On Thu, Sep 27, 2012 at 10:19 AM, Dave Chinner <david@xxxxxxxxxxxxx> wrote: > On Wed, Sep 26, 2012 at 10:53:07AM +0800, Zhi Yong Wu wrote: >> On Tue, Sep 25, 2012 at 5:17 PM, Dave Chinner <david@xxxxxxxxxxxxx> wrote: >> > On Sun, Sep 23, 2012 at 08:56:27PM +0800, zwu.kernel@xxxxxxxxx wrote: >> > I note that the code will always insert range items of a length >> > RANGE_SIZE. This means you have a fixed object granularity and hence >> > you have no need for a range based search. That is, you could use a >> > radix tree where each entry in the radix tree points directly to the >> > range object similar to how the page cache uses a radix tree for >> > indexing pages. That brings the possibility of lockless range item >> > lookups.... >> Great suggestion, but can we temporarily put it in TODO list? because >> it will bring one big code change. > > Sure. I just wanted to point out that there are better choices for > indexing fixed size elements than rb-trees and why it might make > sense to use a different type of tree. Got it, thanks. Moreover, it should also be better to use radix tree to hold hot_inode, not only hot_range. > > Cheers, > > Dave. > -- > Dave Chinner > david@xxxxxxxxxxxxx -- Regards, Zhi Yong Wu -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
