[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20190918170807.GA50966@C02WT3WMHTD6>
Date: Wed, 18 Sep 2019 11:08:07 -0600
From: Keith Busch <kbusch@...nel.org>
To: Christoph Hellwig <hch@....de>
Cc: "Baldyga, Robert" <robert.baldyga@...el.com>,
"axboe@...com" <axboe@...com>,
"sagi@...mberg.me" <sagi@...mberg.me>,
"linux-nvme@...ts.infradead.org" <linux-nvme@...ts.infradead.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"Rakowski, Michal" <michal.rakowski@...el.com>
Subject: Re: [PATCH 0/2] nvme: Add kernel API for admin command
On Wed, Sep 18, 2019 at 03:26:11PM +0200, Christoph Hellwig wrote:
> Even if we had a use case for that the bounce buffer is just too ugly
> to live. And I'm really sick and tired of Intel wasting our time for
> their out of tree monster given that they haven't even tried helping
> to improve the in-kernel write caching layers.
Right, I don't have any stake in that out-of-tree caching either. I am
more just interested in trying to get Linux to generically reach as many
NVMe spec features as possible, and extended formats have been in-spec
since the beginning of nvme.
And yes, that bouncing is really nasty, but it's really only needed for
PRP, so maybe let's just not ignore that transfer mode and support
extended metadata iff the controller supports SGLs. We just need a
special SGL setup routine to weave the data and metadata.
Powered by blists - more mailing lists