From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753436AbdCFNHR (ORCPT ); Mon, 6 Mar 2017 08:07:17 -0500 Received: from mail-wm0-f44.google.com ([74.125.82.44]:38773 "EHLO mail-wm0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752768AbdCFNG4 (ORCPT ); Mon, 6 Mar 2017 08:06:56 -0500 MIME-Version: 1.0 In-Reply-To: References: <1488538070-12549-1-git-send-email-sunil.kovvuri@gmail.com> <1488538070-12549-2-git-send-email-sunil.kovvuri@gmail.com> <20170303.095635.355325147613813698.davem@davemloft.net> From: Sunil Kovvuri Date: Mon, 6 Mar 2017 18:27:52 +0530 Message-ID: Subject: Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults To: Robin Murphy Cc: David Miller , Linux Netdev List , Sunil Goutham , LKML , LAKML Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org >> >> We are seeing a 0.75Mpps drop with IP forwarding rate due to that. >> Hence I have restricted calling DMA interfaces to only when IOMMU is enabled. > > What's 0.07Mpps as a percentage of baseline? On a correctly configured > coherent arm64 system, in the absence of an IOMMU, dma_map_*() is > essentially just virt_to_phys() behind a function call or two, so I'd be > interested to know where any non-trivial overhead might be coming from. It's a 5% drop and yes device is configured as coherent. And the drop is due to additional function calls. Thanks, Sunil. From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sunil Kovvuri Subject: Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults Date: Mon, 6 Mar 2017 18:27:52 +0530 Message-ID: References: <1488538070-12549-1-git-send-email-sunil.kovvuri@gmail.com> <1488538070-12549-2-git-send-email-sunil.kovvuri@gmail.com> <20170303.095635.355325147613813698.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: Linux Netdev List , Sunil Goutham , David Miller , LAKML , LKML To: Robin Murphy Return-path: In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=m.gmane.org@lists.infradead.org List-Id: netdev.vger.kernel.org >> >> We are seeing a 0.75Mpps drop with IP forwarding rate due to that. >> Hence I have restricted calling DMA interfaces to only when IOMMU is enabled. > > What's 0.07Mpps as a percentage of baseline? On a correctly configured > coherent arm64 system, in the absence of an IOMMU, dma_map_*() is > essentially just virt_to_phys() behind a function call or two, so I'd be > interested to know where any non-trivial overhead might be coming from. It's a 5% drop and yes device is configured as coherent. And the drop is due to additional function calls. Thanks, Sunil. From mboxrd@z Thu Jan 1 00:00:00 1970 From: sunil.kovvuri@gmail.com (Sunil Kovvuri) Date: Mon, 6 Mar 2017 18:27:52 +0530 Subject: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults In-Reply-To: References: <1488538070-12549-1-git-send-email-sunil.kovvuri@gmail.com> <1488538070-12549-2-git-send-email-sunil.kovvuri@gmail.com> <20170303.095635.355325147613813698.davem@davemloft.net> Message-ID: To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org >> >> We are seeing a 0.75Mpps drop with IP forwarding rate due to that. >> Hence I have restricted calling DMA interfaces to only when IOMMU is enabled. > > What's 0.07Mpps as a percentage of baseline? On a correctly configured > coherent arm64 system, in the absence of an IOMMU, dma_map_*() is > essentially just virt_to_phys() behind a function call or two, so I'd be > interested to know where any non-trivial overhead might be coming from. It's a 5% drop and yes device is configured as coherent. And the drop is due to additional function calls. Thanks, Sunil.