mm/page_cache_limit: add support for droping caches for target node

hulk inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4HOXK

------------------------------------------

Add a proc interface to drop cache in target node.
Signed-off-by: NChen Wandun <chenwandun@huawei.com>
Reviewed-by: NKefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

fs/drop_caches.c

@@ -9,12 +9,17 @@
 #include <linux/writeback.h>
 #include <linux/sysctl.h>
 #include <linux/gfp.h>
+
+#ifdef CONFIG_SHRINK_PAGECACHE
+#include <linux/page_cache_limit.h>
+#endif
+
 #include "internal.h"
 
 /* A global variable is a bit ugly, but it keeps the code simple */
 int sysctl_drop_caches;
 
-static void drop_pagecache_sb(struct super_block *sb, void *unused)
+static void drop_pagecache_sb(struct super_block *sb, void *nid)
 {
 	struct inode *inode, *toput_inode = NULL;
 
@@ -35,7 +40,12 @@ static void drop_pagecache_sb(struct super_block *sb, void *unused)
 		spin_unlock(&inode->i_lock);
 		spin_unlock(&sb->s_inode_list_lock);
 
-		invalidate_mapping_pages(inode->i_mapping, 0, -1);
+		if (!nid)
+			invalidate_mapping_pages(inode->i_mapping, 0, -1);
+		else
+			node_invalidate_mapping_pages(inode->i_mapping,
+						      *(int *)nid, 0, -1);
+
 		iput(toput_inode);
 		toput_inode = inode;
 
@@ -74,3 +84,25 @@ int drop_caches_sysctl_handler(struct ctl_table *table, int write,
 	}
 	return 0;
 }
+
+#ifdef CONFIG_SHRINK_PAGECACHE
+int proc_shrink_node_caches(struct ctl_table *table, int write,
+		void __user *buffer, size_t *lenp, loff_t *ppos)
+{
+	int ret;
+
+	ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
+	if (ret || !write)
+		return ret;
+
+	if (node_to_shrink >= MAX_NUMNODES)
+		return -EINVAL;
+
+	if (!node_isset(node_to_shrink, node_states[N_MEMORY]))
+		return 0;
+
+	iterate_supers(drop_pagecache_sb, &node_to_shrink);
+
+	return 0;
+}
+#endif

include/linux/fs.h

@@ -2613,6 +2613,15 @@ extern bool is_bad_inode(struct inode *);
 unsigned long invalidate_mapping_pages(struct address_space *mapping,
 					pgoff_t start, pgoff_t end);
 
+#ifdef CONFIG_SHRINK_PAGECACHE
+unsigned long node_invalidate_mapping_pages(struct address_space *mapping,
+					int nid, pgoff_t start, pgoff_t end);
+#else
+static inline unsigned long
+node_invalidate_mapping_pages(struct address_space *mapping, int nid,
+		pgoff_t start, pgoff_t end) { return 0; }
+#endif
+
 void invalidate_mapping_pagevec(struct address_space *mapping,
 				pgoff_t start, pgoff_t end,
 				unsigned long *nr_pagevec);

include/linux/page_cache_limit.h

@@ -12,6 +12,7 @@ enum page_cache_reclaim_flag {
 extern int pagecache_reclaim_enable;
 extern int pagecache_limit_ratio;
 extern int pagecache_reclaim_ratio;
+extern int node_to_shrink;
 
 int proc_page_cache_limit(struct ctl_table *table, int write,
 		void __user *buffer, size_t *lenp, loff_t *ppos);
@@ -20,6 +21,8 @@ unsigned long __shrink_node_page_cache(int nid, gfp_t mask,
 void kpagecache_limitd_stop(int nid);
 int kpagecache_limitd_run(int nid);
 void wakeup_all_kpagecache_limitd(void);
+int proc_shrink_node_caches(struct ctl_table *table, int write,
+		void __user *buffer, size_t *lenp, loff_t *ppos);
 #else
 static inline void kpagecache_limitd_stop(int nid) {}
 static inline int kpagecache_limitd_run(int nid) { return 0; }

kernel/sysctl.c

@@ -3224,6 +3224,14 @@ static struct ctl_table vm_table[] = {
 		.extra1		= SYSCTL_ZERO,
 		.extra2		= (void *)&one_hundred,
 	},
+	{
+		.procname	= "node_drop_caches",
+		.data		= &node_to_shrink,
+		.maxlen		= sizeof(node_to_shrink),
+		.mode		= 0600,
+		.proc_handler	= proc_shrink_node_caches,
+		.extra1		= SYSCTL_ZERO,
+	},
 #endif
 	{ }
 };

mm/page_cache_limit.c

@@ -5,11 +5,13 @@
 #include <linux/module.h>
 #include <linux/err.h>
 #include <linux/swap.h>
+#include <linux/fs.h>
 #include <linux/page_cache_limit.h>
 
 int pagecache_reclaim_enable;
 int pagecache_limit_ratio;
 int pagecache_reclaim_ratio;
+int node_to_shrink;
 
 static unsigned long pagecache_limit_pages;
 static unsigned long node_pagecache_limit_pages[MAX_NUMNODES];

mm/truncate.c

@@ -465,7 +465,7 @@ void truncate_inode_pages_final(struct address_space *mapping)
 EXPORT_SYMBOL(truncate_inode_pages_final);
 
 static unsigned long __invalidate_mapping_pages(struct address_space *mapping,
-		pgoff_t start, pgoff_t end, unsigned long *nr_pagevec)
+		pgoff_t start, pgoff_t end, unsigned long *nr_pagevec, int nid)
 {
 	pgoff_t indices[PAGEVEC_SIZE];
 	struct pagevec pvec;
@@ -487,6 +487,10 @@ static unsigned long __invalidate_mapping_pages(struct address_space *mapping,
 							     page);
 				continue;
 			}
+
+			if (nid != NUMA_NO_NODE && page_to_nid(page) != nid)
+				continue;
+
 			index += thp_nr_pages(page) - 1;
 
 			ret = invalidate_inode_page(page);
@@ -529,10 +533,34 @@ static unsigned long __invalidate_mapping_pages(struct address_space *mapping,
 unsigned long invalidate_mapping_pages(struct address_space *mapping,
 		pgoff_t start, pgoff_t end)
 {
-	return __invalidate_mapping_pages(mapping, start, end, NULL);
+	return __invalidate_mapping_pages(mapping, start, end, NULL, NUMA_NO_NODE);
 }
 EXPORT_SYMBOL(invalidate_mapping_pages);
 
+
+/**
+ * node_invalidate_mapping_pages - Invalidate all the unlocked pages in @nid of one inode
+ * @mapping: the address_space which holds the pages to invalidate
+ * @nid: pages belong to this node will be invalidate
+ * @start: the offset 'from' which to invalidate
+ * @end: the offset 'to' which to invalidate (inclusive)
+ *
+ * This function only removes the unlocked pages, if you want to
+ * remove all the pages of one inode, you must call truncate_inode_pages.
+ *
+ * node_invalidate_mapping_pages() will not block on IO activity. It will not
+ * invalidate pages which are dirty, locked, under writeback or mapped into
+ * pagetables.
+ *
+ * Return: the number of the pages that were invalidated
+ */
+#ifdef CONFIG_SHRINK_PAGECACHE
+unsigned long node_invalidate_mapping_pages(struct address_space *mapping,
+		int nid, pgoff_t start, pgoff_t end)
+{
+	return __invalidate_mapping_pages(mapping, start, end, NULL, nid);
+}
+#endif
 /**
  * This helper is similar with the above one, except that it accounts for pages
  * that are likely on a pagevec and count them in @nr_pagevec, which will used by
@@ -541,7 +569,7 @@ EXPORT_SYMBOL(invalidate_mapping_pages);
 void invalidate_mapping_pagevec(struct address_space *mapping,
 		pgoff_t start, pgoff_t end, unsigned long *nr_pagevec)
 {
-	__invalidate_mapping_pages(mapping, start, end, nr_pagevec);
+	__invalidate_mapping_pages(mapping, start, end, nr_pagevec, NUMA_NO_NODE);
 }
 
 /*