From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga02.intel.com (mga02.intel.com [134.134.136.20]) by dpdk.org (Postfix) with ESMTP id B9B267EB0 for ; Tue, 16 Dec 2014 17:35:32 +0100 (CET) Received: from orsmga001.jf.intel.com ([10.7.209.18]) by orsmga101.jf.intel.com with ESMTP; 16 Dec 2014 08:16:54 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.07,587,1413270000"; d="scan'208";a="624668488" Received: from irsmsx106.ger.corp.intel.com ([163.33.3.31]) by orsmga001.jf.intel.com with ESMTP; 16 Dec 2014 08:16:50 -0800 Received: from irsmsx105.ger.corp.intel.com ([169.254.7.144]) by IRSMSX106.ger.corp.intel.com ([169.254.8.18]) with mapi id 14.03.0195.001; Tue, 16 Dec 2014 16:16:50 +0000 From: "Ananyev, Konstantin" To: Neil Horman Thread-Topic: [dpdk-dev] [PATCH 10/17] librte_acl: add AVX2 as new rte_acl_classify() method Thread-Index: AQHQGIBe6R2bzzgwb0yDNvzVCW5u5ZyQ0+9ggABERICAAUFi0A== Date: Tue, 16 Dec 2014 16:16:48 +0000 Message-ID: <2601191342CEEE43887BDE71AB977258213C12AD@IRSMSX105.ger.corp.intel.com> References: <1418580659-12595-1-git-send-email-konstantin.ananyev@intel.com> <1418580659-12595-11-git-send-email-konstantin.ananyev@intel.com> <20141215160009.GC3803@hmsreliant.think-freely.org> <2601191342CEEE43887BDE71AB977258213C0D9C@IRSMSX105.ger.corp.intel.com> <20141215202043.GD3803@hmsreliant.think-freely.org> In-Reply-To: <20141215202043.GD3803@hmsreliant.think-freely.org> Accept-Language: en-IE, en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [163.33.239.182] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "dev@dpdk.org" Subject: Re: [dpdk-dev] [PATCH 10/17] librte_acl: add AVX2 as new rte_acl_classify() method 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: Tue, 16 Dec 2014 16:35:33 -0000 > -----Original Message----- > From: Neil Horman [mailto:nhorman@tuxdriver.com] > Sent: Monday, December 15, 2014 8:21 PM > To: Ananyev, Konstantin > Cc: dev@dpdk.org > Subject: Re: [dpdk-dev] [PATCH 10/17] librte_acl: add AVX2 as new rte_acl= _classify() method >=20 > On Mon, Dec 15, 2014 at 04:33:47PM +0000, Ananyev, Konstantin wrote: > > Hi Neil, > > > > > -----Original Message----- > > > From: Neil Horman [mailto:nhorman@tuxdriver.com] > > > Sent: Monday, December 15, 2014 4:00 PM > > > To: Ananyev, Konstantin > > > Cc: dev@dpdk.org > > > Subject: Re: [dpdk-dev] [PATCH 10/17] librte_acl: add AVX2 as new rte= _acl_classify() method > > > > > > On Sun, Dec 14, 2014 at 06:10:52PM +0000, Konstantin Ananyev wrote: > > > > Introduce new classify() method that uses AVX2 instructions. > > > > From my measurements: > > > > On HSW boards when processing >=3D 16 packets per call, > > > > AVX2 method outperforms it's SSE counterpart by 10-25%, > > > > (depending on the ruleset). > > > > At runtime, this method is selected as default one on HW that suppo= rts AVX2. > > > > > > > > Signed-off-by: Konstantin Ananyev > > > > --- > > > > lib/librte_acl/Makefile | 9 + > > > > lib/librte_acl/acl.h | 4 + > > > > lib/librte_acl/acl_run.h | 2 +- > > > > lib/librte_acl/acl_run_avx2.c | 58 +++++ > > > > lib/librte_acl/acl_run_avx2.h | 305 ++++++++++++++++++++++++ > > > > lib/librte_acl/acl_run_sse.c | 537 +-----------------------------= ------------ > > > > lib/librte_acl/acl_run_sse.h | 533 ++++++++++++++++++++++++++++++= +++++++++++ > > > > lib/librte_acl/rte_acl.c | 5 +- > > > > lib/librte_acl/rte_acl.h | 2 + > > > > 9 files changed, 917 insertions(+), 538 deletions(-) > > > > create mode 100644 lib/librte_acl/acl_run_avx2.c > > > > create mode 100644 lib/librte_acl/acl_run_avx2.h > > > > create mode 100644 lib/librte_acl/acl_run_sse.h > > > > > > > > diff --git a/lib/librte_acl/Makefile b/lib/librte_acl/Makefile > > > > index 65e566d..223ec31 100644 > > > > --- a/lib/librte_acl/Makefile > > > > +++ b/lib/librte_acl/Makefile > > > > @@ -45,8 +45,17 @@ SRCS-$(CONFIG_RTE_LIBRTE_ACL) +=3D acl_bld.c > > > > SRCS-$(CONFIG_RTE_LIBRTE_ACL) +=3D acl_gen.c > > > > SRCS-$(CONFIG_RTE_LIBRTE_ACL) +=3D acl_run_scalar.c > > > > SRCS-$(CONFIG_RTE_LIBRTE_ACL) +=3D acl_run_sse.c > > > > +SRCS-$(CONFIG_RTE_LIBRTE_ACL) +=3D acl_run_avx2.c > > > > > > > > CFLAGS_acl_run_sse.o +=3D -msse4.1 > > > > +ifeq ($(CC), icc) > > > > +CFLAGS_acl_run_avx2.o +=3D -march=3Dcore-avx2 > > > > +else ifneq ($(shell \ > > > > +test $(GCC_MAJOR_VERSION) -le 4 -a $(GCC_MINOR_VERSION) -le 6 && e= cho 1), 1) > > > > +CFLAGS_acl_run_avx2.o +=3D -mavx2 > > > > +else > > > > +CFLAGS_acl_run_avx2.o +=3D -msse4.1 > > > > +endif > > > > > > > This seems broken. You've unilaterally included acl_run_avx2.c in th= e build > > > list above, but only enable -mavx2 if the compiler is at least gcc 4.= 6. > > > > Actually 4.7 (before that version, as I know, gcc doesn't support avx2= ) > > > > > Unless > > > you want to make gcc 4.6 a requirement for building, > > > > I believe DPDK is required to be buildable by gcc 4.6 > > As I remember, we have to support it all way down to gcc 4.3. > > > > > you need to also exclude > > > the file above from the build list. > > > > That means that for gcc 4.6 and below rte_acl_classify_avx2() would no= t be defined. > > And then at runtime, I have to check for that somehow and (re)populate = classify_fns[]. > > Doesn't seems like a good way to me. > There are plenty of ways around that. >=20 > At a minimum you could make the classify_fns array the one place that you= need > to add an ifdef __AVX__ call. >=20 > You could also create a secondary definition of rte_acl_classify_avx2, an= d mark > it as a weak symbol, which only returns -EOPNOTSUPP. That would be good,= since > the right thing will just automatically happen then if you don't build th= e > actual avx2 classification code >=20 > > Instead, I prefer to always build acl_run_avx2.c, > But you can't do that. You just said above that you need to support down= to gcc > 4.3. I see you've worked around that with some additional ifdef __AVX__ > instructions, but in so doing you ignore the possibiity that sse isn't > supported, so you need to add __SSE__ checks now as well. ifdeffing that= much > just isn't scalable. We don't need to worry about compiler without SSE4.1 support. I believe that all compilers that DDPDK has to build with, do support SSE4.= 1. So for SSE4.1 we only has to worry about situation when target CPU doesn't = support it We manage it by runtime selection. For AVX2 - situation is a bit different: it could be both compiler and targ= et CPU that don't support it. > And for your effort, you get an AVX2 classification path > that potentially doesn't actually do vectorized classification. >=20 > It really seems better to me to not build the code if the compiler doesn'= t > support the instruction set it was meant to enable, and change the > classification function pointer to something that informs the user of the= lack > of support at run time. >=20 > > but for old compilers that don't support AVX2 - > > rte_acl_classify_avx2() would simply be identical to rte_acl_classify_s= se(). > > > That doesn't make sense to me, for two reasons: >=20 > 1) What if the machine being targeted doesn't support sse either? >=20 Exactly the same what is happening now on the machine with now SSE4.1 suppo= rt. There is absolutely no difference here. > 2) If an application selects an AVX2 classifier, I as a developer expect = to > either get AVX2 based classification, or an error indicating that I can't= do > AVX2 classification, not a silent performance degradation down to scalar > classification. In fact I was considering both variants for compilers not supporting AVX2: 1. silently degrade to SSE method. 2. create a dummy function rte_acl_classify_error() and put it into class= ify_fns[RTE_ACL_CLASSIFY_AVX2]. I choose #1 because it seems like a less distraction for the user - all would keep working as before, user just wouldn't see any improvement co= mparing to SSE method.=20 Again didn't want to spread "ifdef __AVX2__" into rte_acl.c Though I don't have any strong opinion here. So if you can provide some good reason why #2 is preferable, I am ok to swi= tch to #2.=20 >=20 > > > That in turn I think allows you to remove a > > > bunch of the ifdeffing that you've done in some of the avx2 specific = files. > > > > Actually there are not many of them. > > One in acl_run_avx2.h and another in acl_run_avx2.c. > > > 2 in acl_run_avx2.h and 1 in rte_acl_osdep_alone.h, which is really 3 mor= e than > you need if you just do an intellegent weak classifier function defintion= . grep -n __AVX2__ lib/librte_acl/*.[c,h] | grep -v endif lib/librte_acl/acl_run_avx2.c:45:#ifdef __AVX2__ lib/librte_acl/acl_run_avx2.h:36:#ifdef __AVX2__ rte_acl_osdep_alone.h - is a different story. It needs to be there anyway, as in rte_common_vect.h. In fact rte_acl_osdep_alone.h is only needed for cases when RTE_LIBRTE_ACL= _STANDALONE=3Dy. That comes from the old days, when we had to to support building librte_acl= library without the rest of DPDK. =20 I think we don't need it anymore and plan to remove it. Just thought it should be in a separate patch.=20 Konstantin >=20 > Neil