[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20080329090434.7bfb7509@mjolnir.drzeus.cx>
Date: Sat, 29 Mar 2008 09:04:34 +0100
From: Pierre Ossman <drzeus@...eus.cx>
To: "Tomas Winkler" <tomasw@...il.com>
Cc: linux-kernel@...r.kernel.org, benzizbit@...il.com
Subject: Re: SDIO: IO-Ready Bit
On Mon, 24 Mar 2008 16:41:25 +0200
"Tomas Winkler" <tomasw@...il.com> wrote:
>
> Anyhow I've seen that concept of parallel probe is not so new. It was
> in 2.6.18 and removed in 2.619.
> Probably different approach, maybe more would be adding a parallel
> initial function.
> Probe will just register to the bus and relevant subsystem, it will be fast.
> Optional Initialization function will run after probe in different
> thread, this would be the place for running FW download.
> Device won't be considered initialized till this function haven't
> completed or is not implemented.
>
I would suspect that would just slow things down in most cases. Most
busses have shared bandwidth, so letting each device take turns is
usually faster than a thundering herd.
Feel free to make your case to the device core maintainers though.
Rgds
Pierre
Download attachment "signature.asc" of type "application/pgp-signature" (198 bytes)
Powered by blists - more mailing lists