From patchwork Mon Oct 7 13:59:31 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Leon Romanovsky X-Patchwork-Id: 1172822 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=kernel.org Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=kernel.org header.i=@kernel.org header.b="Vk1CfCbb"; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46n2Cy2RTmz9sCJ for ; Tue, 8 Oct 2019 00:59:42 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728081AbfJGN7l (ORCPT ); Mon, 7 Oct 2019 09:59:41 -0400 Received: from mail.kernel.org ([198.145.29.99]:44024 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727536AbfJGN7k (ORCPT ); Mon, 7 Oct 2019 09:59:40 -0400 Received: from localhost (unknown [193.47.165.251]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPSA id E08D920867; Mon, 7 Oct 2019 13:59:39 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1570456780; bh=FUHoCM9U2W2uQk6HDXmkiydK6wbVQVxkxHd0kP8tvgI=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=Vk1CfCbbUvTS30TQWOdE5MH+qgFIWc5wK9z0+l+hxxkon6D6XV6PAMIVqGQf0Zz8X 2t34IgVC8oyAiDNB7u0WTFvrRE6dYiypA9v9nTuQycp72l/F0RWplW82z046yAzSlf Fe9qhz1z3ctskYWdUm2IB+2iHKxNAsDSaE/Tu9fk= From: Leon Romanovsky To: Doug Ledford , Jason Gunthorpe , Christoph Hellwig Cc: Leon Romanovsky , RDMA mailing list , Or Gerlitz , Yamin Friedman , Saeed Mahameed , linux-netdev Subject: [PATCH mlx5-next v2 1/3] net/mlx5: Expose optimal performance scatter entries capability Date: Mon, 7 Oct 2019 16:59:31 +0300 Message-Id: <20191007135933.12483-2-leon@kernel.org> X-Mailer: git-send-email 2.21.0 In-Reply-To: <20191007135933.12483-1-leon@kernel.org> References: <20191007135933.12483-1-leon@kernel.org> MIME-Version: 1.0 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org From: Yamin Friedman Expose maximum scatter entries per RDMA READ for optimal performance. Signed-off-by: Yamin Friedman Reviewed-by: Or Gerlitz Signed-off-by: Leon Romanovsky --- include/linux/mlx5/mlx5_ifc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.20.1 diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h index 138c50d5a353..c0bfb1d90dd2 100644 --- a/include/linux/mlx5/mlx5_ifc.h +++ b/include/linux/mlx5/mlx5_ifc.h @@ -1153,7 +1153,7 @@ struct mlx5_ifc_cmd_hca_cap_bits { u8 log_max_srq[0x5]; u8 reserved_at_b0[0x10]; - u8 reserved_at_c0[0x8]; + u8 max_sgl_for_optimized_performance[0x8]; u8 log_max_cq_sz[0x8]; u8 reserved_at_d0[0xb]; u8 log_max_cq[0x5]; From patchwork Mon Oct 7 13:59:32 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Leon Romanovsky X-Patchwork-Id: 1172823 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=kernel.org Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=kernel.org header.i=@kernel.org header.b="mYI/vnt1"; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46n2D20NpBz9sP7 for ; Tue, 8 Oct 2019 00:59:46 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728098AbfJGN7p (ORCPT ); Mon, 7 Oct 2019 09:59:45 -0400 Received: from mail.kernel.org ([198.145.29.99]:44058 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727536AbfJGN7o (ORCPT ); Mon, 7 Oct 2019 09:59:44 -0400 Received: from localhost (unknown [193.47.165.251]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPSA id 35E4C20684; Mon, 7 Oct 2019 13:59:43 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1570456783; bh=KMrUIiNDJqsRDgBEExBmxx1lKczn1G2HUGHDDpCH8mc=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=mYI/vnt190xL1xPfBQsAxrYy6oBs/j3652uW5YysPoZD6bR+RnpLLXbsCREFdZsnQ joUBamTP5MB/n93rlz8gChhh/70fu4ffXqn1+0WrQTOBPGyoC9hNpf8Cg1HVVW+/wr 6qxLgt4UxSgBgMP2kFcnsldxrROMQ+ggyDPN+Ue8= From: Leon Romanovsky To: Doug Ledford , Jason Gunthorpe , Christoph Hellwig Cc: Leon Romanovsky , RDMA mailing list , Or Gerlitz , Yamin Friedman , Saeed Mahameed , linux-netdev Subject: [PATCH rdma-next v2 2/3] RDMA/rw: Support threshold for registration vs scattering to local pages Date: Mon, 7 Oct 2019 16:59:32 +0300 Message-Id: <20191007135933.12483-3-leon@kernel.org> X-Mailer: git-send-email 2.21.0 In-Reply-To: <20191007135933.12483-1-leon@kernel.org> References: <20191007135933.12483-1-leon@kernel.org> MIME-Version: 1.0 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org From: Yamin Friedman If there are more scatter entries than the recommended limit provided by the ib device, UMR registration is used. This will provide optimal performance when performing large RDMA READs over devices that advertise the threshold capability. With ConnectX-5 running NVMeoF RDMA with FIO single QP 128KB writes: Without use of cap: 70Gb/sec With use of cap: 84Gb/sec Signed-off-by: Yamin Friedman Reviewed-by: Or Gerlitz Signed-off-by: Leon Romanovsky Reviewed-by: Christoph Hellwig --- drivers/infiniband/core/rw.c | 25 +++++++++++++++---------- include/rdma/ib_verbs.h | 2 ++ 2 files changed, 17 insertions(+), 10 deletions(-) -- 2.20.1 diff --git a/drivers/infiniband/core/rw.c b/drivers/infiniband/core/rw.c index 5337393d4dfe..c27a543b58ef 100644 --- a/drivers/infiniband/core/rw.c +++ b/drivers/infiniband/core/rw.c @@ -20,14 +20,16 @@ module_param_named(force_mr, rdma_rw_force_mr, bool, 0); MODULE_PARM_DESC(force_mr, "Force usage of MRs for RDMA READ/WRITE operations"); /* - * Check if the device might use memory registration. This is currently only - * true for iWarp devices. In the future we can hopefully fine tune this based - * on HCA driver input. + * Check if the device might use memory registration. This is currently + * true for iWarp devices and devices that have optimized SGL registration + * logic. */ static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num) { if (rdma_protocol_iwarp(dev, port_num)) return true; + if (dev->attrs.max_sgl_rd) + return true; if (unlikely(rdma_rw_force_mr)) return true; return false; @@ -35,17 +37,20 @@ static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num) /* * Check if the device will use memory registration for this RW operation. - * We currently always use memory registrations for iWarp RDMA READs, and - * have a debug option to force usage of MRs. - * - * XXX: In the future we can hopefully fine tune this based on HCA driver - * input. + * For RDMA READs we must use MRs on iWarp and can optionaly use them as an + * optimaztion otherwise. Additionally we have a debug option to force usage + * of MRs to help testing this code path. */ + static inline bool rdma_rw_io_needs_mr(struct ib_device *dev, u8 port_num, enum dma_data_direction dir, int dma_nents) { - if (rdma_protocol_iwarp(dev, port_num) && dir == DMA_FROM_DEVICE) - return true; + if (dir == DMA_FROM_DEVICE) { + if (rdma_protocol_iwarp(dev, port_num)) + return true; + if (dev->attrs.max_sgl_rd && dma_nents > dev->attrs.max_sgl_rd) + return true; + } if (unlikely(rdma_rw_force_mr)) return true; return false; diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 4f671378dbfc..60fd98a9b7e8 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -445,6 +445,8 @@ struct ib_device_attr { struct ib_tm_caps tm_caps; struct ib_cq_caps cq_caps; u64 max_dm_size; + /* Max entries for sgl for optimized performance per READ */ + u32 max_sgl_rd; }; enum ib_mtu { From patchwork Mon Oct 7 13:59:33 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Leon Romanovsky X-Patchwork-Id: 1172824 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=kernel.org Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=kernel.org header.i=@kernel.org header.b="rUd+0Qi6"; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46n2D44nztz9s4Y for ; Tue, 8 Oct 2019 00:59:48 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728126AbfJGN7s (ORCPT ); Mon, 7 Oct 2019 09:59:48 -0400 Received: from mail.kernel.org ([198.145.29.99]:44086 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727536AbfJGN7r (ORCPT ); Mon, 7 Oct 2019 09:59:47 -0400 Received: from localhost (unknown [193.47.165.251]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPSA id 72EBE20867; Mon, 7 Oct 2019 13:59:46 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1570456787; bh=174nUUEc39BnPNjnjG3coCMSkyWpv0morokgjS6xmCA=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=rUd+0Qi6Pt7+4Sre7hWSZNTcQy77wWwBSNlOiOI4HoilPh0ObZwWBCgmeB1VTUphO 5knyoyX/IM2BCrTXlK8P4eIa4nOnZA2tGZ9kLKOd1Gm/uGmoe5XuVuW2WuxyBVLtmX 4BoWx0epSWGQwIFxgKAlgKl7L8cCxenc54KKWuTQ= From: Leon Romanovsky To: Doug Ledford , Jason Gunthorpe , Christoph Hellwig Cc: Leon Romanovsky , RDMA mailing list , Or Gerlitz , Yamin Friedman , Saeed Mahameed , linux-netdev Subject: [PATCH rdma-next v2 3/3] RDMA/mlx5: Add capability for max sge to get optimized performance Date: Mon, 7 Oct 2019 16:59:33 +0300 Message-Id: <20191007135933.12483-4-leon@kernel.org> X-Mailer: git-send-email 2.21.0 In-Reply-To: <20191007135933.12483-1-leon@kernel.org> References: <20191007135933.12483-1-leon@kernel.org> MIME-Version: 1.0 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org From: Yamin Friedman Allows the IB device to provide a value of maximum scatter gather entries per RDMA READ. In certain cases it may be preferable for a device to perform UMR memory registration rather than have many scatter entries in a single RDMA READ. This provides a significant performance increase in devices capable of using different memory registration schemes based on the number of scatter gather entries. This general capability allows each device vendor to fine tune when it is better to use memory registration. Signed-off-by: Yamin Friedman Reviewed-by: Or Gerlitz Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/mlx5/main.c | 2 ++ 1 file changed, 2 insertions(+) -- 2.20.1 diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index fa23c8e7043b..39d54e285ae9 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -1012,6 +1012,8 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, 1 << MLX5_CAP_GEN(mdev, log_max_klm_list_size); props->max_pi_fast_reg_page_list_len = props->max_fast_reg_page_list_len / 2; + props->max_sgl_rd = + MLX5_CAP_GEN(mdev, max_sgl_for_optimized_performance); get_atomic_caps_qp(dev, props); props->masked_atomic_cap = IB_ATOMIC_NONE; props->max_mcast_grp = 1 << MLX5_CAP_GEN(mdev, log_max_mcg);