From nobody Tue Nov 26 14:22:08 2024 Received: from mgamail.intel.com (mgamail.intel.com [192.198.163.7]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 930B7188713; Fri, 18 Oct 2024 02:22:46 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=192.198.163.7 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1729218168; cv=none; b=ikgFeelbP3NEsTLQZywAnxYa0cxfRHXcXuq5PLsMuhVC6J6MWQXqu9DJDDwISZjhB3lgcDyH8iLqRfj7uTCP9p41w7GSczrBv1nU2DETJQgRYPeRAxXW7NVylf6j0XQafV4bzD/zs5lJwzOk2Sa+4UrwEYns7vJY88DcOm9Cw4Y= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1729218168; c=relaxed/simple; bh=vSW/xY/u5gQfdaq8Fb5zDhVy9K06adJcO87Sfqnguxo=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=hs7PBlF8hW6aFPgqA1QeSsaSAgo3i/8vLbtt7x8zMrv2AWZUjIvEEcFz4vLw8xQkLdTZ5pmZkpln+laIQdvBybUbt6dW+5WVysvfi1Q/K+Yq6F7qPZ6nLQdM7iknQ2fr/ive7zD02ZookfFweNeEWZKbRVBW8q7HN7tJ4EjDAWg= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=intel.com; spf=pass smtp.mailfrom=intel.com; dkim=pass (2048-bit key) header.d=intel.com header.i=@intel.com header.b=g/CshZ4c; arc=none smtp.client-ip=192.198.163.7 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=intel.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=intel.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=intel.com header.i=@intel.com header.b="g/CshZ4c" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1729218167; x=1760754167; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=vSW/xY/u5gQfdaq8Fb5zDhVy9K06adJcO87Sfqnguxo=; b=g/CshZ4cwTEiw8Yg6zaoWUhizohWTbcUIAtTMzH3KJdQ2koaE4DLgsez m1k6zOjET1vzqk/qtnhAGx/aXTHL9iHru698I1TV9Cu60J75I9Kk5iaeH xPxPVQTk0a04nHIGF5P9X5/tMKy2+Fv+OZc4eTKhBQK+Wr0p40YxXShjo LZ9NNXqrMAqxUcXrfflwVdSiTx+acn97vH6xZ8Q3nCjcVjeZWVOe2JXk7 sxAML0oUw8aM9vDKArWWcJXoffkIadrRweoEI+QNLJlj3w7c3SWU3O8T6 ufMINBXwC8qG3jPn6a2/26Ymc5N+YY0MuhlT37FBUPt/EoajW5+bCigFY w==; X-CSE-ConnectionGUID: 7SUFslUgQ+WDRNjgtydD6A== X-CSE-MsgGUID: 0wOlB5wiRoKzEV7vs00uWw== X-IronPort-AV: E=McAfee;i="6700,10204,11228"; a="54149721" X-IronPort-AV: E=Sophos;i="6.11,212,1725346800"; d="scan'208";a="54149721" Received: from orviesa010.jf.intel.com ([10.64.159.150]) by fmvoesa101.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 17 Oct 2024 19:22:38 -0700 X-CSE-ConnectionGUID: MvDR+WAaTJKdlQAxE7Q/xw== X-CSE-MsgGUID: snqS7aexSLuXEefTbLjNCQ== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="6.11,212,1725346800"; d="scan'208";a="78697738" Received: from rchatre-desk1.jf.intel.com ([10.165.154.99]) by orviesa010-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 17 Oct 2024 19:22:37 -0700 From: Reinette Chatre To: fenghua.yu@intel.com, shuah@kernel.org, tony.luck@intel.com, peternewman@google.com, babu.moger@amd.com, ilpo.jarvinen@linux.intel.com Cc: maciej.wieczor-retman@intel.com, reinette.chatre@intel.com, linux-kselftest@vger.kernel.org, linux-kernel@vger.kernel.org Subject: [PATCH V3 13/15] selftests/resctrl: Do not compare performance counters and resctrl at low bandwidth Date: Thu, 17 Oct 2024 19:33:28 -0700 Message-ID: X-Mailer: git-send-email 2.46.2 In-Reply-To: References: Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable The MBA test incrementally throttles memory bandwidth, each time followed by a comparison between the memory bandwidth observed by the performance counters and resctrl respectively. While a comparison between performance counters and resctrl is generally appropriate, they do not have an identical view of memory bandwidth. For example RAS features or memory performance features that generate memory traffic may drive accesses that are counted differently by performance counters and MBM respectively, for instance generating "overhead" traffic which is not counted against any specific RMID. As a ratio, this different view of memory bandwidth becomes more apparent at low memory bandwidths. It is not practical to enable/disable the various features that may generate memory bandwidth to give performance counters and resctrl an identical view. Instead, do not compare performance counters and resctrl view of memory bandwidth when the memory bandwidth is low. Bandwidth throttling behaves differently across platforms so it is not appropriate to drop measurement data simply based on the throttling level. Instead, use a threshold of 750MiB that has been observed to support adequate comparison between performance counters and resctrl. Signed-off-by: Reinette Chatre Reviewed-by: Ilpo J=C3=A4rvinen --- Changes since V2: - Add Ilpo's Reviewed-by tag. Changes since V1: - Fix code alignment and spacing. - Modify flow to use "continue" instead of "break" now that earlier changes decreases throttling. - Expand comment of define to elaborate causes of discrepancy between performance counters and MBM. --- tools/testing/selftests/resctrl/mba_test.c | 7 +++++++ tools/testing/selftests/resctrl/resctrl.h | 10 ++++++++++ 2 files changed, 17 insertions(+) diff --git a/tools/testing/selftests/resctrl/mba_test.c b/tools/testing/sel= ftests/resctrl/mba_test.c index bf37f3555660..5b4f0aa7a3a4 100644 --- a/tools/testing/selftests/resctrl/mba_test.c +++ b/tools/testing/selftests/resctrl/mba_test.c @@ -98,6 +98,13 @@ static bool show_mba_info(unsigned long *bw_imc, unsigne= d long *bw_resc) =20 avg_bw_imc =3D sum_bw_imc / (NUM_OF_RUNS - 1); avg_bw_resc =3D sum_bw_resc / (NUM_OF_RUNS - 1); + if (avg_bw_imc < THROTTLE_THRESHOLD || avg_bw_resc < THROTTLE_THRESHOLD)= { + ksft_print_msg("Bandwidth below threshold (%d MiB). Dropping results fr= om MBA schemata %u.\n", + THROTTLE_THRESHOLD, + ALLOCATION_MIN + ALLOCATION_STEP * allocation); + continue; + } + avg_diff =3D (float)labs(avg_bw_resc - avg_bw_imc) / avg_bw_imc; avg_diff_per =3D (int)(avg_diff * 100); =20 diff --git a/tools/testing/selftests/resctrl/resctrl.h b/tools/testing/self= tests/resctrl/resctrl.h index a553fe975938..dab1953fc7a0 100644 --- a/tools/testing/selftests/resctrl/resctrl.h +++ b/tools/testing/selftests/resctrl/resctrl.h @@ -43,6 +43,16 @@ =20 #define MINIMUM_SPAN (250 * MB) =20 +/* + * Memory bandwidth (in MiB) below which the bandwidth comparisons + * between iMC and resctrl are considered unreliable. For example RAS + * features or memory performance features that generate memory traffic + * may drive accesses that are counted differently by performance counters + * and MBM respectively, for instance generating "overhead" traffic which + * is not counted against any specific RMID. + */ +#define THROTTLE_THRESHOLD 750 + /* * fill_buf_param: "fill_buf" benchmark parameters * @buf_size: Size (in bytes) of buffer used in benchmark. --=20 2.46.2