From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-3.8 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS autolearn=no autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 45057C49EA5 for ; Thu, 24 Jun 2021 16:30:36 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 26204613DC for ; Thu, 24 Jun 2021 16:30:36 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231321AbhFXQcy (ORCPT ); Thu, 24 Jun 2021 12:32:54 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:44008 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229464AbhFXQcy (ORCPT ); Thu, 24 Jun 2021 12:32:54 -0400 X-Greylist: delayed 121 seconds by postgrey-1.37 at lindbergh.monkeyblade.net; Thu, 24 Jun 2021 09:30:34 PDT Received: from rin.romanrm.net (rin.romanrm.net [IPv6:2001:bc8:2dd2:1000::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 2CEF1C061574 for ; Thu, 24 Jun 2021 09:30:33 -0700 (PDT) Received: from natsu (natsu2.home.romanrm.net [IPv6:fd39::e99e:8f1b:cfc9:ccb8]) by rin.romanrm.net (Postfix) with SMTP id 25508340 for ; Thu, 24 Jun 2021 16:28:26 +0000 (UTC) Date: Thu, 24 Jun 2021 21:28:25 +0500 From: Roman Mamedov To: kvm@vger.kernel.org Subject: [BUG] vhost-net: Enabling experimental_zcopytx always crashes the host on 5.10, worked fine on 5.4 Message-ID: <20210624212825.32aaa6f9@natsu> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org Hello, I am upgrading my VM host from 5.4.126 to 5.10.46. The host started crashing almost instantly after boot. The issue is traced down to: If the vhost-net module is loaded with "experimental_zcopytx=1", as soon as a Windows KVM guest boots up and starts doing even a light network load (watching YouTube), the entire host crashes with a backtrace pointing to skb_segment, parts of which I was able to capture. This is reproducible 100% of the time, got 3 identical crashes before getting a capture and then flipping the parameter to 0. With it set to 0 the issue goes away. Reported at [1], not sure if that was the best place for the report, thought maybe posting here as well won't hurt. Thanks [1] https://bugzilla.kernel.org/show_bug.cgi?id=213563 -- With respect, Roman