[PATCH] __block_write_full_page speedup
authorAndrew Morton <akpm@osdl.org>
Thu, 5 May 2005 23:15:47 +0000 (16:15 -0700)
committerLinus Torvalds <torvalds@ppc970.osdl.org>
Thu, 5 May 2005 23:36:41 +0000 (16:36 -0700)
Remove all those get_bh()'s and put_bh()'s by extending lock_page() to cover
the troublesome regions.

(get_bh() and put_bh() happen every time whereas contention on a page's lock
in there happens basically never).

Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
fs/buffer.c

index bc75f2e..6f2c330 100644 (file)
@@ -1826,7 +1826,6 @@ static int __block_write_full_page(struct inode *inode, struct page *page,
                }
                if (test_clear_buffer_dirty(bh)) {
                        mark_buffer_async_write(bh);
-                       get_bh(bh);
                        last_bh = bh;
                } else {
                        unlock_buffer(bh);
@@ -1839,20 +1838,19 @@ static int __block_write_full_page(struct inode *inode, struct page *page,
         */
        BUG_ON(PageWriteback(page));
        set_page_writeback(page);
-       unlock_page(page);
 
        do {
                struct buffer_head *next = bh->b_this_page;
                if (buffer_async_write(bh)) {
                        submit_bh(WRITE, bh);
                        nr_underway++;
-                       put_bh(bh);
                        if (bh == last_bh)
                                break;
                }
                bh = next;
        } while (bh != head);
        bh = head;
+       unlock_page(page);
 
        err = 0;
 done:
@@ -1894,7 +1892,6 @@ recover:
                if (buffer_mapped(bh) && buffer_dirty(bh)) {
                        lock_buffer(bh);
                        mark_buffer_async_write(bh);
-                       get_bh(bh);
                        last_bh = bh;
                } else {
                        /*
@@ -1914,7 +1911,6 @@ recover:
                        clear_buffer_dirty(bh);
                        submit_bh(WRITE, bh);
                        nr_underway++;
-                       put_bh(bh);
                        if (bh == last_bh)
                                break;
                }