Fix for a bug where only one process (of multiple) blocked on

maxpages on a zone is woken up, with the rest never being woken up as
a result of the ZFLAG_FULL flag being cleared. Wakeup all such blocked
procsses instead. This change introduces a thundering herd, but since
this should be relatively infrequent, optimizing this (by introducing
a count of blocked processes, for example) may be premature.

Reviewd by: ups@

1 files changed, 7 insertions, 2 deletions

diff --git a/sys/vm/uma_core.c b/sys/vm/uma_core.c
index 0bdcd2f..62e7aff 100644
--- a/sys/vm/uma_core.c
+++ b/sys/vm/uma_core.c
@@ -2475,8 +2475,13 @@ uma_zfree_internal(uma_zone_t zone, void *item, void *udata,
 		if (keg->uk_pages < keg->uk_maxpages)
 			keg->uk_flags &= ~UMA_ZFLAG_FULL;
 
-		/* We can handle one more allocation */
-		wakeup_one(keg);
+		/* 
+		 * We can handle one more allocation. Since we're clearing ZFLAG_FULL,
+		 * wake up all procs blocked on pages. This should be uncommon, so 
+		 * keeping this simple for now (rather than adding count of blocked 
+		 * threads etc).
+		 */
+		wakeup(keg);
 	}
 
 	ZONE_UNLOCK(zone);