From mboxrd@z Thu Jan 1 00:00:00 1970 From: Michael Turquette Subject: Re: [PATCH V3 1/2] clk: imx6: Add SPDIF_GCLK clock in clock tree Date: Tue, 20 Oct 2015 03:23:54 -0700 Message-ID: <20151020102354.20687.97671@quantum> References: <20151019152807.20687.15779@quantum> <20151019153618.GA14709@tiger> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20151019153618.GA14709@tiger> 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 To: Shawn Guo Cc: mark.rutland@arm.com, devicetree@vger.kernel.org, Shengjiu Wang , linux@arm.linux.org.uk, pawel.moll@arm.com, ijc+devicetree@hellion.org.uk, linux-kernel@vger.kernel.org, robh+dt@kernel.org, kernel@pengutronix.de, galak@codeaurora.org, linux-arm-kernel@lists.infradead.org List-Id: devicetree@vger.kernel.org Hi Shawn, Quoting Shawn Guo (2015-10-19 08:36:18) > Hi Mike, > > On Mon, Oct 19, 2015 at 08:28:07AM -0700, Michael Turquette wrote: > > Quoting Shengjiu Wang (2015-10-10 03:15:06) > > > Correct SPDIF clock setting issue in clock tree, the SPDIF_GCLK is also > > > one clock of SPDIF, which is missed before. > > > > > > We found an issue that imx can't enter low power mode with spdif > > > if IMX6x_CLK_SPDIF is used as the core clock of spdif. Because > > > spdif driver will register IMX6x_CLK_SPDIF clock to regmap, regmap will do > > > clk_prepare in init function, then IMX6x_CLK_SPDIF clock is prepared in probe, > > > so its parent clock (PLL clock) is prepared, the prepare operation of > > > PLL clock is to enable the clock. But I.MX needs all PLL clock is disabled, > > > then it can enter low power mode. > > > > > > So we can't use IMX6x_CLK_SPDIF as the core clock of spdif, the correct spdif > > > core clock is SPDIF_GCLK, which share same gate bit with IMX6x_CLK_SPDIF clock. > > > SPDIF_GCLK's parent clock is ipg clock. > > > > I'm confused by this. Is there really a new clock signal to be added, or > > this just to workaround some reference counting problems with regmap? > > I was confused by the previous version of the patch, and asked Shengjiu > to improve the commit log, which seems still not so good. In short, the > patch does add a missing clock, and the missing of the clock is > discovered by a low-power-mode issue when SPDIF driver is enabled. Thank you for the explanation. Sounds good to me. Regards, Mike > > > > > > > > > Signed-off-by: Shengjiu Wang > > > > Please Cc the linux-clk@vger.kernel.org mailing list for future clock > > driver patches. > > Right, should have reminded him. I will keep my eyes more closely on > this. > > Shawn