linux-kernel - Re: [PATCH v4 2/2] cpufreq: tegra124: Allow building as a module

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <bc0a4c75-df5b-4b4d-9c9e-a2c99d3f4f6e@nvidia.com>
Date: Thu, 15 May 2025 07:41:54 +0100
From: Jon Hunter <jonathanh@...dia.com>
To: Aaron Kling <webgeek1234@...il.com>
Cc: "Rafael J. Wysocki" <rafael@...nel.org>,
 Viresh Kumar <viresh.kumar@...aro.org>,
 Thierry Reding <thierry.reding@...il.com>, linux-pm@...r.kernel.org,
 linux-kernel@...r.kernel.org, linux-tegra@...r.kernel.org
Subject: Re: [PATCH v4 2/2] cpufreq: tegra124: Allow building as a module

On 14/05/2025 17:43, Aaron Kling wrote:
> On Mon, May 12, 2025 at 11:26 PM Aaron Kling <webgeek1234@...il.com> wrote:
>>
>> On Fri, May 9, 2025 at 8:38 AM Jon Hunter <jonathanh@...dia.com> wrote:
>>>
>>>
>>>
>>> On 09/05/2025 01:04, Aaron Kling via B4 Relay wrote:
>>>> From: Aaron Kling <webgeek1234@...il.com>
>>>>
>>>> This requires three changes:
>>>> * A soft dependency on cpufreq-dt as this driver only handles power
>>>>     management and cpufreq-dt does the real operations
>>>
>>> Hmmm .. how is this handled for other drivers using the cpufreq-dt
>>> driver? I see the imx driver has a dependency on this.
>>
>> A hard dependency would likely make more sense here. I can update this
>> in a new revision. When I first set the soft dependency, I wasn't
>> certain how the driver worked, so I was trying to be less intrusive.
> 
> I remember why I added this soft dep now. The kconfig already has a
> dependency on cpufreq_dt. However, this driver doesn't call any
> functions directly in that driver. It just builds a platform device
> struct for it, then registers it. This results in depmod not requiring
> cpufreq_dt for tegra124_cpufreq. So I added the softdep to work around
> that, so modprobing tegra124_cpufreq by itself functions properly. Is
> there a better way to make depmod map this as needed?

Yes and that is understood. I see a few drivers calling ...

  platform_device_register_simple("cpufreq-dt", -1, NULL, 0);

One option, and I don't know if this would be acceptable, would be to 
add a new wrapper function in the cpufreq-dt driver for the above that 
other drivers could call and that would create the dependency you need.

Jon

-- 
nvpublic