From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from dpdk.org (dpdk.org [92.243.14.124]) by inbox.dpdk.org (Postfix) with ESMTP id 478CBA32A2 for ; Thu, 24 Oct 2019 19:29:04 +0200 (CEST) Received: from [92.243.14.124] (localhost [127.0.0.1]) by dpdk.org (Postfix) with ESMTP id 9BDE81EBAC; Thu, 24 Oct 2019 19:29:03 +0200 (CEST) Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by dpdk.org (Postfix) with ESMTP id CBEC11EB9A for ; Thu, 24 Oct 2019 19:29:01 +0200 (CEST) X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga007.jf.intel.com ([10.7.209.58]) by orsmga102.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 24 Oct 2019 10:29:00 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.68,225,1569308400"; d="scan'208";a="188658729" Received: from fmsmsx106.amr.corp.intel.com ([10.18.124.204]) by orsmga007.jf.intel.com with ESMTP; 24 Oct 2019 10:29:00 -0700 Received: from fmsmsx120.amr.corp.intel.com (10.18.124.208) by FMSMSX106.amr.corp.intel.com (10.18.124.204) with Microsoft SMTP Server (TLS) id 14.3.439.0; Thu, 24 Oct 2019 10:29:00 -0700 Received: from fmsmsx108.amr.corp.intel.com ([169.254.9.101]) by fmsmsx120.amr.corp.intel.com ([169.254.15.63]) with mapi id 14.03.0439.000; Thu, 24 Oct 2019 10:29:00 -0700 From: "Pathak, Pravin" To: "Trahe, Fiona" , "users@dpdk.org" Thread-Topic: QAT PMD in multi-process mode Thread-Index: AdWKhUF4DlpnJhXiQRaHlCSV9gsEKgABta2AAAEN2CA= Date: Thu, 24 Oct 2019 17:28:59 +0000 Message-ID: <168A68C163D584429EF02A476D52744268BDF7F9@FMSMSX108.amr.corp.intel.com> References: <168A68C163D584429EF02A476D52744268BDF6B3@FMSMSX108.amr.corp.intel.com> <348A99DA5F5B7549AA880327E580B435A217E710@IRSMSX101.ger.corp.intel.com> In-Reply-To: <348A99DA5F5B7549AA880327E580B435A217E710@IRSMSX101.ger.corp.intel.com> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-titus-metadata-40: eyJDYXRlZ29yeUxhYmVscyI6IiIsIk1ldGFkYXRhIjp7Im5zIjoiaHR0cDpcL1wvd3d3LnRpdHVzLmNvbVwvbnNcL0ludGVsMyIsImlkIjoiNWEwMWY4NmUtMWY4Yy00OTJhLTk0NjItMDgzYmZkMzE3Nzg4IiwicHJvcHMiOlt7Im4iOiJDVFBDbGFzc2lmaWNhdGlvbiIsInZhbHMiOlt7InZhbHVlIjoiQ1RQX05UIn1dfV19LCJTdWJqZWN0TGFiZWxzIjpbXSwiVE1DVmVyc2lvbiI6IjE3LjEwLjE4MDQuNDkiLCJUcnVzdGVkTGFiZWxIYXNoIjoiNG45blorMTNqZXpYVG5TNFE0Z2VWRjRIMWtIM0piWENXQ01TTUQwQmc1MEpTeVYrNUg2bXdyMW1QZldwb1FqMSJ9 x-ctpclassification: CTP_NT dlp-product: dlpe-windows dlp-version: 11.2.0.6 dlp-reaction: no-action x-originating-ip: [10.1.200.106] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Subject: Re: [dpdk-users] QAT PMD in multi-process mode X-BeenThere: users@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK usage discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: users-bounces@dpdk.org Sender: "users" Hi Fiona - Thanks. We are thinking of usage where one process only enqueues and other= only dequeues over shared VF. Regards Pravin -----Original Message----- From: Trahe, Fiona=20 Sent: Thursday, October 24, 2019 1:03 PM To: Pathak, Pravin ; users@dpdk.org Cc: Trahe, Fiona Subject: RE: QAT PMD in multi-process mode Hi Pravin, > -----Original Message----- > From: users On Behalf Of Pathak, Pravin > Sent: Thursday, October 24, 2019 5:12 PM > To: users@dpdk.org > Subject: [dpdk-users] QAT PMD in multi-process mode >=20 > Hi - > In multi-process model, if same QAT VF is shared between two processes=20 > which are spawned from separate executables, primary process give segfau= lt while accessing QAT functions. > It looks like PMD installs access functions in shared area so when=20 > secondary process changes those function pointers, they are no longer val= id in primary process. > http://doc.dpdk.org/guides/cryptodevs/qat.html states "Each VF can only be used by one DPDK process. It is not possible to share = the same VF across multiple processes, even if these processes are using di= fferent acceleration services." However, we are currently investigating use of a QAT VF in a primary and se= condary process - but with some constraints about what can be done in each = process, e.g. session init in one process and enqueue/dequeue in the other. > Is QAT PMD going to modified to support this mode in future? > Regards > Pravin >=20 >=20