Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1213257

[PATCH 3/6] RDS: fix fmr pool dirty_count

From Santosh Shilimkar <santosh.shilimkar@oracle.com>
Newsgroups linux.kernel
Subject [PATCH 3/6] RDS: fix fmr pool dirty_count
Date 2015-08-25 21:10 +0200
Message-ID <q1rLA-77n-41@gated-at.bofh.it> (permalink)
References <q1rLz-77n-3@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


From: Wengang Wang <wen.gang.wang@oracle.com>

In rds_ib_flush_mr_pool(), dirty_count accounts the clean ones
which is wrong. This can lead to a negative dirty count value.

Lets fix it.

Signed-off-by: Wengang Wang <wen.gang.wang@oracle.com>
Signed-off-by: Santosh Shilimkar <ssantosh@kernel.org>
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
---
 net/rds/ib_rdma.c | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

diff --git a/net/rds/ib_rdma.c b/net/rds/ib_rdma.c
index 7b7aac8..a275b7d 100644
--- a/net/rds/ib_rdma.c
+++ b/net/rds/ib_rdma.c
@@ -528,11 +528,13 @@ static inline unsigned int rds_ib_flush_goal(struct rds_ib_mr_pool *pool, int fr
 /*
  * given an llist of mrs, put them all into the list_head for more processing
  */
-static void llist_append_to_list(struct llist_head *llist, struct list_head *list)
+static unsigned int llist_append_to_list(struct llist_head *llist,
+					 struct list_head *list)
 {
 	struct rds_ib_mr *ibmr;
 	struct llist_node *node;
 	struct llist_node *next;
+	unsigned int count = 0;
 
 	node = llist_del_all(llist);
 	while (node) {
@@ -540,7 +542,9 @@ static void llist_append_to_list(struct llist_head *llist, struct list_head *lis
 		ibmr = llist_entry(node, struct rds_ib_mr, llnode);
 		list_add_tail(&ibmr->unmap_list, list);
 		node = next;
+		count++;
 	}
+	return count;
 }
 
 /*
@@ -581,7 +585,7 @@ static int rds_ib_flush_mr_pool(struct rds_ib_mr_pool *pool,
 	LIST_HEAD(unmap_list);
 	LIST_HEAD(fmr_list);
 	unsigned long unpinned = 0;
-	unsigned int nfreed = 0, ncleaned = 0, free_goal;
+	unsigned int nfreed = 0, dirty_to_clean = 0, free_goal;
 	int ret = 0;
 
 	rds_ib_stats_inc(s_ib_rdma_mr_pool_flush);
@@ -623,8 +627,8 @@ static int rds_ib_flush_mr_pool(struct rds_ib_mr_pool *pool,
 	/* Get the list of all MRs to be dropped. Ordering matters -
 	 * we want to put drop_list ahead of free_list.
 	 */
-	llist_append_to_list(&pool->drop_list, &unmap_list);
-	llist_append_to_list(&pool->free_list, &unmap_list);
+	dirty_to_clean = llist_append_to_list(&pool->drop_list, &unmap_list);
+	dirty_to_clean += llist_append_to_list(&pool->free_list, &unmap_list);
 	if (free_all)
 		llist_append_to_list(&pool->clean_list, &unmap_list);
 
@@ -652,7 +656,6 @@ static int rds_ib_flush_mr_pool(struct rds_ib_mr_pool *pool,
 			kfree(ibmr);
 			nfreed++;
 		}
-		ncleaned++;
 	}
 
 	if (!list_empty(&unmap_list)) {
@@ -678,7 +681,7 @@ static int rds_ib_flush_mr_pool(struct rds_ib_mr_pool *pool,
 	}
 
 	atomic_sub(unpinned, &pool->free_pinned);
-	atomic_sub(ncleaned, &pool->dirty_count);
+	atomic_sub(dirty_to_clean, &pool->dirty_count);
 	atomic_sub(nfreed, &pool->item_count);
 
 out:
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread


Thread

[PATCH 0/6] RDS: Few more fixes Santosh Shilimkar <santosh.shilimkar@oracle.com> - 2015-08-25 21:10 +0200
  [PATCH 6/6] RDS: remove superfluous from rds_ib_alloc_fmr() Santosh Shilimkar <santosh.shilimkar@oracle.com> - 2015-08-25 21:10 +0200
  [PATCH 1/6] RDS: fix the dangling reference to rds_ib_incoming_slab Santosh Shilimkar <santosh.shilimkar@oracle.com> - 2015-08-25 21:10 +0200
  [PATCH 3/6] RDS: fix fmr pool dirty_count Santosh Shilimkar <santosh.shilimkar@oracle.com> - 2015-08-25 21:10 +0200
  [PATCH 2/6] RDS: Fix rds MR reference count in rds_rdma_unuse() Santosh Shilimkar <santosh.shilimkar@oracle.com> - 2015-08-25 21:10 +0200
  [PATCH 5/6] RDS: flush the FMR pool less often Santosh Shilimkar <santosh.shilimkar@oracle.com> - 2015-08-25 21:10 +0200
  Re: [PATCH 0/6] RDS: Few more fixes David Miller <davem@davemloft.net> - 2015-08-26 01:30 +0200

csiph-web