[PATCHSET,v4,0/8] Turn single segment imports into ITER_UBUF

Message ID	20230328173613.555192-1-axboe@kernel.dk (mailing list archive)
Headers	show Return-Path: <linux-fsdevel-owner@vger.kernel.org> From: Jens Axboe <axboe@kernel.dk> To: linux-fsdevel@vger.kernel.org Cc: torvalds@linux-foundation.org, brauner@kernel.org, viro@zeniv.linux.org.uk Subject: [PATCHSET v4 0/8] Turn single segment imports into ITER_UBUF Date: Tue, 28 Mar 2023 11:36:05 -0600 Message-Id: <20230328173613.555192-1-axboe@kernel.dk> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Precedence: bulk
Series	Turn single segment imports into ITER_UBUF \| expand [PATCHSET,v4,0/8] Turn single segment imports into ITER_UBUF [1/8] iov_iter: teach iov_iter_iovec() to deal with ITER_UBUF [2/8] iov_iter: add iovec_nr_user_vecs() helper [3/8] snd: move mapping an iov_iter to user bufs into a helper [4/8] snd: make snd_map_bufs() deal with ITER_UBUF [5/8] IB/hfi1: make hfi1_write_iter() deal with ITER_UBUF iov_iter [6/8] IB/qib: make qib_write_iter() deal with ITER_UBUF iov_iter [7/8] iov_iter: convert import_single_range() to ITER_UBUF [8/8] iov_iter: import single vector iovecs as ITER_UBUF

Message ID

20230328173613.555192-1-axboe@kernel.dk (mailing list archive)

Headers

From: Jens Axboe <axboe@kernel.dk>
To: linux-fsdevel@vger.kernel.org
Cc: torvalds@linux-foundation.org, brauner@kernel.org,
        viro@zeniv.linux.org.uk
Subject: [PATCHSET v4 0/8] Turn single segment imports into ITER_UBUF
Date: Tue, 28 Mar 2023 11:36:05 -0600
Message-Id: <20230328173613.555192-1-axboe@kernel.dk>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

Series

Turn single segment imports into ITER_UBUF | expand

Message

Jens Axboe March 28, 2023, 5:36 p.m. UTC

Hi,

ather than repeat the same blurb again, see the v2 posting here:

https://lore.kernel.org/linux-fsdevel/20230327180449.87382-1-axboe@kernel.dk/

tldr - turn single segment iovecs into ITER_UBUF rather than ITER_IOVEC,
because they are more efficient.

Main addition since v3 is being careful checking users of iov_iter,
and there are two odd ones - infiniband and sound. I've added a
prep patch to be able to grab the segment count of a user backed
iterator, which sound and IB can then use.

I'm not convinced the sound patch is useful at all, since it LOOKS
like it requires nr_segments > 1 anyway. I'll chat with the folks
on that side to be sure. We may get away with just EINVAL for
this case.

IB is pretty straight forward, just make it deal with ITER_UBUF
as a single segment thing.