[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <yq1h85986fu.fsf@oracle.com>
Date: Wed, 18 Sep 2019 20:40:05 -0400
From: "Martin K. Petersen" <martin.petersen@...cle.com>
To: Christoph Hellwig <hch@....de>
Cc: Keith Busch <kbusch@...nel.org>,
"sagi\@grimberg.me" <sagi@...mberg.me>,
"linux-kernel\@vger.kernel.org" <linux-kernel@...r.kernel.org>,
"linux-nvme\@lists.infradead.org" <linux-nvme@...ts.infradead.org>,
"Rakowski\, Michal" <michal.rakowski@...el.com>,
"axboe\@fb.com" <axboe@...com>,
"Baldyga\, Robert" <robert.baldyga@...el.com>
Subject: Re: [PATCH 0/2] nvme: Add kernel API for admin command
Christoph,
> On Wed, Sep 18, 2019 at 11:08:07AM -0600, Keith Busch wrote:
>> 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.
>
> Well, what is the point? If people really want to use metadata they
> should just buy a drive supporting the separate metadata pointer. In
> fact I haven't had to deal with a drive that only supports interleaved
> metadata so far given how awkward that is to deal with.
Yep. There's a reason we did DIX...
--
Martin K. Petersen Oracle Linux Engineering
Powered by blists - more mailing lists