From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga17.intel.com (mga17.intel.com [192.55.52.151]) by dpdk.org (Postfix) with ESMTP id B190E2862 for ; Tue, 6 Nov 2018 09:04:10 +0100 (CET) X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga001.jf.intel.com ([10.7.209.18]) by fmsmga107.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 06 Nov 2018 00:04:09 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.54,471,1534834800"; d="scan'208";a="105774193" Received: from fmsmsx103.amr.corp.intel.com ([10.18.124.201]) by orsmga001.jf.intel.com with ESMTP; 06 Nov 2018 00:04:09 -0800 Received: from fmsmsx125.amr.corp.intel.com (10.18.125.40) by FMSMSX103.amr.corp.intel.com (10.18.124.201) with Microsoft SMTP Server (TLS) id 14.3.408.0; Tue, 6 Nov 2018 00:04:08 -0800 Received: from HASMSX109.ger.corp.intel.com (10.184.198.21) by FMSMSX125.amr.corp.intel.com (10.18.125.40) with Microsoft SMTP Server (TLS) id 14.3.408.0; Tue, 6 Nov 2018 00:04:07 -0800 Received: from hasmsx112.ger.corp.intel.com ([169.254.11.97]) by hasmsx109.ger.corp.intel.com ([169.254.3.7]) with mapi id 14.03.0415.000; Tue, 6 Nov 2018 10:04:04 +0200 From: "Jozwiak, TomaszX" To: "Verma, Shally" , "dev@dpdk.org" , "Trahe, Fiona" , "akhil.goyal@nxp.com" Thread-Topic: [dpdk-dev] [PATCH 2/3] app/compress-perf: add performance measurement Thread-Index: AQHUYhSjn5ZjkvYiqUCGRBSCtwq5y6U2CV0AgArVnACAAGXs8A== Date: Tue, 6 Nov 2018 08:04:04 +0000 Message-ID: References: <1538400427-20164-1-git-send-email-tomaszx.jozwiak@intel.com> <1538400427-20164-3-git-send-email-tomaszx.jozwiak@intel.com> In-Reply-To: Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: dlp-product: dlpe-windows dlp-version: 11.0.400.15 dlp-reaction: no-action x-originating-ip: [10.104.12.169] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Subject: Re: [dpdk-dev] [PATCH 2/3] app/compress-perf: add performance measurement 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: Tue, 06 Nov 2018 08:04:11 -0000 Hi Shally, Please see my comment inline. > -----Original Message----- > From: Verma, Shally [mailto:Shally.Verma@cavium.com] > Sent: Monday, November 5, 2018 9:34 AM > To: Jozwiak, TomaszX ; dev@dpdk.org; Trahe, > Fiona ; akhil.goyal@nxp.com; De Lara Guarch, Pablo > > Cc: De@dpdk.org; Lara@dpdk.org; Guarch@dpdk.org > Subject: RE: [dpdk-dev] [PATCH 2/3] app/compress-perf: add performance > measurement >=20 >=20 >=20 > >-----Original Message----- > >From: Jozwiak, TomaszX > >Sent: 02 November 2018 15:29 > >To: dev@dpdk.org; Trahe, Fiona ; > >akhil.goyal@nxp.com; Verma, Shally ; De Lara > >Guarch, Pablo > >Cc: De@dpdk.org; Lara@dpdk.org; Guarch@dpdk.org > >Subject: RE: [dpdk-dev] [PATCH 2/3] app/compress-perf: add performance > >measurement > > > >External Email > > > >Hi Shally, > > > >Sorry for delay - I was on sick leave. > >We had some issues with dynamic compression test so I block this test > >in V2. May be there's too late to add this into this release but we've d= ecided > to send this V2 to DPDK. > > > >My comment inline (not all have answer so far, still working on that) > > > >> -----Original Message----- > >> From: Verma, Shally [mailto:Shally.Verma@cavium.com] > >> Sent: Friday, October 12, 2018 12:16 PM > >> To: Jozwiak, TomaszX ; dev@dpdk.org; > >> Trahe, Fiona ; akhil.goyal@nxp.com; De Lara > >> Guarch, Pablo > >> Cc: De@dpdk.org; Lara@dpdk.org; Guarch@dpdk.org > >> Subject: RE: [dpdk-dev] [PATCH 2/3] app/compress-perf: add > >> performance measurement > >> > >> HI TomaszX > >> > >> Sorry for delay in response. Comments inline. > >> > >> >-----Original Message----- > >> >From: dev On Behalf Of Tomasz Jozwiak > >> >Sent: 01 October 2018 18:57 > >> >To: dev@dpdk.org; fiona.trahe@intel.com; tomaszx.jozwiak@intel.com; > >> >akhil.goyal@nxp.com; pablo.de.lara.guarch@intel.com > >> >Cc: De@dpdk.org; Lara@dpdk.org; Guarch@dpdk.org > >> >Subject: [dpdk-dev] [PATCH 2/3] app/compress-perf: add performance > >> >measurement > >> > > >> >External Email > >> > > >> >Added performance measurement part into compression perf. test. > >> > > >> >Signed-off-by: De Lara Guarch, Pablo > >> > > >> >Signed-off-by: Tomasz Jozwiak > >> >--- > >> > app/test-compress-perf/main.c | 844 > >> >++++++++++++++++++++++++++++++++++++++++++ > >> > 1 file changed, 844 insertions(+) > >> > > >> >diff --git a/app/test-compress-perf/main.c > >> >b/app/test-compress-perf/main.c index f52b98d..093dfaf 100644 > >> >--- a/app/test-compress-perf/main.c > >> >+++ b/app/test-compress-perf/main.c > >> >@@ -5,13 +5,721 @@ > >> > #include > >> > #include > >> > #include > >> >+#include > >> > #include > >> > > >> > #include "comp_perf_options.h" > >> > > >> >+#define NUM_MAX_XFORMS 16 > >> >+#define NUM_MAX_INFLIGHT_OPS 512 > >> >+#define EXPANSE_RATIO 1.05 > >> >+#define MIN_ISAL_SIZE 8 > >> >+ > >> >+#define DIV_CEIL(a, b) ((a) / (b) + ((a) % (b) !=3D 0)) > >> >+ > >> >+static int > >> >+param_range_check(uint16_t size, const struct rte_param_log2_range > >> >+*range) { > >> >+ unsigned int next_size; > >> >+ > >> >+ /* Check lower/upper bounds */ > >> >+ if (size < range->min) > >> >+ return -1; > >> >+ > >> >+ if (size > range->max) > >> >+ return -1; > >> >+ > >> >+ /* If range is actually only one value, size is correct */ > >> >+ if (range->increment =3D=3D 0) > >> >+ return 0; > >> >+ > >> >+ /* Check if value is one of the supported sizes */ > >> >+ for (next_size =3D range->min; next_size <=3D range->max; > >> >+ next_size +=3D range->increment) > >> >+ if (size =3D=3D next_size) > >> >+ return 0; > >> >+ > >> >+ return -1; > >> >+} > >> >+ > >> >+static int > >> >+comp_perf_check_capabilities(struct comp_test_data *test_data) { > >> >+ const struct rte_compressdev_capabilities *cap; > >> >+ > >> >+ cap =3D rte_compressdev_capability_get(test_data->cdev_id, > >> >+ RTE_COMP_ALGO_DEFLATE); > >> >+ > >> >+ if (cap =3D=3D NULL) { > >> >+ RTE_LOG(ERR, USER1, > >> >+ "Compress device does not support DEFLATE\n")= ; > >> >+ return -1; > >> >+ } > >> >+ > >> >+ uint64_t comp_flags =3D cap->comp_feature_flags; > >> >+ > >> >+ /* Huffman enconding */ > >> >+ if (test_data->huffman_enc =3D=3D RTE_COMP_HUFFMAN_FIXED && > >> >+ (comp_flags & RTE_COMP_FF_HUFFMAN_FIXED) =3D= =3D 0) { > >> >+ RTE_LOG(ERR, USER1, > >> >+ "Compress device does not supported Fixed Huf= fman\n"); > >> >+ return -1; > >> >+ } > >> >+ > >> >+ if (test_data->huffman_enc =3D=3D RTE_COMP_HUFFMAN_DYNAMIC > && > >> >+ (comp_flags & RTE_COMP_FF_HUFFMAN_DYNAMIC) = =3D=3D 0) > { > >> >+ RTE_LOG(ERR, USER1, > >> >+ "Compress device does not supported Dynamic > Huffman\n"); > >> >+ return -1; > >> >+ } > >> >+ > >> >+ /* Window size */ > >> >+ if (test_data->window_sz !=3D -1) { > >> >+ if (param_range_check(test_data->window_sz, > >> >+ &cap->window_size) > >> What if cap->window_size is 0 i.e. implementation default? > > > >TJ: You probably mean cap->window_size.increment =3D 0 (because > >cap->window_size is a structure). In that case we check if > >test_data->window_sz >=3Dmin and test_data->window_sz <=3D max only, > because increment =3D 0 means (base on compression API) we have only one > value of windows_size (no range is supported). > But PMD can set min and max too 0 for such case. TJ: I can't see any issue in that case too. Maybe I don't understand what y= ou mean but the logic is as follow: 1) if you pass '--window-sz ...' param. into command line your intention = is to force that value of window size during test. We check is this value i= s allow (by param_range_check() function). 2) if you plan to use default value - just don't pass '--window-sz' param. = in command line at all. In that case we get windows size from window_size.m= ax field, so if window_size.min=3D window_size.max=3D0=20 test_data->window_sz will be zero, as well.=20 If you mean that behavior is not good - I will be grateful for other sugges= tions. >=20 > > > > > > > .... >=20 > >> >+ > >> >+ if (fread(data, data_to_read, 1, f) !=3D 1) { > >> >+ RTE_LOG(ERR, USER1, "Input file could not be = read\n"); > >> >+ goto err; > >> >+ } > >> >+ if (fseek(f, 0, SEEK_SET) !=3D 0) { > >> >+ RTE_LOG(ERR, USER1, > >> >+ "Size of input could not be calculate= d\n"); > >> >+ goto err; > >> >+ } > >> >+ remaining_data -=3D data_to_read; > >> >+ data +=3D data_to_read; > >> It looks like it will run 2nd time only if input file size < input > >> data size in which case it will just keep filling input buffer with re= peated > data. > >> Is that the intention here? > > > >TJ: Yes exactly. If test_data->input_data_sz is bigger than > >actual_file_sz then we fill the buffer with repeated data from file to f= ill > whole buffer. > I mentioned in one of the earlier reply, wont that then influence the > compression behaviour and o/p? my suggestion was to work on actual user > provided input to take perf to get actual perf for given content. TJ: You right, but this solution is flexible. You can pass ' --extended-inp= ut-sz" or not, so you can use original input data or extend it if you want. >=20 > > > >> > ... >=20 > >> >+ if (data_addr =3D=3D NULL) { > >> >+ RTE_LOG(ERR, USER1, "Could not > >> >+ append data\n"); > >> Since a new buffer per segment is allocated, so is it possible for > >> append to fail? think, this check is redundant here. > > > >TJ: Yes, you're right, it should never fail. But I think it's good codin= g practice > to add the check just in case. > > > Unless it is called in data path which might cost perf a bit. TJ: prepare_bufs() is out of perf measurement, so shouldn't impact to meas= urements. The performance measurement is inside main_loop() only. Br, Tomek >=20 > Thanks > Shally >=20 > >> >+ return -1; > >> >+ } > >> >+ > >> >+ rte_memcpy(data_addr, input_data_ptr, data_sz= ); > >> >+ input_data_ptr +=3D data_sz; > >> >+ remaining_data -=3D data_sz; > >> >+ > >> >+ if (rte_pktmbuf_chain(test_data->decomp_bufs[= i], > >> >+ next_seg) < 0) { > >> >+ RTE_LOG(ERR, USER1, "Could not chain = mbufs\n"); > >> >+ return -1; > >> >+ } > >> >+ segs_per_mbuf++; > >> >+ } > >> >+ > >> >+ /* Allocate data in output mbuf */ > >> >+ test_data->comp_bufs[i] =3D > >> >+ rte_pktmbuf_alloc(test_data->comp_buf_pool); > >> >+ if (test_data->comp_bufs[i] =3D=3D NULL) { > >> >+ RTE_LOG(ERR, USER1, "Could not allocate mbuf\= n"); > >> >+ return -1; > >> >+ } > >> >+ data_addr =3D (uint8_t *) rte_pktmbuf_append( > >> >+ test_data->comp_bufs[i], > >> >+ test_data->seg_sz); > >> >+ if (data_addr =3D=3D NULL) { > >> >+ RTE_LOG(ERR, USER1, "Could not append data\n"= ); > >> >+ return -1; > >> >+ } > >> >+ > >> >+ /* Chain mbufs if needed for output mbufs */ > >> >+ for (j =3D 1; j < segs_per_mbuf; j++) { > >> >+ struct rte_mbuf *next_seg =3D > >> >+ > >> >+ rte_pktmbuf_alloc(test_data->comp_buf_pool); > >> >+ > >> >+ if (next_seg =3D=3D NULL) { > >> >+ RTE_LOG(ERR, USER1, > >> >+ "Could not allocate mbuf\n"); > >> >+ return -1; > >> >+ } > >> >+ > >> >+ data_addr =3D (uint8_t *)rte_pktmbuf_append(n= ext_seg, > >> >+ test_data->seg_sz); > >> >+ > >> >+ if (data_addr =3D=3D NULL) { > >> >+ RTE_LOG(ERR, USER1, "Could not append= data\n"); > >> >+ return -1; > >> >+ } > >> >+ > >> >+ if (rte_pktmbuf_chain(test_data->comp_bufs[i]= , > >> >+ next_seg) < 0) { > >> >+ RTE_LOG(ERR, USER1, "Could not chain = mbufs\n"); > >> >+ return -1; > >> >+ } > >> >+ } > >> >+ } > >> >+ > >> >+ return 0; > >> >+} > >> >+ > >> >+static void > >> >+free_bufs(struct comp_test_data *test_data) { > >> >+ uint32_t i; > >> >+ > >> >+ for (i =3D 0; i < test_data->total_bufs; i++) { > >> >+ rte_pktmbuf_free(test_data->comp_bufs[i]); > >> >+ rte_pktmbuf_free(test_data->decomp_bufs[i]); > >> >+ } > >> >+ rte_free(test_data->comp_bufs); > >> >+ rte_free(test_data->decomp_bufs); } > >> >+ > >> >+static int > >> >+main_loop(struct comp_test_data *test_data, uint8_t level, > >> >+ enum rte_comp_xform_type type, > >> >+ uint8_t *output_data_ptr, > >> >+ size_t *output_data_sz, > >> >+ unsigned int benchmarking) { > >> >+ uint8_t dev_id =3D test_data->cdev_id; > >> >+ uint32_t i, iter, num_iter; > >> >+ struct rte_comp_op **ops, **deq_ops; > >> >+ void *priv_xform =3D NULL; > >> >+ struct rte_comp_xform xform; > >> >+ size_t output_size =3D 0; > >> >+ struct rte_mbuf **input_bufs, **output_bufs; > >> >+ int res =3D 0; > >> >+ int allocated =3D 0; > >> >+ > >> >+ if (test_data =3D=3D NULL || !test_data->burst_sz) { > >> >+ RTE_LOG(ERR, USER1, > >> >+ "Unknow burst size\n"); > >> >+ return -1; > >> >+ } > >> >+ > >> >+ ops =3D rte_zmalloc_socket(NULL, > >> >+ 2 * test_data->total_bufs * sizeof(struct rte_comp_op= *), > >> >+ 0, rte_socket_id()); > >> >+ > >> >+ if (ops =3D=3D NULL) { > >> >+ RTE_LOG(ERR, USER1, > >> >+ "Can't allocate memory for ops strucures\n"); > >> >+ return -1; > >> >+ } > >> >+ > >> >+ deq_ops =3D &ops[test_data->total_bufs]; > >> >+ > >> >+ if (type =3D=3D RTE_COMP_COMPRESS) { > >> >+ xform =3D (struct rte_comp_xform) { > >> >+ .type =3D RTE_COMP_COMPRESS, > >> >+ .compress =3D { > >> >+ .algo =3D RTE_COMP_ALGO_DEFLATE, > >> >+ .deflate.huffman =3D test_data->huffm= an_enc, > >> >+ .level =3D level, > >> >+ .window_size =3D test_data->window_sz= , > >> >+ .chksum =3D RTE_COMP_CHECKSUM_NONE, > >> >+ .hash_algo =3D RTE_COMP_HASH_ALGO_NON= E > >> >+ } > >> >+ }; > >> >+ input_bufs =3D test_data->decomp_bufs; > >> >+ output_bufs =3D test_data->comp_bufs; > >> >+ } else { > >> >+ xform =3D (struct rte_comp_xform) { > >> >+ .type =3D RTE_COMP_DECOMPRESS, > >> >+ .decompress =3D { > >> >+ .algo =3D RTE_COMP_ALGO_DEFLATE, > >> >+ .chksum =3D RTE_COMP_CHECKSUM_NONE, > >> >+ .window_size =3D test_data->window_sz= , > >> >+ .hash_algo =3D RTE_COMP_HASH_ALGO_NON= E > >> >+ } > >> >+ }; > >> >+ input_bufs =3D test_data->comp_bufs; > >> >+ output_bufs =3D test_data->decomp_bufs; > >> >+ } > >> >+ > >> >+ /* Create private xform */ > >> >+ if (rte_compressdev_private_xform_create(dev_id, &xform, > >> >+ &priv_xform) < 0) { > >> >+ RTE_LOG(ERR, USER1, "Private xform could not be creat= ed\n"); > >> >+ res =3D -1; > >> >+ goto end; > >> >+ } > >> >+ > >> >+ uint64_t tsc_start, tsc_end, tsc_duration; > >> >+ > >> >+ tsc_start =3D tsc_end =3D tsc_duration =3D 0; > >> >+ if (benchmarking) { > >> >+ tsc_start =3D rte_rdtsc(); > >> >+ num_iter =3D test_data->num_iter; > >> >+ } else > >> >+ num_iter =3D 1; > >> Looks like in same code we're doing benchmarking and functional > validation. > >> It can be reorganised to keep validation test separately like done in > >> crypto_perf. > > > >TJ: Ok, makes sense. However in the interests of getting this into the > >18.11 release I'd like to defer this refactoring and the remainder of yo= ur > comments below to the next release. > > > > > >Next comments - WIP > > > > > >Br, Tomek