[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1410468841-320-10-git-send-email-ddstreet@ieee.org>
Date: Thu, 11 Sep 2014 16:54:00 -0400
From: Dan Streetman <ddstreet@...e.org>
To: Minchan Kim <minchan@...nel.org>
Cc: linux-mm@...ck.org, linux-kernel@...r.kernel.org,
Sergey Senozhatsky <sergey.senozhatsky@...il.com>,
Nitin Gupta <ngupta@...are.org>,
Seth Jennings <sjennings@...iantweb.net>,
Andrew Morton <akpm@...ux-foundation.org>,
Dan Streetman <ddstreet@...e.org>
Subject: [PATCH 09/10] zsmalloc: add zs_shrink()
Add function zs_shrink() to implement zs_pool shrinking. This allows
the pool owner to reduce the size of the zs_pool by one zspage, which
contains one or more struct pages. Once the zs_pool is shrunk, the
freed pages are available for system use.
This is used by zswap to limit its total system memory usage to a
user-defined amount, while attempting to keep the most recently stored
pages compressed in memory, and the oldest (or older) pages are evicted
from the zsmalloc zs_pool and written out to swap disk.
Signed-off-by: Dan Streetman <ddstreet@...e.org>
Cc: Minchan Kim <minchan@...nel.org>
---
include/linux/zsmalloc.h | 1 +
mm/zsmalloc.c | 35 +++++++++++++++++++++++++++++++++++
2 files changed, 36 insertions(+)
diff --git a/include/linux/zsmalloc.h b/include/linux/zsmalloc.h
index 2c341d4..07fe84d 100644
--- a/include/linux/zsmalloc.h
+++ b/include/linux/zsmalloc.h
@@ -45,6 +45,7 @@ void zs_destroy_pool(struct zs_pool *pool);
unsigned long zs_malloc(struct zs_pool *pool, size_t size);
void zs_free(struct zs_pool *pool, unsigned long obj);
+int zs_shrink(struct zs_pool *pool);
void *zs_map_object(struct zs_pool *pool, unsigned long handle,
enum zs_mapmode mm);
diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
index 60fd23e..f769c21 100644
--- a/mm/zsmalloc.c
+++ b/mm/zsmalloc.c
@@ -1296,6 +1296,41 @@ void zs_free(struct zs_pool *pool, unsigned long obj)
EXPORT_SYMBOL_GPL(zs_free);
/**
+ * zs_shrink - Shrink the pool
+ * @pool: pool to shrink
+ *
+ * The pool will be shrunk by one zspage, which is some
+ * number of pages in size. On success, the number of freed
+ * pages is returned. On failure, the error is returned.
+ */
+int zs_shrink(struct zs_pool *pool)
+{
+ struct size_class *class;
+ enum fullness_group fullness;
+ struct page *page;
+ int class_idx, ret;
+
+ if (!pool->ops || !pool->ops->evict)
+ return -EINVAL;
+
+ /* if a page is found, the class is locked */
+ page = find_lru_zspage(pool);
+ if (!page)
+ return -ENOENT;
+
+ get_zspage_mapping(page, &class_idx, &fullness);
+ class = &pool->size_class[class_idx];
+
+ /* reclaim_zspage unlocks the class lock */
+ ret = reclaim_zspage(pool, page);
+ if (ret)
+ return ret;
+
+ return class->pages_per_zspage;
+}
+EXPORT_SYMBOL_GPL(zs_shrink);
+
+/**
* zs_map_object - get address of allocated object from handle.
* @pool: pool from which the object was allocated
* @handle: handle returned from zs_malloc
--
1.8.3.1
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists