qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Jan Kiszka <jan.kiszka@web.de>
To: Alexander Graf <agraf@suse.de>
Cc: qemu-devel@nongnu.org
Subject: [Qemu-devel] Re: [PATCH] User Networking: Enable removal of redirections
Date: Thu, 28 May 2009 23:31:09 +0200	[thread overview]
Message-ID: <4A1F029D.7020901@web.de> (raw)
In-Reply-To: <1243335807-9495-1-git-send-email-agraf@suse.de>

[-- Attachment #1: Type: text/plain, Size: 5649 bytes --]

Alexander Graf wrote:
> Using the new host_net_redir command you can easily create redirections
> on the fly while your VM is running.
> 
> While that's great, it's missing the removal of redirections, in case you
> want to have a port closed again at a later point in time.
> 
> This patch adds support for removal of redirections.
> 
> Signed-off-by: Alexander Graf <agraf@suse.de>
> ---
>  monitor.c        |    5 +++--
>  net.c            |   43 ++++++++++++++++++++++++++++++++++++++++++-
>  net.h            |    2 +-
>  slirp/libslirp.h |    1 +
>  slirp/slirp.c    |   23 +++++++++++++++++++++++
>  vl.c             |    2 +-
>  6 files changed, 71 insertions(+), 5 deletions(-)
> 
> diff --git a/monitor.c b/monitor.c
> index 0f38c71..dbab3de 100644
> --- a/monitor.c
> +++ b/monitor.c
> @@ -1759,8 +1759,9 @@ static const mon_cmd_t mon_cmds[] = {
>      { "host_net_remove", "is", net_host_device_remove,
>        "vlan_id name", "remove host VLAN client" },
>  #ifdef CONFIG_SLIRP
> -    { "host_net_redir", "s", net_slirp_redir,
> -      "[tcp|udp]:host-port:[guest-host]:guest-port", "redirect TCP or UDP connections from host to guest (requires -net user)" },
> +    { "host_net_redir", "ss?", net_slirp_redir,
> +      "[tcp|udp]:host-port:[guest-host]:guest-port", "redirect TCP or UDP connections from host to guest (requires -net user)\n"
> +      "host_net_redir remove [tcp:|udp:]host-port -- remove redirection" },
>  #endif
>      { "balloon", "i", do_balloon,
>        "target", "request VM to change it's memory allocation (in MB)" },
> diff --git a/net.c b/net.c
> index 31ee95a..de019b3 100644
> --- a/net.c
> +++ b/net.c
> @@ -568,7 +568,43 @@ static int net_slirp_init(VLANState *vlan, const char *model, const char *name)
>      return 0;
>  }
>  
> -void net_slirp_redir(Monitor *mon, const char *redir_str)
> +static void net_slirp_redir_rm(Monitor *mon, const char *port_str)
> +{
> +    int host_port;
> +    char buf[256] = "";
> +    const char *p = port_str;
> +    int is_udp = 0;
> +    int n;
> +
> +    if (!mon)
> +        return;
> +
> +    if (!port_str || !port_str[0])
> +        goto fail_syntax;
> +
> +    get_str_sep(buf, sizeof(buf), &p, ':');
> +
> +    if (!strcmp(buf, "tcp") || buf[0] == '\0') {
> +        is_udp = 0;
> +    } else if (!strcmp(buf, "udp")) {
> +        is_udp = 1;
> +    } else {
> +        goto fail_syntax;
> +    }
> +
> +    host_port = atoi(p);
> +
> +    n = slirp_redir_rm(is_udp, host_port);
> +
> +    monitor_printf(mon, "removed %d redirections to %s port %d\n", n,
> +                        is_udp ? "udp" : "tcp", host_port);
> +    return;
> +
> + fail_syntax:
> +    monitor_printf(mon, "invalid format\n");
> +}
> +
> +void net_slirp_redir(Monitor *mon, const char *redir_str, const char *redir_opt2)
>  {
>      int is_udp;
>      char buf[256], *r;
> @@ -581,6 +617,11 @@ void net_slirp_redir(Monitor *mon, const char *redir_str)
>          slirp_init(slirp_restrict, slirp_ip);
>      }
>  
> +    if (!strcmp(redir_str, "remove")) {
> +        net_slirp_redir_rm(mon, redir_opt2);
> +        return;
> +    }
> +
>      p = redir_str;
>      if (get_str_sep(buf, sizeof(buf), &p, ':') < 0)
>          goto fail_syntax;
> diff --git a/net.h b/net.h
> index 41a3082..feee021 100644
> --- a/net.h
> +++ b/net.h
> @@ -112,7 +112,7 @@ int net_client_init(const char *device, const char *p);
>  void net_client_uninit(NICInfo *nd);
>  int net_client_parse(const char *str);
>  void net_slirp_smb(const char *exported_dir);
> -void net_slirp_redir(Monitor *mon, const char *redir_str);
> +void net_slirp_redir(Monitor *mon, const char *redir_str, const char *redir_opt2);
>  void net_cleanup(void);
>  int slirp_is_inited(void);
>  void net_client_check(void);
> diff --git a/slirp/libslirp.h b/slirp/libslirp.h
> index a1cd70e..6fc2c32 100644
> --- a/slirp/libslirp.h
> +++ b/slirp/libslirp.h
> @@ -18,6 +18,7 @@ void slirp_input(const uint8_t *pkt, int pkt_len);
>  int slirp_can_output(void);
>  void slirp_output(const uint8_t *pkt, int pkt_len);
>  
> +int slirp_redir_rm(int is_udp, int host_port);
>  int slirp_redir(int is_udp, int host_port,
>                  struct in_addr guest_addr, int guest_port);
>  int slirp_add_exec(int do_pty, const void *args, int addr_low_byte,
> diff --git a/slirp/slirp.c b/slirp/slirp.c
> index 04d3ded..33397c0 100644
> --- a/slirp/slirp.c
> +++ b/slirp/slirp.c
> @@ -734,6 +734,29 @@ void if_encap(const uint8_t *ip_data, int ip_data_len)
>      }
>  }
>  
> +/* Unlistens a redirection
> + *
> + * Return value: number of redirs removed */
> +int slirp_redir_rm(int is_udp, int host_port)
> +{
> +    struct socket *so;
> +    struct socket *head = (is_udp ? &udb : &tcb);
> +    int fport = htons(host_port);
> +    int n = 0;
> +
> + loop_again:
> +    for (so = head->so_next; so != head; so = so->so_next) {
> +        if (so->so_fport == fport) {
> +            close(so->s);
> +            sofree(so);
> +            n++;
> +            goto loop_again;
> +        }
> +    }

Unfortunately, this does not only target host->guest redirection sockets
but also sockets slirp uses for NAT'ing guest originated connections.
The same applies to your "host_net_redir list". So giving this in user
hand, unwanted damaged can be caused to guest network sessions. What we
need is a tag in struct socket to identify static redirection sockets.

What's your current plan regarding these two commits? We should
coordinate as my work touches the same area.

Jan


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 257 bytes --]

  parent reply	other threads:[~2009-05-28 21:31 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-05-26 11:03 [Qemu-devel] [PATCH] User Networking: Enable removal of redirections Alexander Graf
2009-05-26 11:03 ` [Qemu-devel] [PATCH] User networking: Show active connections Alexander Graf
2009-05-27  6:58   ` [Qemu-devel] " Jan Kiszka
2009-05-27  6:58 ` [Qemu-devel] Re: [PATCH] User Networking: Enable removal of redirections Jan Kiszka
2009-05-27  7:21   ` Jan Kiszka
2009-05-27  7:26     ` Alexander Graf
2009-05-27  8:04       ` Jan Kiszka
2009-05-28 21:31 ` Jan Kiszka [this message]
2009-05-28 22:17   ` Alexander Graf
2009-05-29  7:48     ` Jan Kiszka

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=4A1F029D.7020901@web.de \
    --to=jan.kiszka@web.de \
    --cc=agraf@suse.de \
    --cc=qemu-devel@nongnu.org \
    /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).