Patch Detail
get:
Show a patch.
patch:
Update a patch.
put:
Update a patch.
GET /api/patches/2340/?format=api
https://patches.dpdk.org/api/patches/2340/?format=api", "web_url": "https://patches.dpdk.org/project/dpdk/patch/1421632414-10027-4-git-send-email-zhihong.wang@intel.com/", "project": { "id": 1, "url": "https://patches.dpdk.org/api/projects/1/?format=api", "name": "DPDK", "link_name": "dpdk", "list_id": "dev.dpdk.org", "list_email": "dev@dpdk.org", "web_url": "http://core.dpdk.org", "scm_url": "git://dpdk.org/dpdk", "webscm_url": "http://git.dpdk.org/dpdk", "list_archive_url": "https://inbox.dpdk.org/dev", "list_archive_url_format": "https://inbox.dpdk.org/dev/{}", "commit_url_format": "" }, "msgid": "<1421632414-10027-4-git-send-email-zhihong.wang@intel.com>", "list_archive_url": "https://inbox.dpdk.org/dev/1421632414-10027-4-git-send-email-zhihong.wang@intel.com", "date": "2015-01-19T01:53:33", "name": "[dpdk-dev,3/4] app/test: Extended test coverage in test_memcpy_perf.c", "commit_ref": null, "pull_url": null, "state": "superseded", "archived": true, "hash": "ab0cb4db6ded1110808786e6601709406a10ad76", "submitter": { "id": 156, "url": "https://patches.dpdk.org/api/people/156/?format=api", "name": "Zhihong Wang", "email": "zhihong.wang@intel.com" }, "delegate": null, "mbox": "https://patches.dpdk.org/project/dpdk/patch/1421632414-10027-4-git-send-email-zhihong.wang@intel.com/mbox/", "series": [], "comments": "https://patches.dpdk.org/api/patches/2340/comments/", "check": "pending", "checks": "https://patches.dpdk.org/api/patches/2340/checks/", "tags": {}, "related": [], "headers": { "Return-Path": "<dev-bounces@dpdk.org>", "X-Original-To": "patchwork@dpdk.org", "Delivered-To": "patchwork@dpdk.org", "Received": [ "from [92.243.14.124] (localhost [IPv6:::1])\n\tby dpdk.org (Postfix) with ESMTP id 469065A95;\n\tMon, 19 Jan 2015 02:53:52 +0100 (CET)", "from mga01.intel.com (mga01.intel.com [192.55.52.88])\n\tby dpdk.org (Postfix) with ESMTP id B40C65A8E\n\tfor <dev@dpdk.org>; Mon, 19 Jan 2015 02:53:49 +0100 (CET)", "from fmsmga001.fm.intel.com ([10.253.24.23])\n\tby fmsmga101.fm.intel.com with ESMTP; 18 Jan 2015 17:53:48 -0800", "from shvmail01.sh.intel.com ([10.239.29.42])\n\tby fmsmga001.fm.intel.com with ESMTP; 18 Jan 2015 17:53:47 -0800", "from shecgisg004.sh.intel.com (shecgisg004.sh.intel.com\n\t[10.239.29.89])\n\tby shvmail01.sh.intel.com with ESMTP id t0J1rk8X001428;\n\tMon, 19 Jan 2015 09:53:46 +0800", "from shecgisg004.sh.intel.com (localhost [127.0.0.1])\n\tby shecgisg004.sh.intel.com (8.13.6/8.13.6/SuSE Linux 0.8) with ESMTP\n\tid t0J1rhB3010083; Mon, 19 Jan 2015 09:53:45 +0800", "(from zwang84@localhost)\n\tby shecgisg004.sh.intel.com (8.13.6/8.13.6/Submit) id t0J1rho1010079; \n\tMon, 19 Jan 2015 09:53:43 +0800" ], "X-ExtLoop1": "1", "X-IronPort-AV": "E=Sophos;i=\"5.09,422,1418112000\"; d=\"scan'208\";a=\"652964009\"", "From": "zhihong.wang@intel.com", "To": "dev@dpdk.org", "Date": "Mon, 19 Jan 2015 09:53:33 +0800", "Message-Id": "<1421632414-10027-4-git-send-email-zhihong.wang@intel.com>", "X-Mailer": "git-send-email 1.7.4.1", "In-Reply-To": "<1421632414-10027-1-git-send-email-zhihong.wang@intel.com>", "References": "<1421632414-10027-1-git-send-email-zhihong.wang@intel.com>", "Subject": "[dpdk-dev] [PATCH 3/4] app/test: Extended test coverage in\n\ttest_memcpy_perf.c", "X-BeenThere": "dev@dpdk.org", "X-Mailman-Version": "2.1.15", "Precedence": "list", "List-Id": "patches and discussions about DPDK <dev.dpdk.org>", "List-Unsubscribe": "<http://dpdk.org/ml/options/dev>,\n\t<mailto:dev-request@dpdk.org?subject=unsubscribe>", "List-Archive": "<http://dpdk.org/ml/archives/dev/>", "List-Post": "<mailto:dev@dpdk.org>", "List-Help": "<mailto:dev-request@dpdk.org?subject=help>", "List-Subscribe": "<http://dpdk.org/ml/listinfo/dev>,\n\t<mailto:dev-request@dpdk.org?subject=subscribe>", "Errors-To": "dev-bounces@dpdk.org", "Sender": "\"dev\" <dev-bounces@dpdk.org>" }, "content": "Main code changes:\n\n1. Added more typical data points for a thorough performance test\n\n2. Added unaligned test cases since it's common in DPDK usage\n\nSigned-off-by: Zhihong Wang <zhihong.wang@intel.com>\n---\n app/test/test_memcpy_perf.c | 238 +++++++++++++++++++++++++++++---------------\n 1 file changed, 156 insertions(+), 82 deletions(-)", "diff": "diff --git a/app/test/test_memcpy_perf.c b/app/test/test_memcpy_perf.c\nindex 7809610..4875af1 100644\n--- a/app/test/test_memcpy_perf.c\n+++ b/app/test/test_memcpy_perf.c\n@@ -54,9 +54,10 @@\n /* List of buffer sizes to test */\n #if TEST_VALUE_RANGE == 0\n static size_t buf_sizes[] = {\n-\t0, 1, 7, 8, 9, 15, 16, 17, 31, 32, 33, 63, 64, 65, 127, 128, 129, 255,\n-\t256, 257, 320, 384, 511, 512, 513, 1023, 1024, 1025, 1518, 1522, 1600,\n-\t2048, 3072, 4096, 5120, 6144, 7168, 8192\n+\t1, 2, 3, 4, 5, 6, 7, 8, 9, 12, 15, 16, 17, 31, 32, 33, 63, 64, 65, 127, 128,\n+\t129, 191, 192, 193, 255, 256, 257, 319, 320, 321, 383, 384, 385, 447, 448,\n+\t449, 511, 512, 513, 767, 768, 769, 1023, 1024, 1025, 1518, 1522, 1536, 1600,\n+\t2048, 2560, 3072, 3584, 4096, 4608, 5120, 5632, 6144, 6656, 7168, 7680, 8192\n };\n /* MUST be as large as largest packet size above */\n #define SMALL_BUFFER_SIZE 8192\n@@ -78,7 +79,7 @@ static size_t buf_sizes[TEST_VALUE_RANGE];\n #define TEST_BATCH_SIZE 100\n \n /* Data is aligned on this many bytes (power of 2) */\n-#define ALIGNMENT_UNIT 16\n+#define ALIGNMENT_UNIT 32\n \n /*\n * Pointers used in performance tests. The two large buffers are for uncached\n@@ -94,19 +95,19 @@ init_buffers(void)\n {\n \tunsigned i;\n \n-\tlarge_buf_read = rte_malloc(\"memcpy\", LARGE_BUFFER_SIZE, ALIGNMENT_UNIT);\n+\tlarge_buf_read = rte_malloc(\"memcpy\", LARGE_BUFFER_SIZE + ALIGNMENT_UNIT, ALIGNMENT_UNIT);\n \tif (large_buf_read == NULL)\n \t\tgoto error_large_buf_read;\n \n-\tlarge_buf_write = rte_malloc(\"memcpy\", LARGE_BUFFER_SIZE, ALIGNMENT_UNIT);\n+\tlarge_buf_write = rte_malloc(\"memcpy\", LARGE_BUFFER_SIZE + ALIGNMENT_UNIT, ALIGNMENT_UNIT);\n \tif (large_buf_write == NULL)\n \t\tgoto error_large_buf_write;\n \n-\tsmall_buf_read = rte_malloc(\"memcpy\", SMALL_BUFFER_SIZE, ALIGNMENT_UNIT);\n+\tsmall_buf_read = rte_malloc(\"memcpy\", SMALL_BUFFER_SIZE + ALIGNMENT_UNIT, ALIGNMENT_UNIT);\n \tif (small_buf_read == NULL)\n \t\tgoto error_small_buf_read;\n \n-\tsmall_buf_write = rte_malloc(\"memcpy\", SMALL_BUFFER_SIZE, ALIGNMENT_UNIT);\n+\tsmall_buf_write = rte_malloc(\"memcpy\", SMALL_BUFFER_SIZE + ALIGNMENT_UNIT, ALIGNMENT_UNIT);\n \tif (small_buf_write == NULL)\n \t\tgoto error_small_buf_write;\n \n@@ -140,25 +141,25 @@ free_buffers(void)\n \n /*\n * Get a random offset into large array, with enough space needed to perform\n- * max copy size. Offset is aligned.\n+ * max copy size. Offset is aligned, uoffset is used for unalignment setting.\n */\n static inline size_t\n-get_rand_offset(void)\n+get_rand_offset(size_t uoffset)\n {\n-\treturn ((rte_rand() % (LARGE_BUFFER_SIZE - SMALL_BUFFER_SIZE)) &\n-\t ~(ALIGNMENT_UNIT - 1));\n+\treturn (((rte_rand() % (LARGE_BUFFER_SIZE - SMALL_BUFFER_SIZE)) &\n+\t\t\t~(ALIGNMENT_UNIT - 1)) + uoffset);\n }\n \n /* Fill in source and destination addresses. */\n static inline void\n-fill_addr_arrays(size_t *dst_addr, int is_dst_cached,\n-\t\tsize_t *src_addr, int is_src_cached)\n+fill_addr_arrays(size_t *dst_addr, int is_dst_cached, size_t dst_uoffset,\n+\t\t\t\t size_t *src_addr, int is_src_cached, size_t src_uoffset)\n {\n \tunsigned int i;\n \n \tfor (i = 0; i < TEST_BATCH_SIZE; i++) {\n-\t\tdst_addr[i] = (is_dst_cached) ? 0 : get_rand_offset();\n-\t\tsrc_addr[i] = (is_src_cached) ? 0 : get_rand_offset();\n+\t\tdst_addr[i] = (is_dst_cached) ? dst_uoffset : get_rand_offset(dst_uoffset);\n+\t\tsrc_addr[i] = (is_src_cached) ? src_uoffset : get_rand_offset(src_uoffset);\n \t}\n }\n \n@@ -169,16 +170,17 @@ fill_addr_arrays(size_t *dst_addr, int is_dst_cached,\n */\n static void\n do_uncached_write(uint8_t *dst, int is_dst_cached,\n-\t\tconst uint8_t *src, int is_src_cached, size_t size)\n+\t\t\t\t const uint8_t *src, int is_src_cached, size_t size)\n {\n \tunsigned i, j;\n \tsize_t dst_addrs[TEST_BATCH_SIZE], src_addrs[TEST_BATCH_SIZE];\n \n \tfor (i = 0; i < (TEST_ITERATIONS / TEST_BATCH_SIZE); i++) {\n-\t\tfill_addr_arrays(dst_addrs, is_dst_cached,\n-\t\t\t src_addrs, is_src_cached);\n-\t\tfor (j = 0; j < TEST_BATCH_SIZE; j++)\n+\t\tfill_addr_arrays(dst_addrs, is_dst_cached, 0,\n+\t\t\t\t\t\t src_addrs, is_src_cached, 0);\n+\t\tfor (j = 0; j < TEST_BATCH_SIZE; j++) {\n \t\t\trte_memcpy(dst+dst_addrs[j], src+src_addrs[j], size);\n+\t\t}\n \t}\n }\n \n@@ -186,52 +188,129 @@ do_uncached_write(uint8_t *dst, int is_dst_cached,\n * Run a single memcpy performance test. This is a macro to ensure that if\n * the \"size\" parameter is a constant it won't be converted to a variable.\n */\n-#define SINGLE_PERF_TEST(dst, is_dst_cached, src, is_src_cached, size) do { \\\n-\tunsigned int iter, t; \\\n-\tsize_t dst_addrs[TEST_BATCH_SIZE], src_addrs[TEST_BATCH_SIZE]; \\\n-\tuint64_t start_time, total_time = 0; \\\n-\tuint64_t total_time2 = 0; \\\n-\tfor (iter = 0; iter < (TEST_ITERATIONS / TEST_BATCH_SIZE); iter++) { \\\n-\t\tfill_addr_arrays(dst_addrs, is_dst_cached, \\\n-\t\t src_addrs, is_src_cached); \\\n-\t\tstart_time = rte_rdtsc(); \\\n-\t\tfor (t = 0; t < TEST_BATCH_SIZE; t++) \\\n-\t\t\trte_memcpy(dst+dst_addrs[t], src+src_addrs[t], size); \\\n-\t\ttotal_time += rte_rdtsc() - start_time; \\\n-\t} \\\n-\tfor (iter = 0; iter < (TEST_ITERATIONS / TEST_BATCH_SIZE); iter++) { \\\n-\t\tfill_addr_arrays(dst_addrs, is_dst_cached, \\\n-\t\t src_addrs, is_src_cached); \\\n-\t\tstart_time = rte_rdtsc(); \\\n-\t\tfor (t = 0; t < TEST_BATCH_SIZE; t++) \\\n-\t\t\tmemcpy(dst+dst_addrs[t], src+src_addrs[t], size); \\\n-\t\ttotal_time2 += rte_rdtsc() - start_time; \\\n-\t} \\\n-\tprintf(\"%8.0f -\", (double)total_time /TEST_ITERATIONS); \\\n-\tprintf(\"%5.0f\", (double)total_time2 / TEST_ITERATIONS); \\\n+#define SINGLE_PERF_TEST(dst, is_dst_cached, dst_uoffset, \\\n+\t\t\t\t\t\t src, is_src_cached, src_uoffset, size) \\\n+do { \\\n+\tunsigned int iter, t; \\\n+\tsize_t dst_addrs[TEST_BATCH_SIZE], src_addrs[TEST_BATCH_SIZE]; \\\n+\tuint64_t start_time, total_time = 0; \\\n+\tuint64_t total_time2 = 0; \\\n+\tfor (iter = 0; iter < (TEST_ITERATIONS / TEST_BATCH_SIZE); iter++) { \\\n+\t\tfill_addr_arrays(dst_addrs, is_dst_cached, dst_uoffset, \\\n+\t\t\t\t\t\t src_addrs, is_src_cached, src_uoffset); \\\n+\t\tstart_time = rte_rdtsc(); \\\n+\t\tfor (t = 0; t < TEST_BATCH_SIZE; t++) \\\n+\t\t\trte_memcpy(dst+dst_addrs[t], src+src_addrs[t], size); \\\n+\t\ttotal_time += rte_rdtsc() - start_time; \\\n+\t} \\\n+\tfor (iter = 0; iter < (TEST_ITERATIONS / TEST_BATCH_SIZE); iter++) { \\\n+\t\tfill_addr_arrays(dst_addrs, is_dst_cached, dst_uoffset, \\\n+\t\t\t\t\t\t src_addrs, is_src_cached, src_uoffset); \\\n+\t\tstart_time = rte_rdtsc(); \\\n+\t\tfor (t = 0; t < TEST_BATCH_SIZE; t++) \\\n+\t\t\tmemcpy(dst+dst_addrs[t], src+src_addrs[t], size); \\\n+\t\ttotal_time2 += rte_rdtsc() - start_time; \\\n+\t} \\\n+\tprintf(\"%8.0f -\", (double)total_time /TEST_ITERATIONS); \\\n+\tprintf(\"%5.0f\", (double)total_time2 / TEST_ITERATIONS); \\\n } while (0)\n \n-/* Run memcpy() tests for each cached/uncached permutation. */\n-#define ALL_PERF_TESTS_FOR_SIZE(n) do { \\\n-\tif (__builtin_constant_p(n)) \\\n-\t\tprintf(\"\\nC%6u\", (unsigned)n); \\\n-\telse \\\n-\t\tprintf(\"\\n%7u\", (unsigned)n); \\\n-\tSINGLE_PERF_TEST(small_buf_write, 1, small_buf_read, 1, n); \\\n-\tSINGLE_PERF_TEST(large_buf_write, 0, small_buf_read, 1, n); \\\n-\tSINGLE_PERF_TEST(small_buf_write, 1, large_buf_read, 0, n); \\\n-\tSINGLE_PERF_TEST(large_buf_write, 0, large_buf_read, 0, n); \\\n+/* Run aligned memcpy tests for each cached/uncached permutation */\n+#define ALL_PERF_TESTS_FOR_SIZE(n) \\\n+do { \\\n+\tif (__builtin_constant_p(n)) \\\n+\t\tprintf(\"\\nC%6u\", (unsigned)n); \\\n+\telse \\\n+\t\tprintf(\"\\n%7u\", (unsigned)n); \\\n+\tSINGLE_PERF_TEST(small_buf_write, 1, 0, small_buf_read, 1, 0, n); \\\n+\tSINGLE_PERF_TEST(large_buf_write, 0, 0, small_buf_read, 1, 0, n); \\\n+\tSINGLE_PERF_TEST(small_buf_write, 1, 0, large_buf_read, 0, 0, n); \\\n+\tSINGLE_PERF_TEST(large_buf_write, 0, 0, large_buf_read, 0, 0, n); \\\n } while (0)\n \n-/*\n- * Run performance tests for a number of different sizes and cached/uncached\n- * permutations.\n- */\n+/* Run unaligned memcpy tests for each cached/uncached permutation */\n+#define ALL_PERF_TESTS_FOR_SIZE_UNALIGNED(n) \\\n+do { \\\n+\tif (__builtin_constant_p(n)) \\\n+\t\tprintf(\"\\nC%6u\", (unsigned)n); \\\n+\telse \\\n+\t\tprintf(\"\\n%7u\", (unsigned)n); \\\n+\tSINGLE_PERF_TEST(small_buf_write, 1, 1, small_buf_read, 1, 5, n); \\\n+\tSINGLE_PERF_TEST(large_buf_write, 0, 1, small_buf_read, 1, 5, n); \\\n+\tSINGLE_PERF_TEST(small_buf_write, 1, 1, large_buf_read, 0, 5, n); \\\n+\tSINGLE_PERF_TEST(large_buf_write, 0, 1, large_buf_read, 0, 5, n); \\\n+} while (0)\n+\n+/* Run memcpy tests for constant length */\n+#define ALL_PERF_TEST_FOR_CONSTANT \\\n+{ \\\n+\tTEST_CONSTANT(1U); TEST_CONSTANT(2U); TEST_CONSTANT(3U); \\\n+\tTEST_CONSTANT(4U); TEST_CONSTANT(5U); TEST_CONSTANT(6U); \\\n+\tTEST_CONSTANT(7U); TEST_CONSTANT(8U); TEST_CONSTANT(9U); \\\n+\tTEST_CONSTANT(12U); TEST_CONSTANT(15U); TEST_CONSTANT(16U); \\\n+\tTEST_CONSTANT(17U); TEST_CONSTANT(31U); TEST_CONSTANT(32U); \\\n+\tTEST_CONSTANT(33U); TEST_CONSTANT(63U); TEST_CONSTANT(64U); \\\n+\tTEST_CONSTANT(65U); TEST_CONSTANT(127U); TEST_CONSTANT(128U); \\\n+\tTEST_CONSTANT(129U); TEST_CONSTANT(191U); TEST_CONSTANT(192U); \\\n+\tTEST_CONSTANT(193U); TEST_CONSTANT(255U); TEST_CONSTANT(256U); \\\n+\tTEST_CONSTANT(257U); TEST_CONSTANT(319U); TEST_CONSTANT(320U); \\\n+\tTEST_CONSTANT(321U); TEST_CONSTANT(383U); TEST_CONSTANT(384U); \\\n+\tTEST_CONSTANT(385U); TEST_CONSTANT(447U); TEST_CONSTANT(448U); \\\n+\tTEST_CONSTANT(449U); TEST_CONSTANT(511U); TEST_CONSTANT(512U); \\\n+\tTEST_CONSTANT(513U); TEST_CONSTANT(767U); TEST_CONSTANT(768U); \\\n+\tTEST_CONSTANT(769U); TEST_CONSTANT(1023U); TEST_CONSTANT(1024U); \\\n+\tTEST_CONSTANT(1025U); TEST_CONSTANT(1518U); TEST_CONSTANT(1522U); \\\n+\tTEST_CONSTANT(1536U); TEST_CONSTANT(1600U); TEST_CONSTANT(2048U); \\\n+\tTEST_CONSTANT(2560U); TEST_CONSTANT(3072U); TEST_CONSTANT(3584U); \\\n+\tTEST_CONSTANT(4096U); TEST_CONSTANT(4608U); TEST_CONSTANT(5120U); \\\n+\tTEST_CONSTANT(5632U); TEST_CONSTANT(6144U); TEST_CONSTANT(6656U); \\\n+\tTEST_CONSTANT(7168U); TEST_CONSTANT(7680U); TEST_CONSTANT(8192U); \\\n+}\n+\n+/* Run all memcpy tests for aligned constant cases */\n+static inline void\n+perf_test_constant_aligned(void)\n+{\n+#define TEST_CONSTANT ALL_PERF_TESTS_FOR_SIZE\n+\tALL_PERF_TEST_FOR_CONSTANT;\n+#undef TEST_CONSTANT\n+}\n+\n+/* Run all memcpy tests for unaligned constant cases */\n+static inline void\n+perf_test_constant_unaligned(void)\n+{\n+#define TEST_CONSTANT ALL_PERF_TESTS_FOR_SIZE_UNALIGNED\n+\tALL_PERF_TEST_FOR_CONSTANT;\n+#undef TEST_CONSTANT\n+}\n+\n+/* Run all memcpy tests for aligned variable cases */\n+static inline void\n+perf_test_variable_aligned(void)\n+{\n+\tunsigned n = sizeof(buf_sizes) / sizeof(buf_sizes[0]);\n+\tunsigned i;\n+\tfor (i = 0; i < n; i++) {\n+\t\tALL_PERF_TESTS_FOR_SIZE((size_t)buf_sizes[i]);\n+\t}\n+}\n+\n+/* Run all memcpy tests for unaligned variable cases */\n+static inline void\n+perf_test_variable_unaligned(void)\n+{\n+\tunsigned n = sizeof(buf_sizes) / sizeof(buf_sizes[0]);\n+\tunsigned i;\n+\tfor (i = 0; i < n; i++) {\n+\t\tALL_PERF_TESTS_FOR_SIZE_UNALIGNED((size_t)buf_sizes[i]);\n+\t}\n+}\n+\n+/* Run all memcpy tests */\n static int\n perf_test(void)\n {\n-\tconst unsigned num_buf_sizes = sizeof(buf_sizes) / sizeof(buf_sizes[0]);\n-\tunsigned i;\n \tint ret;\n \n \tret = init_buffers();\n@@ -239,7 +318,8 @@ perf_test(void)\n \t\treturn ret;\n \n #if TEST_VALUE_RANGE != 0\n-\t/* Setup buf_sizes array, if required */\n+\t/* Set up buf_sizes array, if required */\n+\tunsigned i;\n \tfor (i = 0; i < TEST_VALUE_RANGE; i++)\n \t\tbuf_sizes[i] = i;\n #endif\n@@ -248,28 +328,23 @@ perf_test(void)\n \tdo_uncached_write(large_buf_write, 0, small_buf_read, 1, SMALL_BUFFER_SIZE);\n \n \tprintf(\"\\n** rte_memcpy() - memcpy perf. tests (C = compile-time constant) **\\n\"\n-\t \"======= ============== ============== ============== ==============\\n\"\n-\t \" Size Cache to cache Cache to mem Mem to cache Mem to mem\\n\"\n-\t \"(bytes) (ticks) (ticks) (ticks) (ticks)\\n\"\n-\t \"------- -------------- -------------- -------------- --------------\");\n-\n-\t/* Do tests where size is a variable */\n-\tfor (i = 0; i < num_buf_sizes; i++) {\n-\t\tALL_PERF_TESTS_FOR_SIZE((size_t)buf_sizes[i]);\n-\t}\n+\t\t \"======= ============== ============== ============== ==============\\n\"\n+\t\t \" Size Cache to cache Cache to mem Mem to cache Mem to mem\\n\"\n+\t\t \"(bytes) (ticks) (ticks) (ticks) (ticks)\\n\"\n+\t\t \"------- -------------- -------------- -------------- --------------\");\n+\n+\tprintf(\"\\n========================== %2dB aligned ============================\", ALIGNMENT_UNIT);\n+\t/* Do aligned tests where size is a variable */\n+\tperf_test_variable_aligned();\n \tprintf(\"\\n------- -------------- -------------- -------------- --------------\");\n-\t/* Do tests where size is a compile-time constant */\n-\tALL_PERF_TESTS_FOR_SIZE(63U);\n-\tALL_PERF_TESTS_FOR_SIZE(64U);\n-\tALL_PERF_TESTS_FOR_SIZE(65U);\n-\tALL_PERF_TESTS_FOR_SIZE(255U);\n-\tALL_PERF_TESTS_FOR_SIZE(256U);\n-\tALL_PERF_TESTS_FOR_SIZE(257U);\n-\tALL_PERF_TESTS_FOR_SIZE(1023U);\n-\tALL_PERF_TESTS_FOR_SIZE(1024U);\n-\tALL_PERF_TESTS_FOR_SIZE(1025U);\n-\tALL_PERF_TESTS_FOR_SIZE(1518U);\n-\n+\t/* Do aligned tests where size is a compile-time constant */\n+\tperf_test_constant_aligned();\n+\tprintf(\"\\n=========================== Unaligned =============================\");\n+\t/* Do unaligned tests where size is a variable */\n+\tperf_test_variable_unaligned();\n+\tprintf(\"\\n------- -------------- -------------- -------------- --------------\");\n+\t/* Do unaligned tests where size is a compile-time constant */\n+\tperf_test_constant_unaligned();\n \tprintf(\"\\n======= ============== ============== ============== ==============\\n\\n\");\n \n \tfree_buffers();\n@@ -277,7 +352,6 @@ perf_test(void)\n \treturn 0;\n }\n \n-\n static int\n test_memcpy_perf(void)\n {\n", "prefixes": [ "dpdk-dev", "3/4" ] }{ "id": 2340, "url": "