[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <57cf3a15-5d9b-7636-4c69-60742e8cfae6@samsung.com>
Date: Mon, 1 Jun 2020 08:28:09 +0200
From: Marek Szyprowski <m.szyprowski@...sung.com>
To: Krzysztof Kozlowski <krzk@...nel.org>
Cc: linux-kernel@...r.kernel.org, linux-pm@...r.kernel.org,
Mark Brown <broonie@...nel.org>,
Dmitry Osipenko <digetx@...il.com>,
Liam Girdwood <lgirdwood@...il.com>,
Lucas Stach <l.stach@...gutronix.de>,
Bartlomiej Zolnierkiewicz <b.zolnierkie@...sung.com>,
Viresh Kumar <viresh.kumar@...aro.org>, peron.clem@...il.com,
Nishanth Menon <nm@...com>, Stephen Boyd <sboyd@...nel.org>,
Vincent Guittot <vincent.guittot@...aro.org>,
Rafael Wysocki <rjw@...ysocki.net>,
linux-samsung-soc@...r.kernel.org,
Chanwoo Choi <cw00.choi@...sung.com>
Subject: Re: [PATCH 2/2] soc: samsung: Add simple voltage coupler for
Exynos5800
Hi Krzysztof,
On 29.05.2020 19:43, Krzysztof Kozlowski wrote:
> On Fri, May 29, 2020 at 02:49:40PM +0200, Marek Szyprowski wrote:
>> Add custom voltage regulator coupler for Exynos5800 SoCs, which require
>> coupling between "vdd_arm" and "vdd_int" regulators. This coupler ensures
>> that coupled regulators voltage balancing is done only when clients for
>> each regulator (cpufreq for "vdd_arm" and devfreq for "vdd_int") apply
>> their constraints, so the voltage values don't go beyond the
>> bootloader-selected operation point during the boot process. This also
>> ensures proper voltage balancing if any of those drivers is missing.
I've intentionally didn't add any new properties nor bindings, because I
assume that the generic regulator coupling bindings fully describe what
is needed for Exynos5800 SoCs. This driver 'fixes' only the issue (or
lets it call undefined behavior) in the regulator core, which is fatal
for Exynos5800.
Please note that this "attach" callback is called only for the
regulators, which are marked as coupled, so for all existing board it
might be even enough to simply return 0 without any check. IMHO for all
existing Exynos5422/5800 the name-based check is enough. I can change it
to a check for particular DT path peculiar for Exynos5422/5800 (the
check if the given regulator mathes the phandle specificied in the
exynos-bus node for "wcore" or A15's CPU).
>> Signed-off-by: Marek Szyprowski <m.szyprowski@...sung.com>
>> ---
>> arch/arm/mach-exynos/Kconfig | 1 +
>> drivers/soc/samsung/Kconfig | 3 +
>> drivers/soc/samsung/Makefile | 1 +
>> .../soc/samsung/exynos-regulator-coupler.c | 59 +++++++++++++++++++
>> 4 files changed, 64 insertions(+)
>> create mode 100644 drivers/soc/samsung/exynos-regulator-coupler.c
>>
>> diff --git a/arch/arm/mach-exynos/Kconfig b/arch/arm/mach-exynos/Kconfig
>> index 76838255b5fa..f185cd3d4c62 100644
>> --- a/arch/arm/mach-exynos/Kconfig
>> +++ b/arch/arm/mach-exynos/Kconfig
>> @@ -118,6 +118,7 @@ config SOC_EXYNOS5800
>> bool "Samsung EXYNOS5800"
>> default y
>> depends on SOC_EXYNOS5420
>> + select EXYNOS_REGULATOR_COUPLER
>>
>> config EXYNOS_MCPM
>> bool
>> diff --git a/drivers/soc/samsung/Kconfig b/drivers/soc/samsung/Kconfig
>> index c7a2003687c7..264185664594 100644
>> --- a/drivers/soc/samsung/Kconfig
>> +++ b/drivers/soc/samsung/Kconfig
>> @@ -37,4 +37,7 @@ config EXYNOS_PM_DOMAINS
>> bool "Exynos PM domains" if COMPILE_TEST
>> depends on PM_GENERIC_DOMAINS || COMPILE_TEST
>>
>> +config EXYNOS_REGULATOR_COUPLER
>> + bool "Exynos SoC Regulator Coupler" if COMPILE_TEST
>> + depends on ARCH_EXYNOS || COMPILE_TEST
>> endif
>> diff --git a/drivers/soc/samsung/Makefile b/drivers/soc/samsung/Makefile
>> index edd1d6ea064d..ecc3a32f6406 100644
>> --- a/drivers/soc/samsung/Makefile
>> +++ b/drivers/soc/samsung/Makefile
>> @@ -9,3 +9,4 @@ obj-$(CONFIG_EXYNOS_PMU) += exynos-pmu.o
>> obj-$(CONFIG_EXYNOS_PMU_ARM_DRIVERS) += exynos3250-pmu.o exynos4-pmu.o \
>> exynos5250-pmu.o exynos5420-pmu.o
>> obj-$(CONFIG_EXYNOS_PM_DOMAINS) += pm_domains.o
>> +obj-$(CONFIG_EXYNOS_REGULATOR_COUPLER) += exynos-regulator-coupler.o
>> diff --git a/drivers/soc/samsung/exynos-regulator-coupler.c b/drivers/soc/samsung/exynos-regulator-coupler.c
>> new file mode 100644
>> index 000000000000..54445918bd75
>> --- /dev/null
>> +++ b/drivers/soc/samsung/exynos-regulator-coupler.c
>> @@ -0,0 +1,59 @@
>> +// SPDX-License-Identifier: GPL-2.0
>> +/*
>> + * Copyright (c) 2020 Samsung Electronics Co., Ltd.
>> + * http://www.samsung.com/
>> + * Author: Marek Szyprowski <m.szyprowski@...sung.com>
>> + *
>> + * Samsung Exynos SoC voltage coupler
>> + */
>> +
>> +#include <linux/init.h>
>> +#include <linux/kernel.h>
>> +#include <linux/of.h>
>> +#include <linux/regulator/coupler.h>
>> +#include <linux/regulator/driver.h>
>> +
>> +static int exynos_coupler_balance_voltage(struct regulator_coupler *coupler,
>> + struct regulator_dev *rdev,
>> + suspend_state_t state)
>> +{
>> + struct coupling_desc *c_desc = &rdev->coupling_desc;
>> + int ret, cons_uV = 0, cons_max_uV = INT_MAX;
>> + bool skip_coupled = false;
>> +
>> + /* get coupled regulator constraints */
>> + ret = regulator_check_consumers(c_desc->coupled_rdevs[1], &cons_uV,
>> + &cons_max_uV, state);
>> + if (ret < 0)
>> + return ret;
>> +
>> + /* skip adjusting coupled regulator if it has no constraints set yet */
>> + if (cons_uV == 0)
>> + skip_coupled = true;
>> +
>> + return regulator_do_balance_voltage(rdev, state, skip_coupled);
>> +}
>> +
>> +static int exynos_coupler_attach(struct regulator_coupler *coupler,
>> + struct regulator_dev *rdev)
>> +{
>> + if (strcmp(rdev_get_name(rdev), "vdd_arm") == 0 ||
>> + strcmp(rdev_get_name(rdev), "vdd_int") == 0)
> Thanks for the patch!
>
> You hard-coded specific names of regulators existing
> only in DTS. They do not come from any driver nor bindings, therefore
> they could change making driver unusable. Someone could add a new DTS
> with different names and expect this to work as well.
>
> You need bindings for this. Something like:
> Documentation/devicetree/bindings/regulator/nvidia,tegra-regulators-coupling.txt
> But better in YAML, if possible.
>
> Best regards,
> Krzysztof
>
Best regards
--
Marek Szyprowski, PhD
Samsung R&D Institute Poland
Powered by blists - more mailing lists