Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <3710261.1691764329@warthog.procyon.org.uk>
In-Reply-To: <3710261.1691764329@warthog.procyon.org.uk>
From:   Linus Torvalds <torvalds@linux-foundation.org>
Date:   Fri, 11 Aug 2023 09:38:46 -0700
Message-ID: <CAHk-=wi1QZ+zdXkjnEY7u1GsVDaBv8yY+m4-9G3R34ihwg9pmQ@mail.gmail.com>
Subject: Re: [RFC PATCH] iov_iter: Convert iterate*() to inline funcs
To:     David Howells <dhowells@redhat.com>
Cc:     Alexander Viro <viro@zeniv.linux.org.uk>,
        Jens Axboe <axboe@kernel.dk>, Christoph Hellwig <hch@lst.de>,
        Christian Brauner <christian@brauner.io>,
        Matthew Wilcox <willy@infradead.org>, jlayton@kernel.org,
        linux-block@vger.kernel.org, linux-fsdevel@vger.kernel.org,
        linux-mm@kvack.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Fri, 11 Aug 2023 at 07:40, David Howells <dhowells@redhat.com> wrote:
>
> Convert the iov_iter iteration macros to inline functions to make the code
> easier to follow.

I like this generally, the code generation deprovement worries me a
bit, but from a quick look on a test-branch it didn't really look all
that bad (but the changes are too big to usefully show up as asm
diffs)

I do note that maybe you should just also mark
copy_to/from/page_user_iter as being always-inlines. clang actually
seems to do that without prompting, gcc apparently not.

Or at *least* do the memcpy_to/from_iter functions, which are only
wrappers around memcpy and are just completely noise. I'm surprised
gcc didn't already inline that. Strange.

            Linus