[PATCH] invalidate_bdev() speedup
authorAndrew Morton <akpm@osdl.org>
Sun, 30 Jul 2006 10:03:28 +0000 (03:03 -0700)
committerLinus Torvalds <torvalds@g5.osdl.org>
Mon, 31 Jul 2006 20:28:38 +0000 (13:28 -0700)
We can immediately bail from invalidate_bdev() if the blockdev has no
pagecache.

This solves the huge IPI storms which hald is causing on the big ia64
machines when it polls CDROM drives.

Acked-by: Jes Sorensen <jes@sgi.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
fs/buffer.c

index 3660dcb975912d5a831a7940192df8f53354024d..71649ef9b6586696c695e2d80c000ceedaa51282 100644 (file)
@@ -470,13 +470,18 @@ out:
    pass does the actual I/O. */
 void invalidate_bdev(struct block_device *bdev, int destroy_dirty_buffers)
 {
+       struct address_space *mapping = bdev->bd_inode->i_mapping;
+
+       if (mapping->nrpages == 0)
+               return;
+
        invalidate_bh_lrus();
        /*
         * FIXME: what about destroy_dirty_buffers?
         * We really want to use invalidate_inode_pages2() for
         * that, but not until that's cleaned up.
         */
-       invalidate_inode_pages(bdev->bd_inode->i_mapping);
+       invalidate_inode_pages(mapping);
 }
 
 /*