On Thu 13 Jun 2024 at 22:46, Uwe Kleine-König <u.kleine-koenig@xxxxxxxxxxxx> wrote:
> Hello,
>
> On Thu, Jun 13, 2024 at 05:54:31PM +0200, Jerome Brunet wrote:
>> > + for (i = 0; i < MESON_NUM_PWMS; i++) {
>> > + meson->channels[i].clk = of_clk_get(np, i);
>> > + if (IS_ERR(meson->channels[i].clk))
>> > + return dev_err_probe(dev,
>> > + PTR_ERR(meson->channels[i].clk),
>> > + "Failed to get clk\n");
>>
>> here it is ok but ..
>>
>> > +
>> > + ret = devm_add_action_or_reset(dev, meson_pwm_s4_put_clk,
>> > + meson->channels[i].clk);
>> > + if (ret)
>> > + return dev_err_probe(dev, ret,
>> > + "Failed to add clk_put action\n");
>>
>> ... here, devm_add_action_or_reset cannot defer, so dev_err_probe is not useful.
>> dev_err() if you must print something. Just "return ret;" would be fine
>> by me
>
> I don't agree. dev_err_probe() has several purposes. Only one of them is
> handling -EPROBE_DEFER. See also commit
> 532888a59505da2a3fbb4abac6adad381cedb374.
I was stuck on the -EPROBE_DEFER usage. Thanks for the heads up
>
> So yes, please use dev_err_probe() also to handle
> devm_add_action_or_reset().
My point here is also that devm_add_action_or_reset() can only fail on
memory allocation, like (devm_)kzalloc. Looking around the kernel, we
tend to not add messages for that and just return the error code,
presumably because those same 'out of memory' messages would proliferate
everywhere.
>
> Best regards
> Uwe
--
Jerome
