From: Eric Dumazet <dada1@cosmosbay.com>
To: David Miller <davem@davemloft.net>
Cc: akpm@linux-foundation.org, Volker.Lendecke@SerNet.DE,
linux-kernel@vger.kernel.org, sfrench@us.ibm.com,
jens.axboe@oracle.com, netdev@vger.kernel.org
Subject: Re: maximum buffer size for splice(2) tcp->pipe?
Date: Thu, 15 Jan 2009 12:47:22 +0100 [thread overview]
Message-ID: <496F224A.9080100@cosmosbay.com> (raw)
In-Reply-To: <20090114.205839.20689555.davem@davemloft.net>
David Miller a écrit :
> From: Eric Dumazet <dada1@cosmosbay.com>
> Date: Wed, 14 Jan 2009 00:38:32 +0100
>
>> [PATCH] net: splice() from tcp to socket should take into account O_NONBLOCK
>>
>> Instead of using SPLICE_F_NONBLOCK to select a non blocking mode both on
>> source tcp socket and pipe destination, we use the underlying file flag (O_NONBLOCK)
>> for selecting a non blocking socket.
>>
>> Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
>
> This needs at least some more thought.
>
> It seems, for one thing, that this change will interfere with the
> intentions of the code in splice_dirt_to_actor which goes:
>
> /*
> * Don't block on output, we have to drain the direct pipe.
> */
> sd->flags &= ~SPLICE_F_NONBLOCK;
Reading splice_direct_to_actor() I see nothing wrong with the patch
(Patch is about splice from socket to pipe, while the sd->flags you mention
in splice_direct_to_actor() only applies to the splice from internal pipe to
something else, as splice_direct_to_actor() allocates an internal pipe to perform
its work.
Also, the meaning of SPLICE_F_NONBLOCK, as explained in include/linux/splice.h is :
#define SPLICE_F_NONBLOCK (0x02) /* don't block on the pipe splicing (but */
/* we may still block on the fd we splice */
/* from/to, of course */
If the comment is still correct, SPLICE_F_NONBLOCK only applies to the pipe implied in
splice() syscall.
For the other file, either its :
- A regular file : nonblocking mode is not available, like a normal read()/write() syscall
- A socket : We should be able to specify if its blocking or not, independantly from
the SPLICE_F_NONBLOCK flag that only applies to the pipe. Normal way
is using ioctl(FIONBIO) or other fcntl() call to change file->f_flags O_NONBLOCK
In order to be able to efficiently use splice() from a socket to a file, we need
to do a loop of :
{
splice(from blocking tcp socket to non blocking pipe, SPLICE_F_NONBLOCK); /* nonblocking pipe or risk deadlock */
splice(from pipe to file)
}
next prev parent reply other threads:[~2009-01-15 11:47 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <E1LKrp3-004Tub-IR@intern.SerNet.DE>
2009-01-13 20:37 ` maximum buffer size for splice(2) tcp->pipe? Andrew Morton
2009-01-13 23:15 ` Eric Dumazet
2009-01-13 23:38 ` Eric Dumazet
2009-01-15 4:58 ` David Miller
2009-01-15 11:47 ` Eric Dumazet [this message]
2009-01-14 7:40 ` Volker Lendecke
2009-01-14 9:13 ` Eric Dumazet
2009-01-14 10:03 ` Volker Lendecke
2009-01-14 10:17 ` Eric Dumazet
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=496F224A.9080100@cosmosbay.com \
--to=dada1@cosmosbay.com \
--cc=Volker.Lendecke@SerNet.DE \
--cc=akpm@linux-foundation.org \
--cc=davem@davemloft.net \
--cc=jens.axboe@oracle.com \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=sfrench@us.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).