[PATCH] mm: dirty_exceeded speedup (e236a166) · Commits · 戴 / test

Ravikiran reports that this variable is bouncing all around nodes on NUMA machines, causing measurable performance problems. Fix that up by only writing to it when it actually changed. And put it in a new cacheline to prevent it sharing with other things (this happened). Signed-off-by:

Ravikiran Thirumalai <kiran@scalex86.org> Signed-off-by:

Andrew Morton <akpm@osdl.org> Signed-off-by:

Linus Torvalds <torvalds@osdl.org>

mm/page-writeback.c

+4 −3

Original line number	Diff line number	Diff line
		@@ -46,7 +46,7 @@
		static long ratelimit_pages = 32;

		static long total_pages; /* The total number of pages in the machine. */
		static int dirty_exceeded; /* Dirty mem may be over limit */
		static int dirty_exceeded __cacheline_aligned_in_smp; /* Dirty mem may be over limit */

		/*
		* When balance_dirty_pages decides that the caller needs to perform some
		@@ -212,6 +212,7 @@ static void balance_dirty_pages(struct address_space *mapping)
		if (nr_reclaimable + wbs.nr_writeback <= dirty_thresh)
		break;

		if (!dirty_exceeded)
		dirty_exceeded = 1;

		/* Note: nr_reclaimable denotes nr_dirty + nr_unstable.
		@@ -234,7 +235,7 @@ static void balance_dirty_pages(struct address_space *mapping)
		blk_congestion_wait(WRITE, HZ/10);
		}

		if (nr_reclaimable + wbs.nr_writeback <= dirty_thresh)
		if (nr_reclaimable + wbs.nr_writeback <= dirty_thresh && dirty_exceeded)
		dirty_exceeded = 0;

		if (writeback_in_progress(bdi))

Admin message