Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 9 Oct 2013 01:57:26 +0300
From:      Alexander Motin <mav@FreeBSD.org>
To:        Ilya Bakulin <ilya@bakulin.de>
Cc:        Adrian Chadd <adrian.chadd@gmail.com>, Alexander Motin <mav@freebsd.org>, freebsd-embedded@freebsd.org
Subject:   Re: SDIO for FreeBSD
Message-ID:  <CAO4K=PUp3tNg1vTTEjKj7fZ0pLeadRV6%2Bf=FWO8ko5xybZDcoA@mail.gmail.com>
In-Reply-To: <5254892B.3050507@bakulin.de>
References:  <5251A9D3.1080803@bakulin.de> <5251C911.6020003@FreeBSD.org> <CAJ-Vmon0xr597We=sF5Uhja2qb=viuiqCbQXZgUvp7VX-FsPfA@mail.gmail.com> <4FB11076-DD88-43F2-A449-E41D2088A4DD@bsdimp.com> <5254892B.3050507@bakulin.de>

next in thread | previous in thread | raw e-mail | index | archive | help
09.10.2013 1:37 =D0=CF=CC=D8=DA=CF=D7=C1=D4=C5=CC=D8 "Ilya Bakulin" <ilya@b=
akulin.de> =CE=C1=D0=C9=D3=C1=CC:
>
> On 07.10.13 01:19, Warner Losh wrote:
> > I've often thought that having an SDIO protocol for CAM would be where
we'd wind up because once you start adding in GPIO pin interrupt
signalling, message "interrupt" signaling, message completion, etc, it
looks a lot like many of the mechanisms we have in CAM...  I'm not at all
sure that USB is the right model to use, frankly...  But it does have some
superficial similarities...
> >
> > Warner
>
> The idea seems to be interesting -- at least we can reuse the CAM queues
> and interrupt handling algorithms...
> One thing that I'd like to ask after initial reading about CAM in FreeBSD=
:
> is it possible for SIM devices to initiate data transfer?
> the CAM subsystem is primarily used for storage devices,
> where the host issues a SCSI command, schedules it to the SIM,
> and then SIM receives an interrupt from the h/w and updates the existing
> CCB,
> signaling command completion.
>
> In SDIO case we can have the interrupts from the card
> that are not associated with any previous command.
> Is it possible to handle such situation using CAM framework?

CAM has mechanism of asynchronous notifications, that is quite alike to
interupts.

On the other side, target mode drivers in CAM are using different concept
with requests that just wait inside the controller or its driver to be
completed on next incoming request, carying requst data inside.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAO4K=PUp3tNg1vTTEjKj7fZ0pLeadRV6%2Bf=FWO8ko5xybZDcoA>