Re: WAPBL/cache flush and mfi(4)

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Fri, Aug 24, 2012 at 06:10:21PM +0000, Michael van Elst wrote:
> bouyer%antioche.eu.org@localhost (Manuel Bouyer) writes:
> 
> >How do you think this should be handled ? have a sysctl to control
> >mfi's behavior per-controller (eventually with default depending on BBU
> >status), or have a vfs.wapbl.flush_disk_cache per mount-point
> >(with a mount option maybe) ? 
> 
> WAPBL really needs the cache flush operations, so while disabling it
> for debugging is probably useful, I don't think that we need to
> offer finer controls.
> 
> You can ignore the flush operation when the controller has a working BBU
> except for shutdown/suspend.

But the mfi(4) driver doesn't know if the request is from a WAPBL
operation, or something else.

> 
> For controllers without BBU, there should be a limit on how much write data
> it is allowed to keep dirty in the cache, so that a flush can finish
> quickly.

But this is hard to track down. The controller doesn't notify when
he decides to flush some data.

> The standard drivers always turn off write caching when the BBU
> is not working.

No, this is something the user can control directly in firmware,
at last with mfi(4) controllers.

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--