Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752990AbdF0QYs (ORCPT ); Tue, 27 Jun 2017 12:24:48 -0400 Received: from mail-wr0-f171.google.com ([209.85.128.171]:33275 "EHLO mail-wr0-f171.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752123AbdF0QYn (ORCPT ); Tue, 27 Jun 2017 12:24:43 -0400 Subject: Re: [PATCH v8 3/3] mailbox: qcom: Add support for APCS clock controller To: Stephen Boyd Cc: jassisinghbrar@gmail.com, bjorn.andersson@linaro.org, robh+dt@kernel.org, mturquette@baylibre.com, linux-clk@vger.kernel.org, linux-kernel@vger.kernel.org, linux-arm-msm@vger.kernel.org, devicetree@vger.kernel.org References: <20170623161533.20449-1-georgi.djakov@linaro.org> <20170623161533.20449-4-georgi.djakov@linaro.org> <20170626224735.GQ4493@codeaurora.org> From: Georgi Djakov Message-ID: <0dc001d7-6ea4-4521-027e-5587a8188b1d@linaro.org> Date: Tue, 27 Jun 2017 19:24:40 +0300 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.1.1 MIME-Version: 1.0 In-Reply-To: <20170626224735.GQ4493@codeaurora.org> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3913 Lines: 149 Hi Stephen, On 06/27/2017 01:47 AM, Stephen Boyd wrote: > On 06/23, Georgi Djakov wrote: >> diff --git a/drivers/mailbox/qcom-apcs-ipc-mailbox.c b/drivers/mailbox/qcom-apcs-ipc-mailbox.c >> index 9924c6d7f05d..da363c6580da 100644 >> --- a/drivers/mailbox/qcom-apcs-ipc-mailbox.c >> +++ b/drivers/mailbox/qcom-apcs-ipc-mailbox.c >> @@ -11,6 +11,8 @@ >> * GNU General Public License for more details. >> */ >> >> +#include >> +#include >> #include >> #include >> #include >> @@ -19,6 +21,34 @@ >> #include >> #include >> #include >> +#include >> + >> +#include "../clk/qcom/clk-regmap.h" >> +#include "../clk/qcom/clk-regmap-mux-div.h" > > Why? > >> + >> + >> +static int msm8916_register_clk(struct device *dev, void __iomem *base) >> +{ >> + struct clk_regmap_mux_div *a53cc; >> + struct clk *pclk; >> + struct regmap *regmap; >> + struct clk_init_data init = { }; >> + int ret; >> + >> + a53cc = devm_kzalloc(dev, sizeof(*a53cc), GFP_KERNEL); >> + if (!a53cc) >> + return -ENOMEM; >> + >> + a53cc->reg_offset = 0x50; >> + a53cc->hid_width = 5; >> + a53cc->hid_shift = 0; >> + a53cc->src_width = 3; >> + a53cc->src_shift = 8; >> + a53cc->parent_map = gpll0_a53cc_map; >> + >> + init.name = "a53mux"; >> + init.parent_names = gpll0_a53cc; >> + init.num_parents = 2; > > ARRAY_SIZE(gpll0_a53cc) instead of 2 please Ok. > >> + init.ops = &clk_regmap_mux_div_ops; >> + init.flags = CLK_SET_RATE_PARENT; >> + a53cc->clkr.hw.init = &init; >> + >> + pclk = __clk_lookup(gpll0_a53cc[1]); > > Urgh.. ok. We can't clk_get()? Ok, will do. > >> + if (!pclk) >> + return -EPROBE_DEFER; >> + >> + a53cc->clk_nb.notifier_call = a53cc_notifier_cb; >> + ret = clk_notifier_register(pclk, &a53cc->clk_nb); >> + if (ret) { >> + dev_err(dev, "failed to register clock notifier: %d\n", ret); >> + return ret; >> + } >> + >> + regmap = devm_regmap_init_mmio(dev, base, &a53cc_regmap_config); >> + if (IS_ERR(regmap)) { >> + ret = PTR_ERR(regmap); >> + dev_err(dev, "failed to init regmap mmio: %d\n", ret); >> + goto err; >> + } >> + >> + a53cc->clkr.regmap = regmap; >> + >> + ret = devm_clk_register_regmap(dev, &a53cc->clkr); > > Regmap is not a requirement to work with the qcom clk driver. > The other option then is to drop the regmap-mux-div patch and switch to readl()/writel(). >> + if (ret) { >> + dev_err(dev, "failed to register regmap clock: %d\n", ret); >> + goto err; >> + } >> + >> + ret = of_clk_add_hw_provider(dev->of_node, of_clk_hw_simple_get, >> + &a53cc->clkr.hw); >> + if (ret) { >> + dev_err(dev, "failed to add clock provider: %d\n", ret); >> + goto err; >> + } >> + >> + return 0; >> + >> +err: >> + clk_notifier_unregister(pclk, &a53cc->clk_nb); >> + return ret; >> +} >> + >> static int qcom_apcs_ipc_probe(struct platform_device *pdev) >> { >> + struct device_node *np = pdev->dev.of_node; >> struct qcom_apcs_ipc *apcs; >> struct resource *res; >> unsigned long offset; >> @@ -63,6 +178,13 @@ static int qcom_apcs_ipc_probe(struct platform_device *pdev) >> if (IS_ERR(base)) >> return PTR_ERR(base); >> >> + if (of_device_is_compatible(np, "qcom,msm8916-apcs-kpss-global")) { >> + /* register the APCS mux and divider clock */ >> + ret = msm8916_register_clk(&pdev->dev, base); > > Register a child platform device here instead of creating clks in the > same driver? Ok, we can register a child platform device and create a separate driver in drivers/clk. If we are dropping regmap, then we can also move the mux-div clk ops and clk registration into the same file. I will give it a try. Thanks, Georgi > >> + if (ret) >> + return ret; >> + } >> + >> offset = (unsigned long)of_device_get_match_data(&pdev->dev); >> >> apcs->reg = base + offset; >