[PATCH] throttle_vm_writeout(): don't loop on GFP_NOFS and GFP_NOIO allocations

throttle_vm_writeout() is designed to wait for the dirty levels to subside. But if the caller holds IO or FS locks, we might be holding up that writeout. So change it to take a single nap to give other devices a chance to clean some memory, then return. Cc: Nick Piggin <nickpiggin@yahoo.com.au> Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> Cc: Kumar Gala <galak@kernel.crashing.org> Cc: Pete Zaitcev <zaitcev@redhat.com> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

[PATCH] throttle_vm_writeout(): don't loop on GFP_NOFS and GFP_NOIO allocations
throttle_vm_writeout() is designed to wait for the dirty levels to subside. But if the caller holds IO or FS locks, we might be holding up that writeout. So change it to take a single nap to give other devices a chance to clean some memory, then return. Cc: Nick Piggin <nickpiggin@yahoo.com.au> Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> Cc: Kumar Gala <galak@kernel.crashing.org> Cc: Pete Zaitcev <zaitcev@redhat.com> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Andrew Morton · Linus Torvalds
1 parent b1a316f6f9
Showing 3 changed files with 13 additions and 4 deletions Side-by-side Diff
include/linux/writeback.h
mm/page-writeback.c
mm/vmscan.c
@@ -84,7 +84,7 @@
 int wakeup_pdflush(long nr_pages);
 void laptop_io_completion(void);
 void laptop_sync_completion(void);
-void throttle_vm_writeout(void);
+void throttle_vm_writeout(gfp_t gfp_mask);
  
 /* These are exported to sysctl. */
 extern int dirty_background_ratio;
@@ -296,11 +296,21 @@
 }
 EXPORT_SYMBOL(balance_dirty_pages_ratelimited_nr);
  
-void throttle_vm_writeout(void)
+void throttle_vm_writeout(gfp_t gfp_mask)
 {
 	long background_thresh;
 	long dirty_thresh;
  
+	if ((gfp_mask & (__GFP_FS|__GFP_IO)) != (__GFP_FS|__GFP_IO)) {
+		/*
+		 * The caller might hold locks which can prevent IO completion
+		 * or progress in the filesystem.  So we cannot just sit here
+		 * waiting for IO to complete.
+		 */
+		congestion_wait(WRITE, HZ/10);
+		return;
+	}
+
         for ( ; ; ) {
 		get_dirty_limits(&background_thresh, &dirty_thresh, NULL);
  
@@ -316,7 +326,6 @@
                 congestion_wait(WRITE, HZ/10);
         }
 }
-
  
 /*
  * writeback at least _min_pages, and keep writing until the amount of dirty
@@ -952,7 +952,7 @@
 		}
 	}
  
-	throttle_vm_writeout();
+	throttle_vm_writeout(sc->gfp_mask);
  
 	atomic_dec(&zone->reclaim_in_progress);
 	return nr_reclaimed;
...	...	@@ -84,7 +84,7 @@
84	84	int wakeup_pdflush(long nr_pages);
85	85	void laptop_io_completion(void);
86	86	void laptop_sync_completion(void);
87		-void throttle_vm_writeout(void);
	87	+void throttle_vm_writeout(gfp_t gfp_mask);
88	88
89	89	/* These are exported to sysctl. */
90	90	extern int dirty_background_ratio;
...	...	@@ -296,11 +296,21 @@
296	296	}
297	297	EXPORT_SYMBOL(balance_dirty_pages_ratelimited_nr);
298	298
299		-void throttle_vm_writeout(void)
	299	+void throttle_vm_writeout(gfp_t gfp_mask)
300	300	{
301	301	long background_thresh;
302	302	long dirty_thresh;
303	303
	304	+ if ((gfp_mask & (__GFP_FS\|__GFP_IO)) != (__GFP_FS\|__GFP_IO)) {
	305	+ /*
	306	+ * The caller might hold locks which can prevent IO completion
	307	+ * or progress in the filesystem. So we cannot just sit here
	308	+ * waiting for IO to complete.
	309	+ */
	310	+ congestion_wait(WRITE, HZ/10);
	311	+ return;
	312	+ }
	313	+
304	314	for ( ; ; ) {
305	315	get_dirty_limits(&background_thresh, &dirty_thresh, NULL);
306	316
...	...	@@ -316,7 +326,6 @@
316	326	congestion_wait(WRITE, HZ/10);
317	327	}
318	328	}
319		-
320	329
321	330	/*
322	331	* writeback at least _min_pages, and keep writing until the amount of dirty
...	...	@@ -952,7 +952,7 @@
952	952	}
953	953	}
954	954
955		- throttle_vm_writeout();
	955	+ throttle_vm_writeout(sc->gfp_mask);
956	956
957	957	atomic_dec(&zone->reclaim_in_progress);
958	958	return nr_reclaimed;