From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dmitry Osipenko Subject: Re: [PATCH v3 09/13] dmaengine: tegra-apb: Remove runtime PM usage Date: Wed, 8 Jan 2020 18:10:30 +0300 Message-ID: References: <20200106011708.7463-1-digetx@gmail.com> <20200106011708.7463-10-digetx@gmail.com> <01660250-0489-870a-6f0e-d74c5041e8e3@nvidia.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Return-path: In-Reply-To: <01660250-0489-870a-6f0e-d74c5041e8e3@nvidia.com> Content-Language: en-US Sender: linux-kernel-owner@vger.kernel.org To: Jon Hunter , Laxman Dewangan , Vinod Koul , Dan Williams , Thierry Reding , =?UTF-8?B?TWljaGHFgiBNaXJvc8WCYXc=?= Cc: dmaengine@vger.kernel.org, linux-tegra@vger.kernel.org, linux-kernel@vger.kernel.org List-Id: linux-tegra@vger.kernel.org 07.01.2020 21:38, Jon Hunter пишет: > > On 07/01/2020 17:12, Dmitry Osipenko wrote: >> 07.01.2020 18:13, Jon Hunter пишет: >>> >>> On 06/01/2020 01:17, Dmitry Osipenko wrote: >>>> There is no benefit from runtime PM usage for the APB DMA driver because >>>> it enables clock at the time of channel's allocation and thus clock stays >>>> enabled all the time in practice, secondly there is benefit from manually >>>> disabled clock because hardware auto-gates it during idle by itself. >>> >>> This assumes that the channel is allocated during a driver >>> initialisation. That may not always be the case. I believe audio is one >>> case where channels are requested at the start of audio playback. >> >> At least serial, I2C, SPI and T20 FUSE are permanently keeping channels >> allocated, thus audio is an exception here. I don't think that it's >> practical to assume that there is a real-world use-case where audio >> driver is the only active DMA client. >> >> The benefits of gating the DMA clock are also dim, do you have any >> power-consumption numbers that show that it's really worth to care about >> the clock-gating? > > No, but at the same time, I really don't see the point in this. In fact, > I think it is a step backwards. If we wanted to only enable clocks while > DMA channels are active we could. So I request you drop this. I'll take a look at making RPM active only during the time of DMA activity, otherwise it's pretty much a dead code as it is now.