Development discussion of WireGuard
 help / color / mirror / Atom feed
From: "Jason A. Donenfeld" <Jason@zx2c4.com>
To: Philipp Czerner <suyjuris.gi@nicze.de>
Cc: WireGuard mailing list <wireguard@lists.zx2c4.com>,
	Simon Rozman <simon@rozman.si>
Subject: Re: [PATCH wireguard-windows] Calculate the actual route metric by summing interface and route metric.
Date: Sun, 29 Mar 2020 23:31:49 -0600	[thread overview]
Message-ID: <CAHmME9r4vDKeWCjp2XF3pKaKUDokDphnVoYBqV5YfLd=gTNbPA@mail.gmail.com> (raw)
In-Reply-To: <20200327170750.5285-1-suyjuris.gi@nicze.de>

On Sun, Mar 29, 2020 at 8:29 PM Philipp Czerner <suyjuris.gi@nicze.de> wrote:
>
> Signed-off-by: Philipp Czerner <suyjuris.gi@nicze.de>
> ---
> Hi,
>
> I had some issues setting up Wireguard behind another VPN. Curiously, it bound the physical interface instead of the other VPN, which was the default route. According to MSDN "the actual route metric used to compute the route preference is the summation of interface metric specified in the Metric member of the MIB_IPINTERFACE_ROW structure and the route metric offset specified in this member" (documentation for MIB_IPFORWARD_ROW2), but the code did not seem to consider this. After I changed the calculation, I got the expected behaviour.
>
> Cheers,
> Philipp
>
>  tunnel/defaultroutemonitor.go | 10 ++++++++--
>  1 file changed, 8 insertions(+), 2 deletions(-)
>
> diff --git a/tunnel/defaultroutemonitor.go b/tunnel/defaultroutemonitor.go
> index c102b64..4b24e8e 100644
> --- a/tunnel/defaultroutemonitor.go
> +++ b/tunnel/defaultroutemonitor.go
> @@ -33,8 +33,14 @@ func bindSocketRoute(family winipcfg.AddressFamily, device *device.Device, ourLU
>                 if err != nil || ifrow.OperStatus != winipcfg.IfOperStatusUp {
>                         continue
>                 }
> -               if r[i].Metric < lowestMetric {
> -                       lowestMetric = r[i].Metric
> +
> +               iface, err := r[i].InterfaceLUID.IPInterface(family)
> +               if err != nil {
> +                       continue
> +               }
> +
> +               if r[i].Metric + iface.Metric < lowestMetric {
> +                       lowestMetric = r[i].Metric + iface.Metric;

Nice semicolon ;-).

Thanks a lot for this patch and triaging the issue. Nice to hear that
nested VPNs work. If you don't mind sharing, I'd be interested to
learn what software sets up its routes in this manner.

Anyway, applied here:
https://git.zx2c4.com/wireguard-windows/commit/?id=66ad537f4e2144112cb6686c4fc4856d70bf6a3d

Jason

  reply	other threads:[~2020-03-30  5:32 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-03-27 17:07 Philipp Czerner
2020-03-30  5:31 ` Jason A. Donenfeld [this message]
2020-03-30  8:40   ` Ludwig Herzog
2020-04-01  6:46     ` Simon Rozman
2020-03-31 12:56   ` Philipp Czerner

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='CAHmME9r4vDKeWCjp2XF3pKaKUDokDphnVoYBqV5YfLd=gTNbPA@mail.gmail.com' \
    --to=jason@zx2c4.com \
    --cc=simon@rozman.si \
    --cc=suyjuris.gi@nicze.de \
    --cc=wireguard@lists.zx2c4.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).