From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga05.intel.com (mga05.intel.com [192.55.52.43]) by dpdk.org (Postfix) with ESMTP id 378522BF3; Fri, 26 Oct 2018 17:41:42 +0200 (CEST) X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from fmsmga006.fm.intel.com ([10.253.24.20]) by fmsmga105.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 26 Oct 2018 08:41:40 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.54,428,1534834800"; d="scan'208";a="275871543" Received: from fmsmsx107.amr.corp.intel.com ([10.18.124.205]) by fmsmga006.fm.intel.com with ESMTP; 26 Oct 2018 08:41:40 -0700 Received: from fmsmsx125.amr.corp.intel.com (10.18.125.40) by fmsmsx107.amr.corp.intel.com (10.18.124.205) with Microsoft SMTP Server (TLS) id 14.3.319.2; Fri, 26 Oct 2018 08:41:40 -0700 Received: from shsmsx104.ccr.corp.intel.com (10.239.4.70) by FMSMSX125.amr.corp.intel.com (10.18.125.40) with Microsoft SMTP Server (TLS) id 14.3.319.2; Fri, 26 Oct 2018 08:41:40 -0700 Received: from shsmsx103.ccr.corp.intel.com ([169.254.4.161]) by SHSMSX104.ccr.corp.intel.com ([169.254.5.117]) with mapi id 14.03.0415.000; Fri, 26 Oct 2018 23:41:38 +0800 From: "Zhang, Qi Z" To: "Ananyev, Konstantin" , "Xing, Beilei" CC: "dev@dpdk.org" , "stable@dpdk.org" Thread-Topic: [dpdk-dev] [PATCH v2] net/i40e: fix Rx instability with vector mode Thread-Index: AQHUbPWr8zsLLC1UvU+LZagc1QKsFqUwsOUAgAD3ZGA= Date: Fri, 26 Oct 2018 15:41:38 +0000 Message-ID: <039ED4275CED7440929022BC67E70611532DBD72@SHSMSX103.ccr.corp.intel.com> References: <1540453252-128654-1-git-send-email-beilei.xing@intel.com> <1540535607-115720-1-git-send-email-beilei.xing@intel.com> <2601191342CEEE43887BDE71AB9772580102FED12C@IRSMSX106.ger.corp.intel.com> In-Reply-To: <2601191342CEEE43887BDE71AB9772580102FED12C@IRSMSX106.ger.corp.intel.com> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-titus-metadata-40: eyJDYXRlZ29yeUxhYmVscyI6IiIsIk1ldGFkYXRhIjp7Im5zIjoiaHR0cDpcL1wvd3d3LnRpdHVzLmNvbVwvbnNcL0ludGVsMyIsImlkIjoiYjQxNzMwYWItZGViNS00NmVmLWI2ODYtNjMyODFlN2JiYzg0IiwicHJvcHMiOlt7Im4iOiJDVFBDbGFzc2lmaWNhdGlvbiIsInZhbHMiOlt7InZhbHVlIjoiQ1RQX05UIn1dfV19LCJTdWJqZWN0TGFiZWxzIjpbXSwiVE1DVmVyc2lvbiI6IjE3LjEwLjE4MDQuNDkiLCJUcnVzdGVkTGFiZWxIYXNoIjoiY2lXak9FdWhER0Q5dm0yXC9LMHRjQnhBcnAyYzBCNHBBQmk2ZHZPXC8xT3NrUDFTZVNTc3lmQTgzSkwzSFRVZlY5In0= x-ctpclassification: CTP_NT dlp-product: dlpe-windows dlp-version: 11.0.400.15 dlp-reaction: no-action x-originating-ip: [10.239.127.40] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Subject: Re: [dpdk-dev] [PATCH v2] net/i40e: fix Rx instability with vector mode 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: Fri, 26 Oct 2018 15:41:43 -0000 > -----Original Message----- > From: Ananyev, Konstantin > Sent: Friday, October 26, 2018 3:47 AM > To: Xing, Beilei ; Zhang, Qi Z > Cc: dev@dpdk.org; stable@dpdk.org > Subject: RE: [dpdk-dev] [PATCH v2] net/i40e: fix Rx instability with vect= or > mode >=20 >=20 > Hi, >=20 > > > > Previously, there is instability during vector Rx if descriptor number > > is not power of 2, e.g. process hang and some Rx packets are > > unexpectedly empty. That's because vector Rx mode assumes Rx > > descriptor number is power of 2 when doing bit mask. > > This patch allows vector mode only when the number of Rx descriptor is > > power of 2. > > > > Fixes: 8e109464c022 ("i40e: allow vector Rx and Tx usage") > > Fixes: a3c83a2527e1 ("net/i40e: enable runtime queue setup") > > Cc: stable@dpdk.org > > > > Signed-off-by: Beilei Xing > > --- > > > > v2 changes: > > - rx_vec_allowed is global configuration, avoid overwrite. > > > > doc/guides/nics/i40e.rst | 7 +++++++ > > drivers/net/i40e/i40e_rxtx.c | 13 ++++++++++++- > > 2 files changed, 19 insertions(+), 1 deletion(-) > > > > diff --git a/doc/guides/nics/i40e.rst b/doc/guides/nics/i40e.rst index > > ab3928a..bfacbd1 100644 > > --- a/doc/guides/nics/i40e.rst > > +++ b/doc/guides/nics/i40e.rst > > @@ -172,6 +172,13 @@ Runtime Config Options > > > > -w 84:00.0,use-latest-supported-vec=3D1 > > > > +Vector RX Pre-conditions > > +~~~~~~~~~~~~~~~~~~~~~~~~ > > +For Vector RX it is assumed that the number of descriptor rings will > > +be a power of 2. With this pre-condition, the ring pointer can easily > > +scroll back to the head after hitting the tail without a conditional > > +check. In addition Vector RX can use this assumption to do a bit mask > using ``ring_size - 1``. > > + > > Driver compilation and testing > > ------------------------------ > > > > diff --git a/drivers/net/i40e/i40e_rxtx.c > > b/drivers/net/i40e/i40e_rxtx.c index a827456..f6dcca0 100644 > > --- a/drivers/net/i40e/i40e_rxtx.c > > +++ b/drivers/net/i40e/i40e_rxtx.c > > @@ -1735,10 +1735,14 @@ i40e_dev_rx_queue_setup_runtime(struct > rte_eth_dev *dev, > > * i40e_set_rx_function. > > */ > > ad->rx_bulk_alloc_allowed =3D true; > > - ad->rx_vec_allowed =3D true; > > dev->data->scattered_rx =3D use_scattered_rx; > > if (use_def_burst_func) > > ad->rx_bulk_alloc_allowed =3D false; > > + /** > > + * Vector mode is allowed only when number of Rx queue > > + * descriptor is a power of 2. > > + */ > > + ad->rx_vec_allowed =3D !(rxq->nb_rx_desc & (rxq->nb_rx_desc - 1)); > > i40e_set_rx_function(dev); > > return 0; > > } >=20 > I think this is not complete - if it is not the first function, we have = to do: > If (if (ad->rx_vec_allowed && (rxq->nb_rx_desc & (rxq->nb_rx_desc - 1) != =3D 0) > { > PMD_DRV_LOG(ERR, ...); > return -EINVAL; > } Yes, we should prevent downgrade if the device is already running. >=20 > BTW, here and below why not to use rte_is_power_of_2()? > Konstantin >=20 > > @@ -1811,6 +1815,13 @@ i40e_dev_rx_queue_setup(struct rte_eth_dev > *dev, > > return -EINVAL; > > } > > > > + /** > > + * Vector mode is allowed only when number of Rx queue > > + * descriptor is a power of 2. > > + */ > > + if (ad->rx_vec_allowed) > > + ad->rx_vec_allowed =3D !(nb_desc & (nb_desc - 1)); Looks like we are not able to get back to vec mode once some queue is set t= o no power of 2, even if all queues has been re-setup to power of 2 again. Seems we still need to check all other queue's configure here. > > + > > /* Free memory if needed */ > > if (dev->data->rx_queues[queue_idx]) { > > i40e_dev_rx_queue_release(dev->data->rx_queues[queue_idx]); >=20 >=20 >=20 >=20 >=20 >=20