timer_thread: do not close pipes around fork

There's actually no need to close the pipes used by the
sleepy timer thread before forking, only to stop the timer
thread itself.

Instead, we only close the parent pipes in the child process,
either via close-on-exec flag or when reinitializing the timer
thread.

This change will be necessary when we allow
rb_wait_for_single_fd and rb_thread_fd_select to wait on the
timer_thread_pipe.normal[0] directly and eliminate timer thread.

I don't anticipate compatibility problems with this change
alone.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63960 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

2 files changed, 51 insertions, 54 deletions

diff --git a/process.c b/process.c
index f1e013a284..f644029641 100644
--- a/process.c
+++ b/process.c
@@ -287,12 +287,30 @@ static ID id_hertz;
 #define ALWAYS_NEED_ENVP 0
 #endif
 
+static void
+assert_close_on_exec(int fd)
+{
+#if VM_CHECK_MODE > 0
+#if defined(HAVE_FCNTL) && defined(F_GETFD) && defined(FD_CLOEXEC)
+        int flags = fcntl(fd, F_GETFD);
+        if (flags == -1) {
+                static const char m[] = "reserved FD closed unexpectedly?\n";
+                write(2, m, sizeof(m) - 1);
+                return;
+        }
+        if (flags & FD_CLOEXEC) return;
+        rb_bug("reserved FD did not have close-on-exec set");
+#else
+        rb_bug("reserved FD without close-on-exec support");
+#endif /* FD_CLOEXEC */
+#endif /* VM_CHECK_MODE */
+}
+
 static inline int
 close_unless_reserved(int fd)
 {
-    /* We should not have reserved FDs at this point */
     if (rb_reserved_fd_p(fd)) { /* async-signal-safe */
-        rb_async_bug_errno("BUG timer thread still running", 0 /* EDOOFUS */);
+        assert_close_on_exec(fd);
         return 0;
     }
     return close(fd); /* async-signal-safe */
diff --git a/thread_pthread.c b/thread_pthread.c
index 3f0ec2e6cc..6420314577 100644
--- a/thread_pthread.c
+++ b/thread_pthread.c
@@ -51,7 +51,9 @@ static void rb_thread_wakeup_timer_thread_low(void);
 #define TIMER_THREAD_SLEEPY  (2|TIMER_THREAD_MASK)
 #define TIMER_THREAD_BUSY    (4|TIMER_THREAD_MASK)
 
-#if defined(HAVE_POLL) && defined(HAVE_FCNTL) && defined(F_GETFL) && defined(F_SETFL) && defined(O_NONBLOCK)
+#if defined(HAVE_POLL) && defined(HAVE_FCNTL) && defined(F_GETFL) && \
+   defined(F_SETFL) && defined(O_NONBLOCK) && \
+   defined(F_GETFD) && defined(F_SETFD) && defined(FD_CLOEXEC)
 /* The timer thread sleeps while only one Ruby thread is running. */
 # define TIMER_IMPL TIMER_THREAD_SLEEPY
 #else
@@ -1199,7 +1201,6 @@ static struct {
 
     /* volatile for signal handler use: */
     volatile rb_pid_t owner_process;
-    rb_atomic_t writing;
 } timer_thread_pipe = {
     {-1, -1},
     {-1, -1}, /* low priority */
@@ -1219,13 +1220,12 @@ async_bug_fd(const char *mesg, int errno_arg, int fd)
 
 /* only use signal-safe system calls here */
 static void
-rb_thread_wakeup_timer_thread_fd(volatile int *fdp)
+rb_thread_wakeup_timer_thread_fd(int fd)
 {
     ssize_t result;
-    int fd = *fdp; /* access fdp exactly once here and do not reread fdp */
 
     /* already opened */
-    if (fd >= 0 && timer_thread_pipe.owner_process == getpid()) {
+    if (fd >= 0) {
 	static const char buff[1] = {'!'};
       retry:
 	if ((result = write(fd, buff, 1)) <= 0) {
@@ -1253,9 +1253,7 @@ rb_thread_wakeup_timer_thread(void)
 {
     /* must be safe inside sighandler, so no mutex */
     if (timer_thread_pipe.owner_process == getpid()) {
-	ATOMIC_INC(timer_thread_pipe.writing);
-	rb_thread_wakeup_timer_thread_fd(&timer_thread_pipe.normal[1]);
-	ATOMIC_DEC(timer_thread_pipe.writing);
+	rb_thread_wakeup_timer_thread_fd(timer_thread_pipe.normal[1]);
     }
 }
 
@@ -1263,9 +1261,7 @@ static void
 rb_thread_wakeup_timer_thread_low(void)
 {
     if (timer_thread_pipe.owner_process == getpid()) {
-	ATOMIC_INC(timer_thread_pipe.writing);
-	rb_thread_wakeup_timer_thread_fd(&timer_thread_pipe.low[1]);
-	ATOMIC_DEC(timer_thread_pipe.writing);
+	rb_thread_wakeup_timer_thread_fd(timer_thread_pipe.low[1]);
     }
 }
 
@@ -1303,9 +1299,9 @@ consume_communication_pipe(int fd)
 #define CLOSE_INVALIDATE(expr) \
     close_invalidate(&timer_thread_pipe.expr,"close_invalidate: "#expr)
 static void
-close_invalidate(volatile int *fdp, const char *msg)
+close_invalidate(int *fdp, const char *msg)
 {
-    int fd = *fdp; /* access fdp exactly once here and do not reread fdp */
+    int fd = *fdp;
 
     *fdp = -1;
     if (close(fd) < 0) {
@@ -1333,6 +1329,12 @@ setup_communication_pipe_internal(int pipes[2])
 {
     int err;
 
+    if (pipes[0] >= 0 || pipes[1] >= 0) {
+        VM_ASSERT(pipes[0] >= 0);
+        VM_ASSERT(pipes[1] >= 0);
+        return 0;
+    }
+
     err = rb_cloexec_pipe(pipes);
     if (err != 0) {
 	rb_warn("pipe creation failed for timer: %s, scheduling broken",
@@ -1350,20 +1352,20 @@ setup_communication_pipe_internal(int pipes[2])
 static int
 setup_communication_pipe(void)
 {
-    VM_ASSERT(timer_thread_pipe.owner_process == 0);
-    VM_ASSERT(timer_thread_pipe.normal[0] == -1);
-    VM_ASSERT(timer_thread_pipe.normal[1] == -1);
-    VM_ASSERT(timer_thread_pipe.low[0] == -1);
-    VM_ASSERT(timer_thread_pipe.low[1] == -1);
+    rb_pid_t owner = timer_thread_pipe.owner_process;
+
+    if (owner && owner != getpid()) {
+        CLOSE_INVALIDATE(normal[0]);
+        CLOSE_INVALIDATE(normal[1]);
+        CLOSE_INVALIDATE(low[0]);
+        CLOSE_INVALIDATE(low[1]);
+    }
 
     if (setup_communication_pipe_internal(timer_thread_pipe.normal) < 0) {
 	return errno;
     }
     if (setup_communication_pipe_internal(timer_thread_pipe.low) < 0) {
-	int e = errno;
-	CLOSE_INVALIDATE(normal[0]);
-	CLOSE_INVALIDATE(normal[1]);
-	return e;
+	return errno;
     }
 
     return 0;
@@ -1532,10 +1534,6 @@ thread_timer(void *p)
         /* wait */
 	timer_thread_sleep(vm);
     }
-#if TIMER_IMPL == TIMER_THREAD_SLEEPY
-    CLOSE_INVALIDATE(normal[0]);
-    CLOSE_INVALIDATE(low[0]);
-#endif
 #if TIMER_IMPL == TIMER_THREAD_BUSY
     rb_native_mutex_unlock(&timer_thread_lock);
     rb_native_cond_destroy(&timer_thread_cond);
@@ -1623,12 +1621,6 @@ rb_thread_create_timer_thread(void)
 		rb_warn("timer thread stack size: system default");
 	    }
 	    VM_ASSERT(err == 0);
-#if TIMER_IMPL == TIMER_THREAD_SLEEPY
-	    CLOSE_INVALIDATE(normal[0]);
-	    CLOSE_INVALIDATE(normal[1]);
-	    CLOSE_INVALIDATE(low[0]);
-	    CLOSE_INVALIDATE(low[1]);
-#endif /* TIMER_THREAD_SLEEPY */
 	    return;
 	}
 #if TIMER_IMPL == TIMER_THREAD_SLEEPY
@@ -1649,31 +1641,18 @@ native_stop_timer_thread(void)
     if (TT_DEBUG) fprintf(stderr, "stop timer thread\n");
     if (stopped) {
 #if TIMER_IMPL == TIMER_THREAD_SLEEPY
-	/* prevent wakeups from signal handler ASAP */
-	timer_thread_pipe.owner_process = 0;
-
-	/*
-	 * however, the above was not enough: the FD may already be
-	 * captured and in the middle of a write while we are running,
-	 * so wait for that to finish:
-	 */
-	while (ATOMIC_CAS(timer_thread_pipe.writing, (rb_atomic_t)0, 0)) {
-	    native_thread_yield();
-	}
-
-	/* stop writing ends of pipes so timer thread notices EOF */
-	CLOSE_INVALIDATE(normal[1]);
-	CLOSE_INVALIDATE(low[1]);
+	/* kick timer thread out of sleep */
+	rb_thread_wakeup_timer_thread_fd(timer_thread_pipe.normal[1]);
 #endif
 
 	/* timer thread will stop looping when system_working <= 0: */
 	native_thread_join(timer_thread.id);
 
-#if TIMER_IMPL == TIMER_THREAD_SLEEPY
-	/* timer thread will close the read end on exit: */
-	VM_ASSERT(timer_thread_pipe.normal[0] == -1);
-	VM_ASSERT(timer_thread_pipe.low[0] == -1);
-#endif
+	/*
+	 * don't care if timer_thread_pipe may fill up at this point.
+	 * If we restart timer thread, signals will be processed, if
+	 * we don't, it's because we're in a different child
+	 */
 
 	if (TT_DEBUG) fprintf(stderr, "joined timer thread\n");
 	timer_thread.created = 0;