[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <200803051102.03478.jbarnes@virtuousgeek.org>
Date: Wed, 5 Mar 2008 11:02:02 -0800
From: Jesse Barnes <jbarnes@...tuousgeek.org>
To: akepner@....com
Cc: James Bottomley <James.Bottomley@...senpartnership.com>,
Grant Grundler <grundler@...isc-linux.org>,
Michael Ellerman <michael@...erman.id.au>,
Tony Luck <tony.luck@...el.com>, Jes Sorensen <jes@....com>,
Randy Dunlap <randy.dunlap@...cle.com>,
Roland Dreier <rdreier@...co.com>,
David Miller <davem@...emloft.net>,
Benjamin Herrenschmidt <benh@...nel.crashing.org>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/3 v3] dma: document dma_{un}map_{single|sg}_attrs() interface
On Wednesday, March 05, 2008 10:13 am akepner@....com wrote:
> On Fri, Feb 29, 2008 at 12:37:56PM -0600, James Bottomley wrote:
> > ....
> > To be honest, I still don't like the name. SYNC_ON_WRITE is the SN2
> > implementation. What it's actually doing is implementing strict
> > ordering semantics. I think it should really be
> > DMA_ATTR_STRICT_ORDERING (with a corresponding
> > DMA_ATTR_RELAXED_ORDERING)....
>
> I've been thinking about a new name, but don't like
> DMA_ATTR_STRICT_ORDERING.
>
> What I'm trying to do is to establish order (across
> a NUMA fabric) of DMA to different memory regions, i.e.,
> DMA to memory region A forces all outstanding DMA (to
> memory regions B, C,....) to complete first.
>
> DMA_ATTR_STRICT_ORDERING sounds like a PCI thing to me,
> and this is a NUMA interconnect thing.
Well, we used to call it a DMA barrier back in the old days, so
DMA_ATTR_BARRIER might work (or DMA_ATTR_INTERCONNECT_BARRIER if you want to
be extra clear about what it's doing).
Jesse
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists