Received: by 2002:a25:8b91:0:0:0:0:0 with SMTP id j17csp555422ybl; Mon, 2 Dec 2019 15:00:13 -0800 (PST) X-Google-Smtp-Source: APXvYqxtoopjQsFHcuGFgVLm3C0Svu4T7R+EJz2QiLbt6OvHMqpteDF21AizUiVcyCOiS4ftCaNF X-Received: by 2002:a50:cb8d:: with SMTP id k13mr2025141edi.284.1575327613234; Mon, 02 Dec 2019 15:00:13 -0800 (PST) ARC-Seal: i=1; a=rsa-sha256; t=1575327613; cv=none; d=google.com; s=arc-20160816; b=x50ZCtcOSOTqDG+5ytRDzBDUzQWKcuBbfC6KMzzrFm5Bull2ZD4ZmcIqS6IG5bUNcy oxE+4XuyJ3Xag9ecJOfLm7K1UdOfqEM7+bJH03bWHPiBUWeGtFS8EFWUuwxQVbA1WtpJ 8378FznSAUHzJGmfoFpz2YxpYJNuIzwUtZq9Hgrq9cfeeSdUt2NPzrCLWWWJZtJGTI1N 2E8GiMcoF4V6GISc9zDbWIuq6r0DWOzIQY+pgrIfD9TNBVjPAFBQs4wAUYXF57qt1xK3 NjI+19itKHnCwlTgy59zFoDGOJksYMptfeTwEt55kYjnQcPIQQPQR3qbfH24HY1PuSbV Y2tA== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:dkim-signature:content-language :content-transfer-encoding:in-reply-to:mime-version:user-agent:date :message-id:from:references:cc:to:subject; bh=c/GyivbDczr8O8WHK5Rd8ruqQg/EBC+dBluKkb5R95w=; b=JxtqlucHMlgq/AU4sc99mdD7+YkUQuLL66EtRg8/fU2+klic1V9saIvwB31WTAOu8E njdNE4yD08C0YTegCmR96I6BrRW2AliqLibgoFCnhp5bB+SdXuOTvblxNsh/820QmPve uYWjXp7QJMB4rcENNFtUfPzVpR8XvPl0HYm56G81cBa+HgD+EmZcArs59LGFQaBYkfV1 P/MVgmKEwaFx+DcFc5Xws5octwmHYXa8gfc6dSWiS/nd53qOC9LaA2K250FKw84hrD8M OHoJrannES10BtAfalI4WIgx2eG+nMaulv+o3M+HMl+f1L6rjQ0vvwJyFZvWBiGbsC5p VtZg== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@nvidia.com header.s=n1 header.b=TjySz6FA; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=nvidia.com Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id dj5si847766edb.40.2019.12.02.14.59.49; Mon, 02 Dec 2019 15:00:13 -0800 (PST) Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67; Authentication-Results: mx.google.com; dkim=pass header.i=@nvidia.com header.s=n1 header.b=TjySz6FA; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=nvidia.com Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1726074AbfLBW6m (ORCPT + 99 others); Mon, 2 Dec 2019 17:58:42 -0500 Received: from hqemgate15.nvidia.com ([216.228.121.64]:16200 "EHLO hqemgate15.nvidia.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1725835AbfLBW6m (ORCPT ); Mon, 2 Dec 2019 17:58:42 -0500 Received: from hqpgpgate101.nvidia.com (Not Verified[216.228.121.13]) by hqemgate15.nvidia.com (using TLS: TLSv1.2, DES-CBC3-SHA) id ; Mon, 02 Dec 2019 14:58:30 -0800 Received: from hqmail.nvidia.com ([172.20.161.6]) by hqpgpgate101.nvidia.com (PGP Universal service); Mon, 02 Dec 2019 14:58:38 -0800 X-PGP-Universal: processed; by hqpgpgate101.nvidia.com on Mon, 02 Dec 2019 14:58:38 -0800 Received: from [10.2.160.125] (10.124.1.5) by HQMAIL107.nvidia.com (172.20.187.13) with Microsoft SMTP Server (TLS) id 15.0.1473.3; Mon, 2 Dec 2019 22:58:37 +0000 Subject: Re: [PATCH v2 02/11] soc: tegra: Add Tegra PMC clock registrations into PMC driver To: Dmitry Osipenko , , , , , , , , CC: , , , , , , , , , , , , , , , , , References: <1574830773-14892-1-git-send-email-skomatineni@nvidia.com> <1574830773-14892-3-git-send-email-skomatineni@nvidia.com> <749de44c-ec59-3cab-c02e-7b8fcb1fb9f4@gmail.com> <3d1492a1-f2a5-2d56-5341-a28fcb73fe64@nvidia.com> <484cb1bb-4fb2-9e71-87be-2bd5bd5b2348@gmail.com>

From: Sowjanya Komatineni Message-ID: Date: Mon, 2 Dec 2019 14:58:38 -0800 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.7.2 MIME-Version: 1.0 In-Reply-To: X-Originating-IP: [10.124.1.5] X-ClientProxiedBy: HQMAIL101.nvidia.com (172.20.187.10) To HQMAIL107.nvidia.com (172.20.187.13) Content-Type: text/plain; charset="utf-8"; format=flowed Content-Transfer-Encoding: quoted-printable Content-Language: en-US DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=nvidia.com; s=n1; t=1575327510; bh=c/GyivbDczr8O8WHK5Rd8ruqQg/EBC+dBluKkb5R95w=; h=X-PGP-Universal:Subject:To:CC:References:From:Message-ID:Date: User-Agent:MIME-Version:In-Reply-To:X-Originating-IP: X-ClientProxiedBy:Content-Type:Content-Transfer-Encoding: Content-Language; b=TjySz6FAQvYr+NmnTmYyT7AFBZ2SXqqBg4IeoaF73Jg3Eag6dCKvw8Ci9B+5PBdM9 FqgY3TiJfB2nm5iQlKoBtKH/fk44mBbjktXKNhum+2M/CQpKGndKNQAT0C8YkO/2Ms NBbLb5YDI91JvhmfgwE8dbXGG/jPGglssN4eofhHf6YLGCqS6kNPVaXFPmvigwHobe LEzqjkX4eXoP0cKd8zwvVwoLcIGFL8sVYREvHINTYsNTB4KmRQoeXfuxtI2xMeZdzP 5xfLhii/5wbRCKYsDoEBGI81OpERBKn3HG+uXUfHBYfm2FWFYAZOo3dDSaBX93sl6F B2jp0Bqxz9YRQ== Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 12/2/19 1:50 PM, Dmitry Osipenko wrote: > 02.12.2019 23:09, Sowjanya Komatineni =D0=BF=D0=B8=D1=88=D0=B5=D1=82: >> On 11/28/19 5:25 AM, Dmitry Osipenko wrote: >>> 28.11.2019 01:57, Sowjanya Komatineni =D0=BF=D0=B8=D1=88=D0=B5=D1=82: >>>> On 11/27/19 7:14 AM, Dmitry Osipenko wrote: >>>>> 27.11.2019 07:59, Sowjanya Komatineni =D0=BF=D0=B8=D1=88=D0=B5=D1=82: >>>>>> Tegra210 and prior Tegra PMC has clk_out_1, clk_out_2, clk_out_3 wit= h >>>>>> mux and gate for each of these clocks. >>>>>> >>>>>> Currently these PMC clocks are registered by Tegra clock driver usin= g >>>>>> clk_register_mux and clk_register_gate by passing PMC base address >>>>>> and register offsets and PMC programming for these clocks happens >>>>>> through direct PMC access by the clock driver. >>>>>> >>>>>> With this, when PMC is in secure mode any direct PMC access from the >>>>>> non-secure world does not go through and these clocks will not be >>>>>> functional. >>>>>> >>>>>> This patch adds these clocks registration with PMC as a clock provid= er >>>>>> for these clocks. clk_ops callback implementations for these clocks >>>>>> uses tegra_pmc_readl and tegra_pmc_writel which supports PMC >>>>>> programming >>>>>> in secure mode and non-secure mode. >>>>>> >>>>>> Signed-off-by: Sowjanya Komatineni >>>>>> --- >>>>>> =C2=A0=C2=A0 drivers/soc/tegra/pmc.c | 330 >>>>>> ++++++++++++++++++++++++++++++++++++++++++++++++ >>>>>> =C2=A0=C2=A0 1 file changed, 330 insertions(+) >>>>>> >>>>>> diff --git a/drivers/soc/tegra/pmc.c b/drivers/soc/tegra/pmc.c >>>>>> index ea0e11a09c12..a353f6d0a832 100644 >>>>>> --- a/drivers/soc/tegra/pmc.c >>>>>> +++ b/drivers/soc/tegra/pmc.c >>>>>> @@ -13,6 +13,9 @@ >>>>>> =C2=A0=C2=A0 =C2=A0 #include >>>>>> =C2=A0=C2=A0 #include >>>>>> +#include >>>>>> +#include >>>>>> +#include >>>>>> =C2=A0=C2=A0 #include >>>>>> =C2=A0=C2=A0 #include >>>>>> =C2=A0=C2=A0 #include >>>>>> @@ -48,6 +51,7 @@ >>>>>> =C2=A0=C2=A0 #include >>>>>> =C2=A0=C2=A0 #include >>>>>> =C2=A0=C2=A0 #include >>>>>> +#include >>>>>> =C2=A0=C2=A0 =C2=A0 #define PMC_CNTRL=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0x0 >>>>>> =C2=A0=C2=A0 #define=C2=A0 PMC_CNTRL_INTR_POLARITY=C2=A0=C2=A0=C2= =A0 BIT(17) /* inverts INTR >>>>>> polarity */ >>>>>> @@ -100,6 +104,7 @@ >>>>>> =C2=A0=C2=A0 #define PMC_WAKE2_STATUS=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0 0x168 >>>>>> =C2=A0=C2=A0 #define PMC_SW_WAKE2_STATUS=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0 0x16c >>>>>> =C2=A0=C2=A0 +#define PMC_CLK_OUT_CNTRL=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0 0x1a8 >>>>>> =C2=A0=C2=A0 #define PMC_SENSOR_CTRL=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0x1b0 >>>>>> =C2=A0=C2=A0 #define=C2=A0 PMC_SENSOR_CTRL_SCRATCH_WRITE=C2=A0=C2= =A0=C2=A0 BIT(2) >>>>>> =C2=A0=C2=A0 #define=C2=A0 PMC_SENSOR_CTRL_ENABLE_RST=C2=A0=C2=A0= =C2=A0 BIT(1) >>>>>> @@ -155,6 +160,91 @@ >>>>>> =C2=A0=C2=A0 #define=C2=A0 TEGRA_SMC_PMC_READ=C2=A0=C2=A0=C2=A0 0xa= a >>>>>> =C2=A0=C2=A0 #define=C2=A0 TEGRA_SMC_PMC_WRITE=C2=A0=C2=A0=C2=A0 0x= bb >>>>>> =C2=A0=C2=A0 +struct pmc_clk_mux { >>>>>> +=C2=A0=C2=A0=C2=A0 struct clk_hw=C2=A0=C2=A0=C2=A0 hw; >>>>>> +=C2=A0=C2=A0=C2=A0 unsigned long=C2=A0=C2=A0=C2=A0 offs; >>>>>> +=C2=A0=C2=A0=C2=A0 u32=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 ma= sk; >>>>>> +=C2=A0=C2=A0=C2=A0 u32=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 sh= ift; >>>>>> +}; >>>>>> + >>>>>> +#define to_pmc_clk_mux(_hw) container_of(_hw, struct pmc_clk_mux, h= w) >>>>>> + >>>>>> +struct pmc_clk_gate { >>>>>> +=C2=A0=C2=A0=C2=A0 struct clk_hw=C2=A0=C2=A0=C2=A0 hw; >>>>>> +=C2=A0=C2=A0=C2=A0 unsigned long=C2=A0=C2=A0=C2=A0 offs; >>>>>> +=C2=A0=C2=A0=C2=A0 u32=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 sh= ift; >>>>>> +}; >>>>>> + >>>>>> +#define to_pmc_clk_gate(_hw) container_of(_hw, struct >>>>>> pmc_clk_gate, hw) >>>>>> + >>>>>> +struct pmc_clk_init_data { >>>>>> +=C2=A0=C2=A0=C2=A0 char *mux_name; >>>>>> +=C2=A0=C2=A0=C2=A0 char *gate_name; >>>>>> +=C2=A0=C2=A0=C2=A0 const char **parents; >>>>>> +=C2=A0=C2=A0=C2=A0 int num_parents; >>>>>> +=C2=A0=C2=A0=C2=A0 int mux_id; >>>>>> +=C2=A0=C2=A0=C2=A0 int gate_id; >>>>>> +=C2=A0=C2=A0=C2=A0 char *dev_name; >>>>>> +=C2=A0=C2=A0=C2=A0 u8 mux_shift; >>>>>> +=C2=A0=C2=A0=C2=A0 u8 gate_shift; >>>>>> +=C2=A0=C2=A0=C2=A0 u8 init_parent_index; >>>>>> +=C2=A0=C2=A0=C2=A0 int init_state; >>>>>> +}; >>>>>> + >>>>>> +static const char *clk_out1_parents[] =3D { "clk_m", "clk_m_div2", >>>>>> +=C2=A0=C2=A0=C2=A0 "clk_m_div4", "extern1", >>>>>> +}; >>>>>> + >>>>>> +static const char *clk_out2_parents[] =3D { "clk_m", "clk_m_div2", >>>>>> +=C2=A0=C2=A0=C2=A0 "clk_m_div4", "extern2", >>>>>> +}; >>>>>> + >>>>>> +static const char *clk_out3_parents[] =3D { "clk_m", "clk_m_div2", >>>>>> +=C2=A0=C2=A0=C2=A0 "clk_m_div4", "extern3", >>>>>> +}; >>>>>> + >>>>>> +static struct pmc_clk_init_data tegra_pmc_clks_data[] =3D { >>>>>> +=C2=A0=C2=A0=C2=A0 { >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .mux_name =3D "clk_out_1= _mux", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .gate_name =3D "clk_out_= 1", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .parents =3D clk_out1_pa= rents, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .num_parents =3D ARRAY_S= IZE(clk_out1_parents), >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .mux_id =3D TEGRA_PMC_CL= K_OUT_1_MUX, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .gate_id =3D TEGRA_PMC_C= LK_OUT_1, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .dev_name =3D "extern1", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .mux_shift =3D 6, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .gate_shift =3D 2, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .init_parent_index =3D 3= , >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .init_state =3D 1, >>>>>> +=C2=A0=C2=A0=C2=A0 }, >>>>>> +=C2=A0=C2=A0=C2=A0 { >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .mux_name =3D "clk_out_2= _mux", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .gate_name =3D "clk_out_= 2", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .parents =3D clk_out2_pa= rents, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .num_parents =3D ARRAY_S= IZE(clk_out2_parents), >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .mux_id =3D TEGRA_PMC_CL= K_OUT_2_MUX, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .gate_id =3D TEGRA_PMC_C= LK_OUT_2, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .dev_name =3D "extern2", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .mux_shift =3D 14, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .gate_shift =3D 10, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .init_parent_index =3D 0= , >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .init_state =3D 0, >>>>>> +=C2=A0=C2=A0=C2=A0 }, >>>>>> +=C2=A0=C2=A0=C2=A0 { >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .mux_name =3D "clk_out_3= _mux", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .gate_name =3D "clk_out_= 3", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .parents =3D clk_out3_pa= rents, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .num_parents =3D ARRAY_S= IZE(clk_out3_parents), >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .mux_id =3D TEGRA_PMC_CL= K_OUT_3_MUX, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .gate_id =3D TEGRA_PMC_C= LK_OUT_3, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .dev_name =3D "extern3", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .mux_shift =3D 22, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .gate_shift =3D 18, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .init_parent_index =3D 0= , >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .init_state =3D 0, >>>>>> +=C2=A0=C2=A0=C2=A0 }, >>>>>> +}; >>>>>> + >>>>>> =C2=A0=C2=A0 struct tegra_powergate { >>>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 struct generic_pm_domain genpd= ; >>>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 struct tegra_pmc *pmc; >>>>>> @@ -254,6 +344,9 @@ struct tegra_pmc_soc { >>>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 */ >>>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 const struct tegra_wake_event = *wake_events; >>>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 unsigned int num_wake_events; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 struct pmc_clk_init_data *pmc_clks_data; >>>>>> +=C2=A0=C2=A0=C2=A0 unsigned int num_pmc_clks; >>>>>> =C2=A0=C2=A0 }; >>>>>> =C2=A0=C2=A0 =C2=A0 static const char * const tegra186_reset_source= s[] =3D { >>>>>> @@ -2163,6 +2256,228 @@ static int tegra_pmc_clk_notify_cb(struct >>>>>> notifier_block *nb, >>>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return NOTIFY_OK; >>>>>> =C2=A0=C2=A0 } >>>>>> =C2=A0=C2=A0 +static void pmc_clk_fence_udelay(u32 offset) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 tegra_pmc_readl(pmc, offset); >>>>>> +=C2=A0=C2=A0=C2=A0 /* pmc clk propagation delay 2 us */ >>>>>> +=C2=A0=C2=A0=C2=A0 udelay(2); >>>>>> +} >>>>>> + >>>>>> +static u8 pmc_clk_mux_get_parent(struct clk_hw *hw) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 struct pmc_clk_mux *mux =3D to_pmc_clk_mux(hw); >>>>>> +=C2=A0=C2=A0=C2=A0 int num_parents =3D clk_hw_get_num_parents(hw); >>>>>> +=C2=A0=C2=A0=C2=A0 u32 val; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 val =3D tegra_pmc_readl(pmc, mux->offs) >> mux->= shift; >>>>>> +=C2=A0=C2=A0=C2=A0 val &=3D mux->mask; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 if (val >=3D num_parents) >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return -EINVAL; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 return val; >>>>>> +} >>>>>> + >>>>>> +static int pmc_clk_mux_set_parent(struct clk_hw *hw, u8 index) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 struct pmc_clk_mux *mux =3D to_pmc_clk_mux(hw); >>>>>> +=C2=A0=C2=A0=C2=A0 u32 val; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 val =3D tegra_pmc_readl(pmc, mux->offs); >>>>>> +=C2=A0=C2=A0=C2=A0 val &=3D ~(mux->mask << mux->shift); >>>>>> +=C2=A0=C2=A0=C2=A0 val |=3D index << mux->shift; >>>>>> +=C2=A0=C2=A0=C2=A0 tegra_pmc_writel(pmc, val, mux->offs); >>>>>> +=C2=A0=C2=A0=C2=A0 pmc_clk_fence_udelay(mux->offs); >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 return 0; >>>>>> +} >>>>>> + >>>>>> +static const struct clk_ops pmc_clk_mux_ops =3D { >>>>>> +=C2=A0=C2=A0=C2=A0 .get_parent =3D pmc_clk_mux_get_parent, >>>>>> +=C2=A0=C2=A0=C2=A0 .set_parent =3D pmc_clk_mux_set_parent, >>>>>> +=C2=A0=C2=A0=C2=A0 .determine_rate =3D __clk_mux_determine_rate, >>>>>> +}; >>>>>> + >>>>>> +static struct clk * >>>>>> +tegra_pmc_clk_mux_register(const char *name, const char * const >>>>>> *parent_names, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0 int num_parents, unsigned long flags, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0 unsigned long offset, u32 shift, u32 mask) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 struct clk_init_data init; >>>>>> +=C2=A0=C2=A0=C2=A0 struct pmc_clk_mux *mux; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 mux =3D kzalloc(sizeof(*mux), GFP_KERNEL); >>>>>> +=C2=A0=C2=A0=C2=A0 if (!mux) >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return ERR_PTR(-ENOMEM); >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 init.name =3D name; >>>>>> +=C2=A0=C2=A0=C2=A0 init.ops =3D &pmc_clk_mux_ops; >>>>>> +=C2=A0=C2=A0=C2=A0 init.parent_names =3D parent_names; >>>>>> +=C2=A0=C2=A0=C2=A0 init.num_parents =3D num_parents; >>>>>> +=C2=A0=C2=A0=C2=A0 init.flags =3D flags; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 mux->hw.init =3D &init; >>>>>> +=C2=A0=C2=A0=C2=A0 mux->offs =3D offset; >>>>>> +=C2=A0=C2=A0=C2=A0 mux->mask =3D mask; >>>>>> +=C2=A0=C2=A0=C2=A0 mux->shift =3D shift; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 return clk_register(NULL, &mux->hw); >>>>>> +} >>>>>> + >>>>>> +static int pmc_clk_is_enabled(struct clk_hw *hw) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 struct pmc_clk_gate *gate =3D to_pmc_clk_gate(hw= ); >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 return tegra_pmc_readl(pmc, gate->offs) & BIT(ga= te->shift) ? 1 >>>>>> : 0; >>>>>> +} >>>>>> + >>>>>> +static void pmc_clk_set_state(struct clk_hw *hw, int state) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 struct pmc_clk_gate *gate =3D to_pmc_clk_gate(hw= ); >>>>>> +=C2=A0=C2=A0=C2=A0 u32 val; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 val =3D tegra_pmc_readl(pmc, gate->offs); >>>>>> +=C2=A0=C2=A0=C2=A0 val =3D state ? (val | BIT(gate->shift)) : (val = & >>>>>> ~BIT(gate->shift)); >>>>>> +=C2=A0=C2=A0=C2=A0 tegra_pmc_writel(pmc, val, gate->offs); >>>>>> +=C2=A0=C2=A0=C2=A0 pmc_clk_fence_udelay(gate->offs); >>>>>> +} >>>>>> + >>>>>> +static int pmc_clk_enable(struct clk_hw *hw) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 pmc_clk_set_state(hw, 1); >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 return 0; >>>>>> +} >>>>>> + >>>>>> +static void pmc_clk_disable(struct clk_hw *hw) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 pmc_clk_set_state(hw, 0); >>>>>> +} >>>>>> + >>>>>> +static const struct clk_ops pmc_clk_gate_ops =3D { >>>>>> +=C2=A0=C2=A0=C2=A0 .is_enabled =3D pmc_clk_is_enabled, >>>>>> +=C2=A0=C2=A0=C2=A0 .enable =3D pmc_clk_enable, >>>>>> +=C2=A0=C2=A0=C2=A0 .disable =3D pmc_clk_disable, >>>>>> +}; >>>>>> + >>>>>> +static struct clk * >>>>>> +tegra_pmc_clk_gate_register(const char *name, const char >>>>>> *parent_name, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0 unsigned long flags, unsigned long offset, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0 u32 shift) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 struct clk_init_data init; >>>>>> +=C2=A0=C2=A0=C2=A0 struct pmc_clk_gate *gate; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 gate =3D kzalloc(sizeof(*gate), GFP_KERNEL); >>>>>> +=C2=A0=C2=A0=C2=A0 if (!gate) >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return ERR_PTR(-ENOMEM); >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 init.name =3D name; >>>>>> +=C2=A0=C2=A0=C2=A0 init.ops =3D &pmc_clk_gate_ops; >>>>>> +=C2=A0=C2=A0=C2=A0 init.parent_names =3D &parent_name; >>>>>> +=C2=A0=C2=A0=C2=A0 init.num_parents =3D 1; >>>>>> +=C2=A0=C2=A0=C2=A0 init.flags =3D flags; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 gate->hw.init =3D &init; >>>>>> +=C2=A0=C2=A0=C2=A0 gate->offs =3D offset; >>>>>> +=C2=A0=C2=A0=C2=A0 gate->shift =3D shift; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 return clk_register(NULL, &gate->hw); >>>>>> +} >>>>>> + >>>>>> +static void tegra_pmc_clock_register(struct tegra_pmc *pmc, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 struct device_node *= np) >>>>>> +{ >>>>>> +=C2=A0=C2=A0=C2=A0 struct clk *clkmux, *clk, *parent; >>>>>> +=C2=A0=C2=A0=C2=A0 struct clk_onecell_data *clk_data; >>>>>> +=C2=A0=C2=A0=C2=A0 unsigned int num_clks; >>>>>> +=C2=A0=C2=A0=C2=A0 int i, ret; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 /* each pmc clock output has a mux and a gate */ >>>>>> +=C2=A0=C2=A0=C2=A0 num_clks =3D pmc->soc->num_pmc_clks * 2; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 if (!num_clks) >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 clk_data =3D kmalloc(sizeof(*clk_data), GFP_KERN= EL); >>>>>> +=C2=A0=C2=A0=C2=A0 if (!clk_data) >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 clk_data->clks =3D kcalloc(TEGRA_PMC_CLK_MAX, >>>>>> sizeof(*clk_data->clks), >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 GFP_KERNEL); >>>>>> +=C2=A0=C2=A0=C2=A0 if (!clk_data->clks) >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 goto free_clkdata; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 clk_data->clk_num =3D num_clks; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0 for (i =3D 0; i < pmc->soc->num_pmc_clks; i++) { >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 struct pmc_clk_init_data= *data; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 data =3D pmc->soc->pmc_c= lks_data + i; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 clkmux =3D tegra_pmc_clk= _mux_register(data->mux_name, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 data->parents, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 data->num_parents, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 CLK_SET_RATE_NO_REPARENT | >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 CLK_SET_RATE_PARENT, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 PMC_CLK_OUT_CNTRL, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 data->mux_shift, 3); >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 if (IS_ERR(clkmux)) >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 = goto free_clks; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 clk_data->clks[data->mux= _id] =3D clkmux; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 clk =3D tegra_pmc_clk_ga= te_register(data->gate_name, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0 data->mux_name, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0 CLK_SET_RATE_PARENT, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0 PMC_CLK_OUT_CNTRL, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0 data->gate_shift); >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 if (IS_ERR(clk)) >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 = goto free_clks; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 clk_data->clks[data->gat= e_id] =3D clk; >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 ret =3D clk_set_parent(c= lk, clkmux); >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 if (ret < 0) { >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 = pr_err("failed to set parent of %s to %s\n", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 __func__, __clk_get_name(clk), >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 __clk_get_name(clkmux)); >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 } >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 clk_register_clkdev(clk,= data->dev_name, data->gate_name); >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 /* configure initial clo= ck parent and state */ >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 parent =3D clk_get_sys(d= ata->gate_name, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 data->parents[data->= init_parent_index]); > Couldn't the default parent be defined using "assigned clock" in a > device-tree? Please see "Assigned clock parents and rates" in the doc. > > https://www.kernel.org/doc/Documentation/devicetree/bindings/clock/clock-= bindings.txt > > Then you could simply use of_clk_set_defaults(pmc->dev->of_node, true). Yes, of_clk_add_provider() does of_clk_set_defaults which sets based on=20 assigned parents and clock rates. This need device tree to specify assigned clock parent properties. Will=20 update device tree and remove init parent from the driver. > >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 if (!IS_ERR(parent)) { >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 = ret =3D clk_set_parent(clkmux, parent); >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 = if (ret < 0) { >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0 pr_err("failed to set parent of %s to %s\n", >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 __func__= , __clk_get_name(clkmux), >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 __clk_ge= t_name(parent)); >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0 WARN_ON(1); >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 = } >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 } >>>>>> + >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 if (data->init_state) { >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 = if (clk_prepare_enable(clk)) { >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0 pr_err("failed to enable %s\n", __func__, >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 __clk_ge= t_name(clk)); >>>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0 WARN_ON(1); >>> Alternatively you could write it like this: >>> >>> =C2=A0=C2=A0=C2=A0=C2=A0err =3D clk_prepare_enable(clk); >>> >>> =C2=A0=C2=A0=C2=A0=C2=A0WARN_ON(err, "failed to enable %s: %d\n", >>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 __clk_get_name(clk), err); >>> >>>>> Should be a bit better to move the WARN_ON to the end of errors >>>>> handling >>>>> in order to catch all possible errors: >>>>> >>>>> @@ -2510,6 +2510,7 @@ static void tegra_pmc_clock_register(struct >>>>> tegra_pmc *pmc, >>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return; >>>>> >>>>> =C2=A0=C2=A0 free_clks: >>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 WARN_ON(1); >>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 kfree(clk_dat= a->clks); >>>>> =C2=A0=C2=A0 free_clkdata: >>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 kfree(clk_dat= a); >>>> Reason I had WARN_ON right during clk_set_parent failure is to have th= e >>>> loop continue for subsequence pmc clocks registration instead of >>>> terminating all pmc clocks registration. >>> Ah, okay. Nevertheless this WARN_ON in the end shouldn't be the least >>> (IMO). >> Hi Dmitry, Just want to be clear on the above comment. Are you >> suggesting to add additional WARN_ON at the end? > Yes, it was my suggestion. > >> Thought WARN_ON right during corresponding clock failure with warn >> message showing clock names will be clear and also other clocks still >> should be registered. >> >> To add additional WARN_ON at the end need to track status of each clock >> and use that to as warn condition. > You could add a warning/error message to every point of failure. > > Primarily, it is important not to miss a error. Secondarily, it is > important to make diagnostic message meaningful. > > Realistically, I doubt that this chunk of code will ever fail once it is > known to work well. So it will be nice to have a more detailed > diagnostics (just in a case), but it shouldn't be a must. OK, Will add additional WARN message "failed registering PMC clocks" at=20 the end.