From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga01.intel.com (mga01.intel.com [192.55.52.88]) by dpdk.org (Postfix) with ESMTP id 30E8B68CD for ; Fri, 5 Dec 2014 18:07:49 +0100 (CET) Received: from fmsmga001.fm.intel.com ([10.253.24.23]) by fmsmga101.fm.intel.com with ESMTP; 05 Dec 2014 09:07:46 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.07,522,1413270000"; d="scan'208";a="633346488" Received: from irsmsx102.ger.corp.intel.com ([163.33.3.155]) by fmsmga001.fm.intel.com with ESMTP; 05 Dec 2014 09:07:45 -0800 Received: from irsmsx105.ger.corp.intel.com ([169.254.7.144]) by IRSMSX102.ger.corp.intel.com ([169.254.2.93]) with mapi id 14.03.0195.001; Fri, 5 Dec 2014 17:07:45 +0000 From: "Ananyev, Konstantin" To: Jean-Mickael Guerin Thread-Topic: [dpdk-dev] [RFC PATCH] ixgbe: ixgbe_recv_pkts_vec shouldn't override mbuf buffer length Thread-Index: AQHQEJ8OviAXTJbJGk65klUTfQ/TFJyBOEsAgAABP4A= Date: Fri, 5 Dec 2014 17:07:44 +0000 Message-ID: <2601191342CEEE43887BDE71AB977258213BD098@IRSMSX105.ger.corp.intel.com> References: <1417792834-20590-1-git-send-email-konstantin.ananyev@intel.com> <5481E456.1050001@6wind.com> In-Reply-To: <5481E456.1050001@6wind.com> Accept-Language: en-IE, en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [163.33.239.180] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "dev@dpdk.org" Subject: Re: [dpdk-dev] [RFC PATCH] ixgbe: ixgbe_recv_pkts_vec shouldn't override mbuf buffer length X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 05 Dec 2014 17:07:50 -0000 > -----Original Message----- > From: Jean-Mickael Guerin [mailto:jean-mickael.guerin@6wind.com] > Sent: Friday, December 05, 2014 4:59 PM > To: Ananyev, Konstantin > Cc: dev@dpdk.org > Subject: Re: [dpdk-dev] [RFC PATCH] ixgbe: ixgbe_recv_pkts_vec shouldn't = override mbuf buffer length >=20 > On 05/12/2014 16:20, Konstantin Ananyev wrote: > > That's an alternative way to fix the problem described in the patch: > > http://dpdk.org/ml/archives/dev/2014-December/009394.html. > > The main difference is: > > - move buf_len fields out of rearm_data marker. > > - make ixgbe_recv_pkts_vec() not touch buf_len field at all > > (as all other RX functions behave). > > > > Signed-off-by: Konstantin Ananyev > > --- > > lib/librte_mbuf/rte_mbuf.h | 7 +++++-- > > lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c | 20 +++++++++++++++----- > > 2 files changed, 20 insertions(+), 7 deletions(-) > > > > diff --git a/lib/librte_mbuf/rte_mbuf.h b/lib/librte_mbuf/rte_mbuf.h > > index 2e5fce5..bb88318 100644 > > --- a/lib/librte_mbuf/rte_mbuf.h > > +++ b/lib/librte_mbuf/rte_mbuf.h > > @@ -179,6 +179,8 @@ const char *rte_get_tx_ol_flag_name(uint64_t mask); > > typedef void *MARKER[0]; /**< generic marker for a point in a st= ructure */ > > typedef uint64_t MARKER64[0]; /**< marker that allows us to overwrite= 8 bytes > > * with a single assignment */ > > +typedef uint8_t MARKER8[0]; /**< generic marker with 1B alignment */ > > + > > /** > > * The generic rte_mbuf, containing a packet mbuf. > > */ > > @@ -188,9 +190,10 @@ struct rte_mbuf { > > void *buf_addr; /**< Virtual address of segment buffer. */ > > phys_addr_t buf_physaddr; /**< Physical address of segment buffer. *= / > > > > - /* next 8 bytes are initialised on RX descriptor rearm */ > > - MARKER64 rearm_data; > > uint16_t buf_len; /**< Length of segment buffer. */ > > + > > + /* next 6 bytes are initialised on RX descriptor rearm */ > > + MARKER8 rearm_data; > > uint16_t data_off; > > > > /** > > diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c b/lib/librte_pmd_ixg= be/ixgbe_rxtx_vec.c > > index 579bc46..d5fc0cc 100644 > > --- a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c > > +++ b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c > > @@ -79,13 +79,22 @@ ixgbe_rxq_rearm(struct igb_rx_queue *rxq) > > /* Initialize the mbufs in vector, process 2 mbufs in one loop */ > > for (i =3D 0; i < RTE_IXGBE_RXQ_REARM_THRESH; i +=3D 2, rxep +=3D 2)= { > > __m128i vaddr0, vaddr1; > > + uintptr_t p0, p1; > > > > mb0 =3D rxep[0].mbuf; > > mb1 =3D rxep[1].mbuf; > > > > - /* flush mbuf with pkt template */ > > - mb0->rearm_data[0] =3D rxq->mbuf_initializer; > > - mb1->rearm_data[0] =3D rxq->mbuf_initializer; > > + /* > > + * Flush mbuf with pkt template. > > + * Data to be rearmed is 6 bytes long. > > + * Though, RX will overwrite ol_flags that are coming next > > + * anyway. So overwrite whole 8 bytes with one load: > > + * 6 bytes of rearm_data plus first 2 bytes of ol_flags. > > + */ > > + p0 =3D (uintptr_t)&mb0->rearm_data; > > + *(uint64_t *)p0 =3D rxq->mbuf_initializer; > > + p1 =3D (uintptr_t)&mb1->rearm_data; > > + *(uint64_t *)p1 =3D rxq->mbuf_initializer; > > > > /* load buf_addr(lo 64bit) and buf_physaddr(hi 64bit) */ > > vaddr0 =3D _mm_loadu_si128((__m128i *)&(mb0->buf_addr)); > > @@ -732,14 +741,15 @@ static struct ixgbe_txq_ops vec_txq_ops =3D { > > int > > ixgbe_rxq_vec_setup(struct igb_rx_queue *rxq) > > { > > + uintptr_t p; > > struct rte_mbuf mb_def =3D { .buf_addr =3D 0 }; /* zeroed mbuf */ > > > > mb_def.nb_segs =3D 1; > > mb_def.data_off =3D RTE_PKTMBUF_HEADROOM; > > - mb_def.buf_len =3D rxq->mb_pool->elt_size - sizeof(struct rte_mbuf); > > mb_def.port =3D rxq->port_id; > > rte_mbuf_refcnt_set(&mb_def, 1); > > - rxq->mbuf_initializer =3D *((uint64_t *)&mb_def.rearm_data); > > + p =3D (uintptr_t)&mb_def.rearm_data; > > + rxq->mbuf_initializer =3D *(uint64_t *)p; > > return 0; > > } > > > > >=20 > The patch introduces writes on unaligned data, but we can assume no > performance penalty on intel hw, correct? >=20 Yes to both: it introduces 64bit unaligned store. I run performance test on IVB board, didn't see any degradation. Konstantin=20