+	return table;
+}
+
+/** Free a signal table. */
+static void sigtable_free(struct sigtable *table) {
+	if (!table) {
+		return;
+	}
+
+	for (size_t i = 0; i < table->size; ++i) {
+		struct arc *arc = &table->hooks[i];
+		arc_destroy(arc);
+	}
+	free(table);
+}
+
+/** Add a hook to a signal table, growing it if necessary. */
+static int rcu_sigtable_add(struct rcu *rcu, struct sighook *hook) {
+	struct sigtable *prev = rcu_peek(rcu);
+	if (sigtable_add(prev, hook) == 0) {
+		return 0;
+	}
+
+	struct sigtable *next = sigtable_grow(prev);
+	if (!next) {
+		return -1;
+	}
+
+	bfs_verify(sigtable_add(next, hook) == 0);
+	rcu_update(rcu, next);
+	sigtable_free(prev);
+	return 0;
+}
+
+/** The global table of signal hooks. */
+static struct rcu rcu_sighooks;
+/** The global table of atsigexit() hooks. */
+static struct rcu rcu_exithooks;
+
+/** Mutex for initialization and RCU writer exclusion. */
+static pthread_mutex_t sigmutex = PTHREAD_MUTEX_INITIALIZER;
+
+/** Check if a signal was generated by userspace. */
+static bool is_user_generated(const siginfo_t *info) {
+	// https://pubs.opengroup.org/onlinepubs/9699919799/functions/V2_chap02.html#tag_15_04_03
+	//
+	//     If si_code is SI_USER or SI_QUEUE, or any value less than or
+	//     equal to 0, then the signal was generated by a process ...
+	int code = info->si_code;
+	return code == SI_USER || code == SI_QUEUE || code <= 0;
+}
+
+/** Check if a signal is caused by a fault. */
+static bool is_fault(const siginfo_t *info) {
+	int sig = info->si_signo;
+	if (sig == SIGBUS || sig == SIGFPE || sig == SIGILL || sig == SIGSEGV) {
+		return !is_user_generated(info);
+	} else {
+		return false;
+	}
+}
+
+// https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/signal.h.html
+static const int FATAL_SIGNALS[] = {
+	SIGABRT,
+	SIGALRM,
+	SIGBUS,
+	SIGFPE,
+	SIGHUP,
+	SIGILL,
+	SIGINT,
+	SIGPIPE,
+	SIGQUIT,
+	SIGSEGV,
+	SIGTERM,
+	SIGUSR1,
+	SIGUSR2,
+#ifdef SIGPOLL
+	SIGPOLL,
+#endif
+#ifdef SIGPROF
+	SIGPROF,
+#endif
+#ifdef SIGSYS
+	SIGSYS,
+#endif
+	SIGTRAP,
+#ifdef SIGVTALRM
+	SIGVTALRM,
+#endif
+	SIGXCPU,
+	SIGXFSZ,
+};
+
+/** Check if a signal's default action is to terminate the process. */
+static bool is_fatal(int sig) {
+	for (size_t i = 0; i < countof(FATAL_SIGNALS); ++i) {
+		if (sig == FATAL_SIGNALS[i]) {
+			return true;
+		}
+	}
+
+#ifdef SIGRTMIN
+	// https://pubs.opengroup.org/onlinepubs/9699919799/functions/V2_chap02.html#tag_15_04_03
+	//
+	//     The default actions for the realtime signals in the range
+	//     SIGRTMIN to SIGRTMAX shall be to terminate the process
+	//     abnormally.
+	if (sig >= SIGRTMIN && sig <= SIGRTMAX) {
+		return true;
+	}
+#endif
+
+	return false;
+}
+
+/** Reraise a fatal signal. */
+static noreturn void reraise(int sig) {
+	// Restore the default signal action
+	if (signal(sig, SIG_DFL) == SIG_ERR) {
+		goto fail;
+	}
+
+	// Unblock the signal, since we didn't set SA_NODEFER
+	sigset_t mask;
+	if (sigemptyset(&mask) != 0
+	    || sigaddset(&mask, sig) != 0
+	    || pthread_sigmask(SIG_UNBLOCK, &mask, NULL) != 0) {
+		goto fail;
+	}
+
+	raise(sig);
+fail:
+	abort();
+}
+
+/** Find any matching hooks and run them. */
+static enum sigflags run_hooks(struct rcu *rcu, int sig, siginfo_t *info) {
+	enum sigflags ret = 0;
+	struct arc *slot;
+	struct sigtable *table = rcu_read(rcu, &slot);
+	if (!table) {
+		goto done;
+	}
+
+	for (size_t i = 0; i < table->size; ++i) {
+		struct arc *arc = &table->hooks[i];
+		struct sighook *hook = arc_get(arc);
+		if (!hook) {
+			continue;
+		}
+
+		if (hook->sig == sig || hook->sig == 0) {
+			hook->fn(sig, info, hook->arg);
+			ret |= hook->flags;
+		}
+		arc_put(arc);
+	}
+
+done:
+	arc_put(slot);
+	return ret;
+}
+
+/** Dispatches a signal to the registered handlers. */
+static void sigdispatch(int sig, siginfo_t *info, void *context) {
+	// https://pubs.opengroup.org/onlinepubs/9699919799/functions/V2_chap02.html#tag_15_04_03
+	//
+	//     The behavior of a process is undefined after it returns normally
+	//     from a signal-catching function for a SIGBUS, SIGFPE, SIGILL, or
+	//     SIGSEGV signal that was not generated by kill(), sigqueue(), or
+	//     raise().
+	if (is_fault(info)) {
+		reraise(sig);
+	}
+
+	// https://pubs.opengroup.org/onlinepubs/9699919799/functions/V2_chap02.html#tag_15_04_03
+	//
+	//     After returning from a signal-catching function, the value of
+	//     errno is unspecified if the signal-catching function or any
+	//     function it called assigned a value to errno and the signal-
+	//     catching function did not save and restore the original value of
+	//     errno.
+	int error = errno;
+
+	// Run the normal hooks
+	enum sigflags flags = run_hooks(&rcu_sighooks, sig, info);
+
+	// Run the atsigexit() hooks, if we're exiting
+	if (!(flags & SH_CONTINUE) && is_fatal(sig)) {
+		run_hooks(&rcu_exithooks, sig, info);
+		reraise(sig);
+	}
+
+	errno = error;
+}
+
+/** Make sure our signal handler is installed for a given signal. */
+static int siginit(int sig) {
+	static struct sigaction action = {
+		.sa_sigaction = sigdispatch,
+		.sa_flags = SA_RESTART | SA_SIGINFO,
+	};
+
+	static sigset_t signals;
+	static bool initialized = false;
+
+	if (!initialized) {
+		if (sigemptyset(&signals) != 0
+		    || sigemptyset(&action.sa_mask) != 0) {
+			return -1;
+		}
+		rcu_init(&rcu_sighooks);
+		rcu_init(&rcu_exithooks);
+		initialized = true;
+	}
+
+	int installed = sigismember(&signals, sig);
+	if (installed < 0) {
+		return -1;
+	} else if (installed) {
+		return 0;
+	}
+
+	if (sigaction(sig, &action, NULL) != 0) {
+		return -1;
+	}
+
+	if (sigaddset(&signals, sig) != 0) {
+		return -1;
+	}
+
+	return 0;
+}
+
+/** Shared sighook()/atsigexit() implementation. */
+static struct sighook *sighook_impl(struct rcu *rcu, int sig, sighook_fn *fn, void *arg, enum sigflags flags) {
+	struct sighook *hook = ALLOC(struct sighook);
+	if (!hook) {
+		return NULL;
+	}
+
+	hook->sig = sig;
+	hook->fn = fn;
+	hook->arg = arg;
+	hook->flags = flags;
+
+	if (rcu_sigtable_add(rcu, hook) != 0) {
+		free(hook);
+		return NULL;
+	}
+
+	return hook;
+}
+
+struct sighook *sighook(int sig, sighook_fn *fn, void *arg, enum sigflags flags) {
+	mutex_lock(&sigmutex);
+
+	struct sighook *ret = NULL;
+	if (siginit(sig) != 0) {
+		goto done;
+	}
+
+	ret = sighook_impl(&rcu_sighooks, sig, fn, arg, flags);
+done:
+	mutex_unlock(&sigmutex);
+	return ret;
+}
+
+struct sighook *atsigexit(sighook_fn *fn, void *arg) {
+	mutex_lock(&sigmutex);
+
+	struct sighook *ret = NULL;
+
+	for (size_t i = 0; i < countof(FATAL_SIGNALS); ++i) {
+		if (siginit(FATAL_SIGNALS[i]) != 0) {
+			goto done;
+		}
+	}
+
+#ifdef SIGRTMIN
+	for (int i = SIGRTMIN; i <= SIGRTMAX; ++i) {
+		if (siginit(i) != 0) {
+			goto done;
+		}
+	}
+#endif
+
+	ret = sighook_impl(&rcu_exithooks, 0, fn, arg, 0);
+done:
+	mutex_unlock(&sigmutex);
+	return ret;
+}
+
+void sigunhook(struct sighook *hook) {
+	mutex_lock(&sigmutex);
+
+	struct rcu *rcu = hook->sig ? &rcu_sighooks : &rcu_exithooks;
+	struct sigtable *table = rcu_peek(rcu);
+	bfs_verify(sigtable_del(table, hook) == 0);
+
+	if (table->filled == 0) {
+		rcu_update(rcu, NULL);
+		sigtable_free(table);
+	}
+
+	mutex_unlock(&sigmutex);
+	free(hook);
+}
diff --git a/src/sighook.h b/src/sighook.h
new file mode 100644
index 0000000..74d18c0
--- /dev/null
+++ b/src/sighook.h
@@ -0,0 +1,73 @@
+// Copyright © Tavian Barnes <tavianator@tavianator.com>
+// SPDX-License-Identifier: 0BSD
+
+/**
+ * Signal hooks.
+ */
+
+#ifndef BFS_SIGHOOK_H
+#define BFS_SIGHOOK_H
+
+#include <signal.h>
+
+/**
+ * A dynamic signal hook.
+ */
+struct sighook;
+
+/**
+ * Signal hook flags.
+ */
+enum sigflags {
+	/** Suppress the default action for this signal. */
+	SH_CONTINUE = 1 << 0,
+};
+
+/**
+ * A signal hook callback.  Hooks are executed from a signal handler, so must
+ * only call async-signal-safe functions.
+ *
+ * @param sig
+ *         The signal number.
+ * @param info
+ *         Additional information about the signal.
+ * @param arg
+ *         An arbitrary pointer passed to the hook.
+ */
+typedef void sighook_fn(int sig, siginfo_t *info, void *arg);
+
+/**
+ * Install a hook for a signal.
+ *
+ * @param sig
+ *         The signal to hook.
+ * @param fn
+ *         The function to call.
+ * @param arg
+ *         An argument passed to the function.
+ * @param flags
+ *         Flags for the new hook.
+ * @return
+ *         The installed hook, or NULL on failure.
+ */
+struct sighook *sighook(int sig, sighook_fn *fn, void *arg, enum sigflags flags);
+
+/**
+ * On a best-effort basis, invoke the given hook just before the program is
+ * abnormally terminated by a signal.
+ *
+ * @param fn
+ *         The function to call.
+ * @param arg
+ *         An argument passed to the function.
+ * @return
+ *         The installed hook, or NULL on failure.
+ */
+struct sighook *atsigexit(sighook_fn *fn, void *arg);
+
+/**
+ * Remove a signal hook.
+ */
+void sigunhook(struct sighook *hook);
+
+#endif // BFS_SIGHOOK_H