[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <NC3QCS.LEAYSBJ8Y8GC1@lyndeno.ca>
Date: Mon, 29 Apr 2024 15:21:59 -0600
From: Lyndon Sanche <lsanche@...deno.ca>
To: Mario Limonciello <mario.limonciello@....com>
Cc: pali@...nel.org, W_Armin@....de, srinivas.pandruvada@...ux.intel.com,
ilpo.jarvinen@...ux.intel.com, Matthew Garrett <mjg59@...f.ucam.org>,
Hans de Goede <hdegoede@...hat.com>, platform-driver-x86@...r.kernel.org,
linux-kernel@...r.kernel.org, Dell.Client.Kernel@...l.com
Subject: Re: [PATCH v3] platform/x86: dell-laptop: Implement platform_profile
On Mon, Apr 29 2024 at 12:45:19 PM -05:00:00, Mario Limonciello
<mario.limonciello@....com> wrote:
> On 4/29/2024 11:48, Lyndon Sanche wrote:
>> #include <linux/i8042.h>
>> #include <linux/debugfs.h>
>> #include <linux/seq_file.h>
>> +#include <linux/platform_profile.h>
>> +#include <linux/bitfield.h>
>
> These should be inserted in alphabetical order.
Agree
>> +
>> + // Clean up but do not fail
>
> Switch comment style to /* */
Agree
>
>> + if (platform_profile_register(thermal_handler))
>> + kfree(thermal_handler);
>
> Don't you also want to return an error in this case? Because this
> means that the platform supports thermal modes but it failed to setup
> due to other issues.
>
> It's different than the case of no supported modes in which case
> returning 0 makes sense.
>
> Maybe like this:
>
>
> ret = platform_profile_register(thermal_handler);
> if (ret)
> kfree(thermal_handler);
>
> return ret;
Good idea, I will propogate this error. Failure of module will then
occur on allocation or platform_profile error.
>
>
>>
>> goto fail_rfkill;
>> }
>> + // Do not fail module if thermal modes not supported, just skip
>
> Switch comment style to /* */
Agree.
Thank you for this feedback.
Lyndon
>
Powered by blists - more mailing lists