Group short-lived and reclaimable kernel allocations

This patch marks a number of allocations that are either short-lived such as
network buffers or are reclaimable such as inode allocations.  When something
like updatedb is called, long-lived and unmovable kernel allocations tend to
be spread throughout the address space which increases fragmentation.

This patch groups these allocations together as much as possible by adding a
new MIGRATE_TYPE.  The MIGRATE_RECLAIMABLE type is for allocations that can be
reclaimed on demand, but not moved.  i.e.  they can be migrated by deleting
them and re-reading the information from elsewhere.

Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Cc: Andy Whitcroft <apw@shadowen.org>
Cc: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

6 files changed, 17 insertions, 13 deletions

diff --git a/fs/buffer.c b/fs/buffer.c
index a406cfd89e3b..faceb5eecca9 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -3169,7 +3169,8 @@ static void recalc_bh_state(void)
 	
 struct buffer_head *alloc_buffer_head(gfp_t gfp_flags)
 {
-	struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags);
+	struct buffer_head *ret = kmem_cache_zalloc(bh_cachep,
+				set_migrateflags(gfp_flags, __GFP_RECLAIMABLE));
 	if (ret) {
 		INIT_LIST_HEAD(&ret->b_assoc_buffers);
 		get_cpu_var(bh_accounting).nr++;
diff --git a/fs/dcache.c b/fs/dcache.c
index 678d39deb607..7da0cf50873e 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -903,7 +903,7 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
 	struct dentry *dentry;
 	char *dname;
 
-	dentry = kmem_cache_alloc(dentry_cache, GFP_KERNEL); 
+	dentry = kmem_cache_alloc(dentry_cache, GFP_KERNEL);
 	if (!dentry)
 		return NULL;
 
diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c
index 06ab3c10b1b8..a6be78c05dce 100644
--- a/fs/jbd/journal.c
+++ b/fs/jbd/journal.c
@@ -1710,7 +1710,7 @@ static int journal_init_journal_head_cache(void)
 	journal_head_cache = kmem_cache_create("journal_head",
 				sizeof(struct journal_head),
 				0,		/* offset */
-				0,		/* flags */
+				SLAB_TEMPORARY,	/* flags */
 				NULL);		/* ctor */
 	retval = 0;
 	if (journal_head_cache == 0) {
@@ -2006,7 +2006,7 @@ static int __init journal_init_handle_cache(void)
 	jbd_handle_cache = kmem_cache_create("journal_handle",
 				sizeof(handle_t),
 				0,		/* offset */
-				0,		/* flags */
+				SLAB_TEMPORARY,	/* flags */
 				NULL);		/* ctor */
 	if (jbd_handle_cache == NULL) {
 		printk(KERN_EMERG "JBD: failed to create handle cache\n");
diff --git a/fs/jbd/revoke.c b/fs/jbd/revoke.c
index 62e13c8db132..ad2eacf570c6 100644
--- a/fs/jbd/revoke.c
+++ b/fs/jbd/revoke.c
@@ -170,13 +170,15 @@ int __init journal_init_revoke_caches(void)
 {
 	revoke_record_cache = kmem_cache_create("revoke_record",
 					   sizeof(struct jbd_revoke_record_s),
-					   0, SLAB_HWCACHE_ALIGN, NULL);
+					   0,
+					   SLAB_HWCACHE_ALIGN|SLAB_TEMPORARY,
+					   NULL);
 	if (revoke_record_cache == 0)
 		return -ENOMEM;
 
 	revoke_table_cache = kmem_cache_create("revoke_table",
 					   sizeof(struct jbd_revoke_table_s),
-					   0, 0, NULL);
+					   0, SLAB_TEMPORARY, NULL);
 	if (revoke_table_cache == 0) {
 		kmem_cache_destroy(revoke_record_cache);
 		revoke_record_cache = NULL;
diff --git a/fs/proc/base.c b/fs/proc/base.c
index e5d0953d4db1..78fdfea1a7f8 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -492,7 +492,7 @@ static ssize_t proc_info_read(struct file * file, char __user * buf,
 		count = PROC_BLOCK_SIZE;
 
 	length = -ENOMEM;
-	if (!(page = __get_free_page(GFP_KERNEL)))
+	if (!(page = __get_free_page(GFP_TEMPORARY)))
 		goto out;
 
 	length = PROC_I(inode)->op.proc_read(task, (char*)page);
@@ -532,7 +532,7 @@ static ssize_t mem_read(struct file * file, char __user * buf,
 		goto out;
 
 	ret = -ENOMEM;
-	page = (char *)__get_free_page(GFP_USER);
+	page = (char *)__get_free_page(GFP_TEMPORARY);
 	if (!page)
 		goto out;
 
@@ -602,7 +602,7 @@ static ssize_t mem_write(struct file * file, const char __user *buf,
 		goto out;
 
 	copied = -ENOMEM;
-	page = (char *)__get_free_page(GFP_USER);
+	page = (char *)__get_free_page(GFP_TEMPORARY);
 	if (!page)
 		goto out;
 
@@ -788,7 +788,7 @@ static ssize_t proc_loginuid_write(struct file * file, const char __user * buf,
 		/* No partial writes. */
 		return -EINVAL;
 	}
-	page = (char*)__get_free_page(GFP_USER);
+	page = (char*)__get_free_page(GFP_TEMPORARY);
 	if (!page)
 		return -ENOMEM;
 	length = -EFAULT;
@@ -954,7 +954,8 @@ static int do_proc_readlink(struct dentry *dentry, struct vfsmount *mnt,
 			    char __user *buffer, int buflen)
 {
 	struct inode * inode;
-	char *tmp = (char*)__get_free_page(GFP_KERNEL), *path;
+	char *tmp = (char*)__get_free_page(GFP_TEMPORARY);
+	char *path;
 	int len;
 
 	if (!tmp)
@@ -1726,7 +1727,7 @@ static ssize_t proc_pid_attr_write(struct file * file, const char __user * buf,
 		goto out;
 
 	length = -ENOMEM;
-	page = (char*)__get_free_page(GFP_USER);
+	page = (char*)__get_free_page(GFP_TEMPORARY);
 	if (!page)
 		goto out;
 
diff --git a/fs/proc/generic.c b/fs/proc/generic.c
index b5e7155d30d8..1bdb62435758 100644
--- a/fs/proc/generic.c
+++ b/fs/proc/generic.c
@@ -74,7 +74,7 @@ proc_file_read(struct file *file, char __user *buf, size_t nbytes,
 		nbytes = MAX_NON_LFS - pos;
 
 	dp = PDE(inode);
-	if (!(page = (char*) __get_free_page(GFP_KERNEL)))
+	if (!(page = (char*) __get_free_page(GFP_TEMPORARY)))
 		return -ENOMEM;
 
 	while ((nbytes > 0) && !eof) {