It's a straight take from ext2, and seems to pass the seven-second
test. You should get a 15-20% drop in CPU load from this.
Could it please be merged up sometime?
--- 2.5.31/fs/jfs/inode.c~jfs-bio Wed Aug 21 13:12:03 2002
+++ 2.5.31-akpm/fs/jfs/inode.c Wed Aug 21 13:12:03 2002
@@ -18,6 +18,7 @@
*/
#include <linux/fs.h>
+#include <linux/mpage.h>
#include <linux/buffer_head.h>
#include "jfs_incore.h"
#include "jfs_filsys.h"
@@ -271,9 +272,20 @@ static int jfs_writepage(struct page *pa
return block_write_full_page(page, jfs_get_block);
}
+static int jfs_writepages(struct address_space *mapping, int *nr_to_write)
+{
+ return mpage_writepages(mapping, nr_to_write, jfs_get_block);
+}
+
static int jfs_readpage(struct file *file, struct page *page)
{
- return block_read_full_page(page, jfs_get_block);
+ return mpage_readpage(page, jfs_get_block);
+}
+
+static int jfs_readpages(struct address_space *mapping,
+ struct list_head *pages, unsigned nr_pages)
+{
+ return mpage_readpages(mapping, pages, nr_pages, jfs_get_block);
}
static int jfs_prepare_write(struct file *file,
@@ -308,7 +320,9 @@ static int jfs_direct_IO(int rw, struct
struct address_space_operations jfs_aops = {
.readpage = jfs_readpage,
+ .readpages = jfs_readpages,
.writepage = jfs_writepage,
+ .writepages = jfs_writepages,
.sync_page = block_sync_page,
.prepare_write = jfs_prepare_write,
.commit_write = generic_commit_write,
.
_______________________________________________
Jfs-discussion mailing list
[EMAIL PROTECTED]
http://www-124.ibm.com/developerworks/oss/mailman/listinfo/jfs-discussion
