Add a deferred free mechanism for freeing swap space that does not require

an exclusive object lock.

Previously swap space was freed on a best effort basis when a page that
had valid swap was dirtied, thus invalidating the swap copy.  This may be
done inconsistently and requires the object lock which is not always
convenient.

Instead, track when swap space is present.  The first dirty is responsible
for deleting space or setting PGA_SWAP_FREE which will trigger background
scans to free the swap space.

Simplify the locking in vm_fault_dirty() now that we can reliably identify
the first dirty.

Discussed with:	alc, kib, markj
Differential Revision:	https://reviews.freebsd.org/D22654

1 files changed, 2 insertions, 8 deletions

diff --git a/sys/dev/md/md.c b/sys/dev/md/md.c
index fa122a5cd8e6..da0ce6361856 100644
--- a/sys/dev/md/md.c
+++ b/sys/dev/md/md.c
@@ -1118,10 +1118,7 @@ mdstart_swap(struct md_s *sc, struct bio *bp)
 			}
 
 			vm_page_valid(m);
-			if (m->dirty != VM_PAGE_BITS_ALL) {
-				vm_page_dirty(m);
-				vm_pager_page_unswapped(m);
-			}
+			vm_page_set_dirty(m);
 		} else if (bp->bio_cmd == BIO_DELETE) {
 			if (len == PAGE_SIZE || vm_page_all_valid(m))
 				rv = VM_PAGER_OK;
@@ -1138,10 +1135,7 @@ mdstart_swap(struct md_s *sc, struct bio *bp)
 				/* Page is valid. */
 				if (len != PAGE_SIZE) {
 					pmap_zero_page_area(m, offs, len);
-					if (m->dirty != VM_PAGE_BITS_ALL) {
-						vm_page_dirty(m);
-						vm_pager_page_unswapped(m);
-					}
+					vm_page_set_dirty(m);
 				} else {
 					vm_pager_page_unswapped(m);
 					vm_page_free(m);