Hi,
On Wed, Sep 18, 2019 at 01:37:38PM +0300, Samuli Seppänen wrote:
> We are considering removing the --disable-server option from OpenVPN in 2.5.
>
> Do you use (and need) it, or know of somebody using (and needing) it?
Just rediscovered this thread... nobody??
gert
--
"If was one thing al
So a recent question on the openvpn-users caused me to have found
that, this "reserve-one-more-address" is seemingly related to Windows
/ --ip-win32. Commit 251cc8f made a "correction" to the
manual/documentation which sort of implied that the address is
reserved as the default address chosen by `-
If it isn't a typo, I wonder if it is the equivalence of
`pool_end_reserve`. As with the subnet topology, each client takes up
1 address in contrast to 4 with net30.
So perhaps the question is, what's the purpose of pool_end_reserve?
On Wed, 13 Nov 2019 at 18:07, Tom Yan wrote:
>
> It appears to
Hi,
> What I do not really like is the inflation of the code with
>
> if (!tt->wintun)
>
> statements now. I think this should be refactored out into an
> "open_tun_wintun()" and an "open_tun_tap_windows()" function, where
> all the bits that are now inside an "if (!tt->wintun)" get thei
From: Lev Stipakov
This makes Windows's tun_open() method easier to read
by factoring out blocks of code, which perform certain task,
into separate functions. This also minimizes inflation of
if (!tt->wintun) { }
blocks.
While patch looks big and scary, there are no functional changes
at al
As /31 subnet now works (as we stop setting broadcast address), the server
directives can be fixed for it as well.
---
src/openvpn/helper.c | 19 ---
1 file changed, 12 insertions(+), 7 deletions(-)
diff --git a/src/openvpn/helper.c b/src/openvpn/helper.c
index 470b9120..5b2ca0cc
It appears to be a copy-and-paste kind of typo (pool start is network address +
2).
---
src/openvpn/helper.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/openvpn/helper.c b/src/openvpn/helper.c
index 5b2ca0cc..6e2f0891 100644
--- a/src/openvpn/helper.c
+++ b/src/openvpn
It also avoids limitation (i.e. netbits <= 29) that does not apply to the
subnet topology.
---
src/openvpn/helper.c | 74
1 file changed, 27 insertions(+), 47 deletions(-)
diff --git a/src/openvpn/helper.c b/src/openvpn/helper.c
index ff9df506..470b91
The setting probably just slipped out of the if-block over rewrites.
---
src/openvpn/helper.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/openvpn/helper.c b/src/openvpn/helper.c
index 6e2f0891..a36bc63a 100644
--- a/src/openvpn/helper.c
+++ b/src/openvpn/helper.c
@@ -35
On Wed, 13 Nov 2019 at 17:18, Gert Doering wrote:
>
> Hi,
>
> It might be related to the way the windows tap6 driver always needs a
> gateway address even in tun mode (and uses the last address from the
> subnet for this). There's very few typos in OpenVPN code regarding
> *this* type of math, an
Hi,
On 13/11/2019 10:18, Gert Doering wrote:
>> By the way, why does `o->ifconfig_pool_netmask` need to be set even
>> when `nopool` is set?
>
> Not sure. This code has been rewritten a number of times over the
> years, so maybe the assignment slipped outside of an if() block...
>
> Things like
Hi,
On Wed, Nov 13, 2019 at 05:13:09PM +0800, Tom Yan wrote:
> For the record, as I don't see why `o->ifconfig_pool_end` was
> `(o->server_network | ~o->server_netmask) - 2` for tun + subnet while
> it was `(o->server_network | ~o->server_netmask) - 1` for tap, I
> assume the former is a typo.
It
For the record, as I don't see why `o->ifconfig_pool_end` was
`(o->server_network | ~o->server_netmask) - 2` for tun + subnet while
it was `(o->server_network | ~o->server_netmask) - 1` for tap, I
assume the former is a typo.
By the way, why does `o->ifconfig_pool_netmask` need to be set even
when
As /31 subnet now works (as we stop setting broadcast address), the server
directives can be fixed for it as well. Also stop repeating code for tap and
tun + subnet.
---
src/openvpn/helper.c | 90 ++--
1 file changed, 37 insertions(+), 53 deletions(-)
dif
14 matches
Mail list logo
