Messages in this thread | | | Date | Sun, 16 Dec 2012 16:48:59 +0800 | From | Zheng Liu <> | Subject | Re: [PATCH] fadvise: perform WILLNEED readahead in a workqueue |
| |
On Sun, Dec 16, 2012 at 03:15:49PM +1100, Dave Chinner wrote: > On Sun, Dec 16, 2012 at 03:35:49AM +0000, Eric Wong wrote: > > Dave Chinner <david@fromorbit.com> wrote: > > > On Sun, Dec 16, 2012 at 12:25:49AM +0000, Eric Wong wrote: > > > > Alan Cox <alan@lxorguk.ukuu.org.uk> wrote: > > > > > On Sat, 15 Dec 2012 00:54:48 +0000 > > > > > Eric Wong <normalperson@yhbt.net> wrote: > > > > > > > > > > > Applications streaming large files may want to reduce disk spinups and > > > > > > I/O latency by performing large amounts of readahead up front > > > This could also be a use case for an audio/video player. > > Sure, but this can all be handled by a userspace application. If you > want to avoid/batch IO to enable longer spindown times, then you > have to load the file into RAM somewhere, and you don't need special > kernel support for that. > > > So no, there's no difference that matters between the approaches. > > But I think doing this in the kernel is easier for userspace users. > > The kernel provides mechanisms for applications to use. You have not > mentioned anything new that requires a new kernel mechanism to > acheive - you just need to have the knowledge to put the pieces > together properly. People have been solving this same problem for > the last 20 years without needing to tweak fadvise(). Or even having > an fadvise() syscall... > > Nothing about low latency IO or streaming IO is simple or easy, and > changing how readahead works doesn't change that fact. All it does > is change the behaviour of every other application that uses > fadvise() to minimise IO latency....
Hi Dave,
I am wondering this patch might be a good idea to reduce the latency of fadvise() syscall itself. I do a really simple test in my desktop to measure the latency of fadvise syscall. Before applying this patch, fadvise syscall takes 32 microseconds. After applying the patch, it only takes 4 microseconds. (I was surprised that it takes a very long time!)
Actually we observe a latency after using fadvise. But I don't find a proper time to look at this problem. So I guess this patch might be useful to reduce latency.
Regards, - Zheng
| |