From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx1.redhat.com (mx1.redhat.com [209.132.183.28]) by dpdk.org (Postfix) with ESMTP id 67B434CA0; Wed, 3 Oct 2018 10:03:06 +0200 (CEST) Received: from smtp.corp.redhat.com (int-mx10.intmail.prod.int.phx2.redhat.com [10.5.11.25]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 897963082B56; Wed, 3 Oct 2018 08:03:05 +0000 (UTC) Received: from [10.36.112.44] (ovpn-112-44.ams2.redhat.com [10.36.112.44]) by smtp.corp.redhat.com (Postfix) with ESMTPS id 73A262010D08; Wed, 3 Oct 2018 08:02:56 +0000 (UTC) To: Ilya Maximets , dev@dpdk.org, tiwei.bie@intel.com, zhihong.wang@intel.com, jfreimann@redhat.com, nicknickolaev@gmail.com, bruce.richardson@intel.com, alejandro.lucero@netronome.com Cc: dgilbert@redhat.com, stable@dpdk.org, "Michael S. Tsirkin" References: <20181002093651.24795-1-maxime.coquelin@redhat.com> <20181002093651.24795-2-maxime.coquelin@redhat.com> <20181002141315eucas1p16c87759329eeb374528bcb70a2d71ee4~Z0CLLGpN92076620766eucas1p1R@eucas1p1.samsung.com> <2a57953d-67c6-26f3-f65f-4e5a1dcf1474@redhat.com> <20181003075530eucas1p1dd7191a728c1129fd5d9dbaed5fa1047~aChpN_6wn1404114041eucas1p1H@eucas1p1.samsung.com> From: Maxime Coquelin Message-ID: Date: Wed, 3 Oct 2018 10:02:54 +0200 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.9.1 MIME-Version: 1.0 In-Reply-To: <20181003075530eucas1p1dd7191a728c1129fd5d9dbaed5fa1047~aChpN_6wn1404114041eucas1p1H@eucas1p1.samsung.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 8bit X-Scanned-By: MIMEDefang 2.84 on 10.5.11.25 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.45]); Wed, 03 Oct 2018 08:03:05 +0000 (UTC) Subject: Re: [dpdk-dev] [PATCH v2 01/17] vhost: fix messages error checks X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 03 Oct 2018 08:03:06 -0000 On 10/03/2018 09:57 AM, Ilya Maximets wrote: > On 03.10.2018 10:50, Maxime Coquelin wrote: >> >> >> On 10/02/2018 04:15 PM, Ilya Maximets wrote: >>> On 02.10.2018 12:36, Maxime Coquelin wrote: >>>> Return of message handling has now changed to an enum that can >>>> take non-negative value that is not zero in case a reply is >>>> needed. But the code checking the variable afterwards has not >>>> been updated, leading to success messages handling being >>>> treated as errors. >>>> >>>> Fixes: 4e601952cae6 ("vhost: message handling implemented as a callback array") >>>> >>>> Signed-off-by: Maxime Coquelin >>>> --- >>>>   lib/librte_vhost/vhost_user.c | 6 +++--- >>>>   1 file changed, 3 insertions(+), 3 deletions(-) >>>> >>>> diff --git a/lib/librte_vhost/vhost_user.c b/lib/librte_vhost/vhost_user.c >>>> index 7ef3fb4a4..060b41893 100644 >>>> --- a/lib/librte_vhost/vhost_user.c >>>> +++ b/lib/librte_vhost/vhost_user.c >>>> @@ -1783,7 +1783,7 @@ vhost_user_msg_handler(int vid, int fd) >>>>       } >>>>     skip_to_post_handle: >>>> -    if (!ret && dev->extern_ops.post_msg_handle) { >>>> +    if (ret != VH_RESULT_ERR && dev->extern_ops.post_msg_handle) { >>>>           uint32_t need_reply; >>>>             ret = (*dev->extern_ops.post_msg_handle)( >>>> @@ -1800,10 +1800,10 @@ vhost_user_msg_handler(int vid, int fd) >>>>           vhost_user_unlock_all_queue_pairs(dev); >>>>         if (msg.flags & VHOST_USER_NEED_REPLY) { >>> >>> Maybe we need to reply here only if we didn't reply >>> already (not VH_RESULT_REPLY) ? Otherwise, we could >>> reply twice (with payload and with return code). >> >> Well, if the master sets this bit, it means it is waiting for >> a "reply-ack", so not sending it would cause the master to wait >> forever. >> >> It is the master responsibility to not set this bit for requests >> already expecting a non "reply-ack" reply (as you fixed it for >> postcopy's set mem table case). > > vhost-user docs in QEMU says: > " > For the message types that already solicit a reply from the client, the > presence of VHOST_USER_PROTOCOL_F_REPLY_ACK or need_reply bit being set brings > no behavioural change. > " > i.e. even if QEMU sets the need_reply flag, vhost should not reply twice. > Am I missing something? Oh, right. Thanks for pointing it out. So coming back to the DPDK implementation, I just had a look again, and it seems that we don't send a reply twice, as send_vhost_reply takes care of clearing the VHOST_USER_NEED_REPLY flag. Do you confirm my understanding is correct? >> >>>> -        msg.payload.u64 = !!ret; >>>> +        msg.payload.u64 = ret == VH_RESULT_ERR; >>>>           msg.size = sizeof(msg.payload.u64); >>>>           send_vhost_reply(fd, &msg); >>>> -    } else if (ret) { >>>> +    } else if (ret == VH_RESULT_ERR) { >>>>           RTE_LOG(ERR, VHOST_CONFIG, >>>>               "vhost message handling failed.\n"); >>>>           return -1; >>>> >> >>