From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: jens@viisauksena.de Received: from viisauksena.de (v32412.1blu.de [178.254.39.111]) by krantz.zx2c4.com (ZX2C4 Mail Server) with ESMTP id e387e7f8 for ; Wed, 10 Aug 2016 01:36:13 +0000 (UTC) To: "Jason A. Donenfeld" References: <7a999008-9396-c82d-1764-d658873af7fc@viisauksena.de>

<8eebb32d-b302-5975-726b-874a9ad12a84@viisauksena.de> <20160803064853.Horde.ZExiUp0by8SskYUmZW7CFq4@www.vdorst.com> <20160803201014.Horde.RctXDT7A7Fi6hFbWMcqDmJs@www.vdorst.com> <923cdc92-aa01-758a-5741-99a3f6009dc5@viisauksena.de>

From: jens Message-ID: <90a667ed-4f48-a0cc-1e93-472fcc5ef67f@viisauksena.de> Date: Wed, 10 Aug 2016 03:37:00 +0200 MIME-Version: 1.0 In-Reply-To: Content-Type: multipart/alternative; boundary="------------F04F4078619B3ED3C105AD7F" Cc: WireGuard mailing list Subject: Re: [WireGuard] LEDE / OpenWrt test - on TP-Link841Nv11 - 15MBs and crash List-Id: Development discussion of WireGuard List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , This is a multi-part message in MIME format. --------------F04F4078619B3ED3C105AD7F Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable On 04.08.2016 13:34, jens wrote: > [ X ] I have tried removing "sk_clear_memalloc(sock);" and > "sk_set_memalloc(sock->sk);" from socket.c, and the problem still > continued. > [ ] I have tried removing "sock->sk->sk_sndbuf =3D INT_MAX;" from > socket.c, and the problem still continued. we just want to inform you that we want to run some more tests in the next days, including your suggestions and compare all with fastd (which we also test), and everything also on a different router - we write to the list as soon we have some results on this. Testsetup will stay the same: (see first post for details) laptop (iperf3) > tp841v11 <- WireGuard -> tp841v11 < laptop (iperf3) we will test fastd, default lede Wireguard(20160630), latest wireguard(20160808), latest wireguard without memalloc(step1) , latest wireguard without sndbuf =3D INT_MAX (step2) we will also compare the setup to a similar router : TPLink cpe210 , with 64MB ram and 8 mb flash with similar cpu (maybe not with all tests so far) the ready to use builded packages (Firmware and ipk) are already laying here http://openfreiburg.de/freifunk/firmware/ if some want to play around with them. if there are any suggestions and ideas we could/should add, we are happy to hear from you (its all about hunting the out of memory issue on heavy udp traffic through wireguard tunnel) greetz --=20 make the world nicer, please use PGP encryption --------------F04F4078619B3ED3C105AD7F Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: 8bit On 04.08.2016 13:34, jens wrote:

[ X ] I have tried removing "sk_clear_memalloc(sock);" and "sk_set_memalloc(sock->sk);" from socket.c, and the problem still continued.

[ ] I have tried removing "sock->sk->sk_sndbuf = INT_MAX;" from socket.c, and the problem still continued.

we just want to inform you that we want to run some more tests in the next days, including your suggestions and compare all with fastd (which we also test), and everything also on a different router - we write to the list as soon we have some results on this.

Testsetup will stay the same: (see first post for details)
laptop (iperf3) > tp841v11 <- WireGuard -> tp841v11 < laptop (iperf3)

we will test
    fastd,
    default lede Wireguard(20160630),
    latest wireguard(20160808),
    latest wireguard without memalloc(step1) ,
    latest wireguard without sndbuf = INT_MAX (step2)

we will also compare the setup to a similar router : TPLink cpe210 , with 64MB ram and 8 mb flash with similar cpu (maybe not with all tests so far)

the ready to use builded packages (Firmware and ipk) are already laying here http://openfreiburg.de/freifunk/firmware/
if some want to play around with them.

if there are any suggestions and ideas we could/should add, we are happy to hear from you
(its all about hunting the out of memory issue on heavy udp traffic through wireguard tunnel)

greetz

-- 
make the world nicer, please use PGP encryption

--------------F04F4078619B3ED3C105AD7F--