ANDROID: mm: lru_cache_disable skips lru cache drainnig

lru_cache_disable is not trivial cost since it should run work
from every cores in the system. Thus, repeated call of the
function whenever alloc_contig_range in the cma's allocation loop
is called is expensive.

This patch makes the lru_cache_disable smarter in that it will
not run __lru_add_drain_all since it knows the cache was already
disabled by someone else.
With that, user of alloc_contig_range can disable the lru cache
in advance in their context so that subsequent alloc_contig_range
for user's operation will avoid the costly function call.

Bug: 313795505
Change-Id: Icbb0e6dbf74644d45f562fd1d845888ca1f1f347
Signed-off-by: Richard Chang <richardycc@google.com>

mm/swap.c

@@ -912,6 +912,7 @@ void lru_add_drain_all(void)
 #endif /* CONFIG_SMP */
 
 atomic_t lru_disable_count = ATOMIC_INIT(0);
+EXPORT_SYMBOL_GPL(lru_disable_count);
 
 /*
  * lru_cache_disable() needs to be called before we start compiling
@@ -923,7 +924,12 @@ atomic_t lru_disable_count = ATOMIC_INIT(0);
  */
 void lru_cache_disable(void)
 {
-	atomic_inc(&lru_disable_count);
+	/*
+	 * If someone is already disabled lru_cache, just return with
+	 * increasing the lru_disable_count.
+	 */
+	if (atomic_inc_not_zero(&lru_disable_count))
+		return;
 	/*
 	 * Readers of lru_disable_count are protected by either disabling
 	 * preemption or rcu_read_lock:
@@ -943,7 +949,9 @@ void lru_cache_disable(void)
 #else
 	lru_add_and_bh_lrus_drain();
 #endif
+	atomic_inc(&lru_disable_count);
 }
+EXPORT_SYMBOL_GPL(lru_cache_disable);
 
 /**
  * release_pages - batched put_page()