[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <3072d5a0-43c7-3396-c57f-6af83621b71c@nvidia.com>
Date: Mon, 24 Aug 2020 19:07:02 -0700
From: John Hubbard <jhubbard@...dia.com>
To: Al Viro <viro@...iv.linux.org.uk>
CC: Andrew Morton <akpm@...ux-foundation.org>,
Christoph Hellwig <hch@...radead.org>,
Ilya Dryomov <idryomov@...il.com>,
Jens Axboe <axboe@...nel.dk>, Jeff Layton <jlayton@...nel.org>,
<linux-xfs@...r.kernel.org>, <linux-fsdevel@...r.kernel.org>,
<linux-block@...r.kernel.org>, <ceph-devel@...r.kernel.org>,
<linux-mm@...ck.org>, LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 0/5] bio: Direct IO: convert to pin_user_pages_fast()
On 8/24/20 6:54 PM, Al Viro wrote:
> On Fri, Aug 21, 2020 at 09:20:54PM -0700, John Hubbard wrote:
>
>> Direct IO behavior:
>>
>> ITER_IOVEC:
>> pin_user_pages_fast();
>> break;
>>
>> ITER_KVEC: // already elevated page refcount, leave alone
>> ITER_BVEC: // already elevated page refcount, leave alone
>> ITER_PIPE: // just, no :)
>
> Why? What's wrong with splice to O_DIRECT file?
>
Oh! I'll take a look. Is this the fs/splice.c stuff? I ruled this out early
mainly based on Christoph's comment in [1] ("ITER_PIPE is rejected іn the
direct I/O path"), but if it's supportable then I'll hook it up.
(As you can see, I'm still very much coming up to speed on the various things
that invoke iov_iter_get_pages*().)
[1] https://lore.kernel.org/kvm/20190724061750.GA19397@infradead.org/
thanks,
--
John Hubbard
NVIDIA
Powered by blists - more mailing lists