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: <20230629155433.4170837-1-dhowells@redhat.com> <20230629155433.4170837-2-dhowells@redhat.com>
 <CAJfpegsJuvXJDcXpo9T19Gw0tDuvyOJdv44Y2bt04MEf1JLxGg@mail.gmail.com>
 <c634a18e-9f2b-4746-bd8f-aa1d41e6ddf7@mattwhitlock.name> <CAJfpegvq4M_Go7fHiWVBBkrK6h4ChLqQTd0+EOKbRWZDcVerWA@mail.gmail.com>
 <ZLg9HbhOVnLk1ogA@casper.infradead.org> <CAHk-=wiq95bWiWLyz96ombPfpy=PNrc2KKyzJ2d+WMrxi6=OVA@mail.gmail.com>
 <63041.1690191864@warthog.procyon.org.uk>
In-Reply-To: <63041.1690191864@warthog.procyon.org.uk>
From:   Miklos Szeredi <miklos@szeredi.hu>
Date:   Mon, 24 Jul 2023 15:55:21 +0200
Message-ID: <CAJfpegstr2CwC2ZL4-y_bAjS3hqF_vta5e4XQneJYmxz9rhVpA@mail.gmail.com>
Subject: Re: [RFC PATCH 1/4] splice: Fix corruption of spliced data after
 splice() returns
To:     David Howells <dhowells@redhat.com>
Cc:     Linus Torvalds <torvalds@linux-foundation.org>,
        Matthew Wilcox <willy@infradead.org>,
        Matt Whitlock <kernel@mattwhitlock.name>,
        netdev@vger.kernel.org, Dave Chinner <david@fromorbit.com>,
        Jens Axboe <axboe@kernel.dk>, linux-fsdevel@kvack.org,
        linux-mm@kvack.org, linux-kernel@vger.kernel.org,
        Christoph Hellwig <hch@lst.de>, linux-fsdevel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, 24 Jul 2023 at 11:44, David Howells <dhowells@redhat.com> wrote:
>
> Linus Torvalds <torvalds@linux-foundation.org> wrote:
>
> > > So what's the API that provides the semantics of _copying_?
> >
> > It's called "read()" and "write()".
>
> What about copy_file_range()?  That seems to fall back to splicing if not
> directly implemented by the filesystem.  It looks like the manpage for that
> needs updating too - or should that actually copy?

Both source and destination of copy_file_range() are regular files and
do_splice_direct() is basically equivalent to write(dest, mmap of
source), no refd buffers remain beyond the end of the syscall.  What
is it that should be updated in the manpage?

Thanks,
Miklos