Fix deadlock danger when atomic ops are done under spinlock.

This was a danger only for --disable-spinlocks in combination with
atomic operations unsupported by the current platform.

While atomics.c was careful to signal that a separate semaphore ought
to be used when spinlock emulation is active, spin.c didn't actually
implement that mechanism. That's my (Andres') fault, it seems to have
gotten lost during the development of the atomic operations support.

Fix that issue and add test for nesting atomic operations inside a
spinlock.

Author: Andres Freund
Discussion: https://postgr.es/m/20200605023302.g6v3ydozy5txifji@alap3.anarazel.de
Backpatch: 9.5-

src/backend/storage/lmgr/spin.c

@@ -28,8 +28,24 @@
 
 
 #ifndef HAVE_SPINLOCKS
+
+/*
+ * No TAS, so spinlocks are implemented as PGSemaphores.
+ */
+
+#ifndef HAVE_ATOMICS
+#define NUM_EMULATION_SEMAPHORES (NUM_SPINLOCK_SEMAPHORES + NUM_ATOMICS_SEMAPHORES)
+#else
+#define NUM_EMULATION_SEMAPHORES (NUM_SPINLOCK_SEMAPHORES)
+#endif /* DISABLE_ATOMICS */
+
 PGSemaphore *SpinlockSemaArray;
-#endif
+
+#else							/* !HAVE_SPINLOCKS */
+
+#define NUM_EMULATION_SEMAPHORES 0
+
+#endif							/* HAVE_SPINLOCKS */
 
 /*
  * Report the amount of shared memory needed to store semaphores for spinlock
@@ -38,34 +54,19 @@ PGSemaphore *SpinlockSemaArray;
 Size
 SpinlockSemaSize(void)
 {
-	return SpinlockSemas() * sizeof(PGSemaphore);
+	return NUM_EMULATION_SEMAPHORES * sizeof(PGSemaphore);
 }
 
-#ifdef HAVE_SPINLOCKS
-
 /*
  * Report number of semaphores needed to support spinlocks.
  */
 int
 SpinlockSemas(void)
 {
-	return 0;
+	return NUM_EMULATION_SEMAPHORES;
 }
-#else							/* !HAVE_SPINLOCKS */
 
-/*
- * No TAS, so spinlocks are implemented as PGSemaphores.
- */
-
-
-/*
- * Report number of semaphores needed to support spinlocks.
- */
-int
-SpinlockSemas(void)
-{
-	return NUM_SPINLOCK_SEMAPHORES + NUM_ATOMICS_SEMAPHORES;
-}
+#ifndef HAVE_SPINLOCKS
 
 /*
  * Initialize spinlock emulation.
@@ -92,23 +93,59 @@ SpinlockSemaInit(void)
 /*
  * s_lock.h hardware-spinlock emulation using semaphores
  *
- * We map all spinlocks onto a set of NUM_SPINLOCK_SEMAPHORES semaphores.
- * It's okay to map multiple spinlocks onto one semaphore because no process
- * should ever hold more than one at a time.  We just need enough semaphores
- * so that we aren't adding too much extra contention from that.
+ * We map all spinlocks onto NUM_EMULATION_SEMAPHORES semaphores.  It's okay to
+ * map multiple spinlocks onto one semaphore because no process should ever
+ * hold more than one at a time.  We just need enough semaphores so that we
+ * aren't adding too much extra contention from that.
+ *
+ * There is one exception to the restriction of only holding one spinlock at a
+ * time, which is that it's ok if emulated atomic operations are nested inside
+ * spinlocks. To avoid the danger of spinlocks and atomic using the same sema,
+ * we make sure "normal" spinlocks and atomics backed by spinlocks use
+ * distinct semaphores (see the nested argument to s_init_lock_sema).
  *
  * slock_t is just an int for this implementation; it holds the spinlock
- * number from 1..NUM_SPINLOCK_SEMAPHORES.  We intentionally ensure that 0
+ * number from 1..NUM_EMULATION_SEMAPHORES.  We intentionally ensure that 0
  * is not a valid value, so that testing with this code can help find
  * failures to initialize spinlocks.
  */
 
+static inline void
+s_check_valid(int lockndx)
+{
+	if (unlikely(lockndx <= 0 || lockndx > NUM_EMULATION_SEMAPHORES))
+		elog(ERROR, "invalid spinlock number: %d", lockndx);
+}
+
 void
 s_init_lock_sema(volatile slock_t *lock, bool nested)
 {
 	static uint32 counter = 0;
+	uint32		offset;
+	uint32		sema_total;
+	uint32		idx;
 
-	*lock = ((++counter) % NUM_SPINLOCK_SEMAPHORES) + 1;
+	if (nested)
+	{
+		/*
+		 * To allow nesting atomics inside spinlocked sections, use a
+		 * different spinlock. See comment above.
+		 */
+		offset = 1 + NUM_SPINLOCK_SEMAPHORES;
+		sema_total = NUM_ATOMICS_SEMAPHORES;
+	}
+	else
+	{
+		offset = 1;
+		sema_total = NUM_SPINLOCK_SEMAPHORES;
+	}
+
+	idx = (counter++ % sema_total) + offset;
+
+	/* double check we did things correctly */
+	s_check_valid(idx);
+
+	*lock = idx;
 }
 
 void
@@ -116,8 +153,8 @@ s_unlock_sema(volatile slock_t *lock)
 {
 	int			lockndx = *lock;
 
-	if (lockndx <= 0 || lockndx > NUM_SPINLOCK_SEMAPHORES)
-		elog(ERROR, "invalid spinlock number: %d", lockndx);
+	s_check_valid(lockndx);
+
 	PGSemaphoreUnlock(SpinlockSemaArray[lockndx - 1]);
 }
 
@@ -134,8 +171,8 @@ tas_sema(volatile slock_t *lock)
 {
 	int			lockndx = *lock;
 
-	if (lockndx <= 0 || lockndx > NUM_SPINLOCK_SEMAPHORES)
-		elog(ERROR, "invalid spinlock number: %d", lockndx);
+	s_check_valid(lockndx);
+
 	/* Note that TAS macros return 0 if *success* */
 	return !PGSemaphoreTryLock(SpinlockSemaArray[lockndx - 1]);
 }