From: MORITA Kazutaka <morita.kazutaka@gmail.com>
To: Stefan Hajnoczi <stefanha@gmail.com>
Cc: kwolf@redhat.com, stefanha@redhat.com, qemu-devel@nongnu.org,
MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Subject: Re: [Qemu-devel] [PATCH] sheepdog: add support for connecting to unix domain socket
Date: Sun, 20 Jan 2013 00:13:38 +0900 [thread overview]
Message-ID: <m2obglp5nh.wl%morita.kazutaka@gmail.com> (raw)
In-Reply-To: <20130116142109.GA9300@stefanha-thinkpad.redhat.com>
At Wed, 16 Jan 2013 15:21:09 +0100,
Stefan Hajnoczi wrote:
>
> On Tue, Jan 15, 2013 at 11:12:40PM +0900, MORITA Kazutaka wrote:
> > +static int set_nodelay(int fd)
> > +{
> > + int opt = 1;
> > + return setsockopt(fd, IPPROTO_TCP, TCP_NODELAY, (char *)&opt, sizeof(opt));
> > +}
> > +
>
> Please split this into a separate patch that moves the function to
> util/osdep.c and names it socket_set_nodelay(). You can put it below
> socket_set_cork().
Agreed.
>
> > @@ -804,23 +781,14 @@ static int set_nodelay(int fd)
> > */
> > static int get_sheep_fd(BDRVSheepdogState *s)
> > {
> > - int ret, fd;
> > + int fd;
> >
> > - fd = connect_to_sdog(s->addr, s->port);
> > + fd = connect_to_sdog(s->host_spec);
>
> The patch would be easier to review if you split out a separate patch to
> move from addr/port to host_spec. Then the final patch can focus just
> on UNIX domain sockets without all the addr/port to host_spec changes.
Agreed.
>
> > if (fd < 0) {
> > error_report("%s", strerror(errno));
>
> connect_to_sdog() does not set errno and it already reports errors
> internally. Can this error_report() be dropped?
Yes, I'll remove it.
>
> > return fd;
> > }
> >
> > - socket_set_nonblock(fd);
>
> Where is nonblock being set now that you've removed this?
Oops, I thought that unix_connect and inet_connect return nonblocking
sockets, but it was wrong.
> >
> > -``sheepdog:<host>:<port>:<vdiname>:<tag>''
> > +using Unix Domain Socket:
> > +@example
> > +sheepdog:unix:<domain-socket>:<vdiname>[:<snapid or tag>]
> > +@end example
>
> Please document that <domain-socket> must be an absolute path. This
> will prevent confusing users who try a relative path.
Okay.
Thanks for your comments, I'll send v2.
Kazutaka
prev parent reply other threads:[~2013-01-19 15:13 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-01-15 14:12 [Qemu-devel] [PATCH] sheepdog: add support for connecting to unix domain socket MORITA Kazutaka
2013-01-16 14:21 ` Stefan Hajnoczi
2013-01-19 15:13 ` MORITA Kazutaka [this message]
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=m2obglp5nh.wl%morita.kazutaka@gmail.com \
--to=morita.kazutaka@gmail.com \
--cc=kwolf@redhat.com \
--cc=morita.kazutaka@lab.ntt.co.jp \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@gmail.com \
--cc=stefanha@redhat.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).