Re: [PATCH 5/5] ext4: add new ioctl EXT4_IOC_PRECACHE_EXTENTS

[Zheng Liu <gnehzuil.liu@gmail.com>]

On Tue, Jul 16, 2013 at 11:18:03AM -0400, Theodore Ts'o wrote:
> Add a new ioctl which forces the all of the extents in an inode to be
> cached in the extent_status tree.  This is critically important when
> using AIO to a preallocated file, since if we need to read in blocks
> from the extent tree, the io_submit(2) system call becomes
> synchronous, and the AIO is no longer "A", which is bad.
> 
> In addition, for most files which have an external leaf tree block,
> the cost of caching the information in the extent status tree will be
> less than caching the entire 4k block in the buffer cache.  So it is
> generally a win to keep the extent information cached.
> ---
[...]
> @@ -1069,10 +1072,16 @@ static int __es_try_to_reclaim_extents(struct ext4_inode_info *ei,
>  	struct rb_node *node;
>  	struct extent_status *es;
>  	int nr_shrunk = 0;
> +	static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL,
> +	                              DEFAULT_RATELIMIT_BURST);
>  
>  	if (ei->i_es_lru_nr == 0)
>  		return 0;
>  
> +	if (ext4_test_inode_state(inode, EXT4_STATE_EXT_PRECACHED) &&
> +	    __ratelimit(&_rs))
> +		ext4_warning(inode->i_sb, "forced shrink of precached extents");

If I understand correctly, we don't want to reclaim from an inode with
EXT4_STATE_EXT_PRECACHED flag when __ratelimit() returns 0, right?  So
I propose the following code:

        if (ext4_test_inode_state(inode, EXT4_STATE_EXT_PRECACHED) {
                if (!__ratelimit(&_rs))
                        return 0;
                ext4_warning(inode->i_sb, "forced shrink of precached extents");
        }

                                                - Zheng

> +
>  	node = rb_first(&tree->root);
>  	while (node != NULL) {
>  		es = rb_entry(node, struct extent_status, rb_node);