Set spl_kmem_cache_reclaim=0 to default

Reinstate the correct default behavior of returning the number of objects
in the cache for reclaim.  This behavior was disabled in recent releases
to do occasional reports of spinning in shrink_slabs().  Those issues have
been resolved and can no longer can be reproduced.  See commit 376dc35.

Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: DHE <git@dehacked.net>
Issue #358
Closes #379

diff --git a/module/spl/spl-kmem.c b/module/spl/spl-kmem.c
index 6389dc5aaba76086cc368d844df216b95c942533..0d1da437865e1aa47392c2e91d6709b9ff0c82fb 100644 (file)
--- a/module/spl/spl-kmem.c
+++ b/module/spl/spl-kmem.c
@@ -58,12 +58,13 @@ module_param(spl_kmem_cache_expire, uint, 0644);
 MODULE_PARM_DESC(spl_kmem_cache_expire, "By age (0x1) or low memory (0x2)");
 
 /*
- * KMC_RECLAIM_ONCE is set as the default until zfsonlinux/spl#268 is
- * definitively resolved.  Depending on the system configuration and
- * workload this may increase the likelihood of out of memory events.
- * For those cases it is advised that this option be set to zero.
+ * The default behavior is to report the number of objects remaining in the
+ * cache.  This allows the Linux VM to repeatedly reclaim objects from the
+ * cache when memory is low satisfy other memory allocations.  Alternately,
+ * setting this value to KMC_RECLAIM_ONCE limits how aggressively the cache
+ * is reclaimed.  This may increase the likelihood of out of memory events.
  */
-unsigned int spl_kmem_cache_reclaim = KMC_RECLAIM_ONCE;
+unsigned int spl_kmem_cache_reclaim = 0;
 module_param(spl_kmem_cache_reclaim, uint, 0644);
 MODULE_PARM_DESC(spl_kmem_cache_reclaim, "Single reclaim pass (0x1)");