Quoting Daniel P. Berrange (berrange(a)redhat.com):
On Thu, Oct 31, 2013 at 04:36:24PM -0500, Serge Hallyn wrote:
> This will properly lock libvirt's usage of iptables with
> others (like ufw).
>
> (See
https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1245322)
>
> Signed-off-by: Serge Hallyn <serge.hallyn(a)ubuntu.com>
> ---
> src/util/viriptables.c | 30 ++++++++++++++++++++++++++----
> 1 file changed, 26 insertions(+), 4 deletions(-)
>
> diff --git a/src/util/viriptables.c b/src/util/viriptables.c
> index 16f571e..30d59b6 100644
> --- a/src/util/viriptables.c
> +++ b/src/util/viriptables.c
> @@ -50,19 +50,25 @@
> #include "virstring.h"
> #include "virutil.h"
>
> +bool iptables_supports_xlock = false;
> +
> #if HAVE_FIREWALLD
> static char *firewall_cmd_path = NULL;
> +#endif
>
> static int
> virIpTablesOnceInit(void)
> {
> + virCommandPtr cmd;
> + int status;
> +
> +#if HAVE_FIREWALLD
> firewall_cmd_path = virFindFileInPath("firewall-cmd");
> if (!firewall_cmd_path) {
> VIR_INFO("firewall-cmd not found on system. "
> "firewalld support disabled for iptables.");
> } else {
> - virCommandPtr cmd = virCommandNew(firewall_cmd_path);
> - int status;
> + virCommandNew(firewall_cmd_path);
>
> virCommandAddArgList(cmd, "--state", NULL);
> if (virCommandRun(cmd, &status) < 0 || status != 0) {
> @@ -74,13 +80,26 @@ virIpTablesOnceInit(void)
> }
> virCommandFree(cmd);
> }
> +
> + if (firewall_cmd_path)
> + return 0;
> +
> +#endif
> +
> + cmd = virCommandNew(IPTABLES_PATH);
> + virCommandAddArgList(cmd, "-w", "-L", "-n",
NULL);
What version of iptables actually has this '-w' flag ? The ubuntu bug
It was introduced in 1.4.20, specifically with
commit 93587a04d0f2511e108bbc4d87a8b9d28a5c5dd8
Author: Phil Oester <kernel(a)linuxace.com>
Date: Fri May 31 09:07:04 2013 -0400
ip[6]tables: Add locking to prevent concurrent instances
points to an upstream patch, which does locking unconditionally, no
mention of any '-w' flag.
Well yes it locks unconditionally, but without -w it won't wait. So
if there is another app using iptables at the time, then the call fails
and libvirt fails to create the interface.
Looking on a per-rule basis is pretty lame locking, since changes
to iptables are often comprised of many rule changes :-(
> + if (virCommandRun(cmd, &status) < 0 || status != 0) {
> + VIR_INFO("xtables locking not supported by your iptables");
> + } else {
> + VIR_INFO("using xtables locking for iptables");
> + iptables_supports_xlock = true;
> + }
> + virCommandFree(cmd);
> return 0;
> }
>
> VIR_ONCE_GLOBAL_INIT(virIpTables)
>
> -#endif
> -
The call to virIptablesInitialize is protected by
#if HAVE_FIREWALLD too, so this code will not run
on older distros.
I thought I changed that, but I may not properly understand the
libvirt init code. I moved the VIR_ONCE_GLOBAL_INIT(virIptTables)
out from under that ifdef.
-serge