Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH 2/2] opp: Don't create an OPP table from
 dev_pm_opp_get_opp_table()
To:     Viresh Kumar <viresh.kumar@linaro.org>
Cc:     "Rafael J. Wysocki" <rjw@rjwysocki.net>,
        Kevin Hilman <khilman@kernel.org>,
        Ulf Hansson <ulf.hansson@linaro.org>,
        Len Brown <len.brown@intel.com>, Pavel Machek <pavel@ucw.cz>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Viresh Kumar <vireshk@kernel.org>, Nishanth Menon <nm@ti.com>,
        Stephen Boyd <sboyd@kernel.org>, linux-pm@vger.kernel.org,
        Vincent Guittot <vincent.guittot@linaro.org>,
        linux-kernel@vger.kernel.org
References: <684ff01900180c0a40ec307dacc673b24eab593b.1604643714.git.viresh.kumar@linaro.org>
 <1012a98950355bd5a52424668050a17c3430cbe0.1604643714.git.viresh.kumar@linaro.org>
 <ec9839dd-5d2d-0d6b-6563-b14da4af1a57@gmail.com>
 <20201109043457.xf55kufhjjz2fvct@vireshk-i7>
 <c5078503-ee14-0f84-85fd-9c6e55d2e897@gmail.com>
 <20201109045740.y5kpd6tjscoqxhi5@vireshk-i7>
From:   Dmitry Osipenko <digetx@gmail.com>
Message-ID: <7a0a904d-8ee2-d1d6-509c-950bb408e0c5@gmail.com>
Date:   Mon, 9 Nov 2020 08:29:50 +0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.10.0
MIME-Version: 1.0
In-Reply-To: <20201109045740.y5kpd6tjscoqxhi5@vireshk-i7>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk

09.11.2020 07:57, Viresh Kumar пишет:
> On 09-11-20, 07:41, Dmitry Osipenko wrote:
>> 09.11.2020 07:34, Viresh Kumar пишет:
>>> On 06-11-20, 16:18, Dmitry Osipenko wrote:
>>>> 06.11.2020 09:24, Viresh Kumar пишет:
>>>>> It has been found that some users (like cpufreq-dt and others on LKML)
>>>>> have abused the helper dev_pm_opp_get_opp_table() to create the OPP
>>>>> table instead of just finding it, which is the wrong thing to do. This
>>>>> routine was meant for OPP core's internal working and exposed the whole
>>>>> functionality by mistake.
>>>>>
>>>>> Change the scope of dev_pm_opp_get_opp_table() to only finding the
>>>>> table. The internal helpers _opp_get_opp_table*() are thus renamed to
>>>>> _add_opp_table*(), dev_pm_opp_get_opp_table_indexed() is removed (as we
>>>>> don't need the index field for finding the OPP table) and so the only
>>>>> user, genpd, is updated.
>>>>>
>>>>> Note that the prototype of _add_opp_table() was already left in opp.h by
>>>>> mistake when it was removed earlier and so we weren't required to add it
>>>>> now.
>>>>
>>>> Hello Viresh,
>>>>
>>>> It looks like this is not an entirely correct change because previously
>>>> it was possible to get an empty opp_table in order to use it for the
>>>> dev_pm_opp_set_rate(), which would fall back to clk_set_rate if table is
>>>> empty.
>>>>
>>>> Now it's not possible to get an empty table and
>>>> dev_pm_opp_of_add_table() would error out if OPPs are missing in a
>>>> device-tree. Hence it's not possible to implement a fall back without
>>>> abusing opp_set_regulators() or opp_set_supported_hw() for getting the
>>>> empty table. Or am I missing something?
>>>
>>> For that case you were always required to call
>>> dev_pm_opp_set_clkname(), otherwise how would the OPP core know which
>>> clock to set ? And the same shall work now as well.
>>
>> Why _allocate_opp_table() grabs the first default clk of a device and
>> assigns it to the created table?
> 
> Right, it was there so everybody isn't required to call
> dev_pm_opp_set_clkname() if they don't need to pass a connection id
> while getting the clock. But for the case of supporting empty OPP
> tables for cases where we just want dev_pm_opp_set_rate() to act as
> clk_set_rate() (in order for the drivers to avoid supporting two
> different ways of doing so), we do need that call to be made.
> 
> We need to add the OPP table explicitly and what happened with
> dev_pm_opp_get_opp_table() was accidental and not what I wanted.
> 
> drivers/mmc/host/sdhci-msm.c has an example of how this is done.
> 

Alright, thank you for the clarification.