Re: [PATCH v2] laptop-mode: Make flushes per-device

[Matthew Garrett <mjg59@srcf.ucam.org>]

On Mon, Dec 14, 2009 at 08:58:06PM +0100, Jens Axboe wrote:
> On Mon, Dec 14 2009, Matthew Garrett wrote:
> > One of the features of laptop-mode is that it forces a writeout of dirty
> > pages if something else triggers a physical read or write from a device.
> > The current implementation flushes pages on all devices, rather than only
> > the one that triggered the flush. This patch alters the behaviour so that
> > only the recently accessed block device is flushed, preventing other
> > disks being spun up for no terribly good reason.
> 
> A few comments... Perhaps the timer deletion should go into the backing,
> since that is where it's sitting?

That was for symmetry with the setup, but I'm not married to it.

> Also, I think it would be cleaner to queue the flush work from the timer
> on the per-bdi thread, instead of having a work struct allocated and
> using that work item to simply call bdi writeback instead.

I had a vague recollection of context awkwardness, but I may have been 
wrong there. I'll try that.

-- 
Matthew Garrett | mjg59@srcf.ucam.org