Re: [PATCH] opp: Don't ignore clk_get() errors other than -ENODEV

From: Dmitry Osipenko
Date: Fri Jan 29 2021 - 14:27:21 EST


29.01.2021 20:46, Viresh Kumar пишет:
> On 29-01-21, 18:23, Dmitry Osipenko wrote:
>> 29.01.2021 13:51, Viresh Kumar пишет:
>>> Not all devices that need to use OPP core need to have clocks, a missing
>>> clock is fine in which case -ENODEV shall be returned by clk_get().
>>>
>>> Anything else is an error and must be handled properly.
>>>
>>> Reported-by: Dmitry Osipenko <digetx@xxxxxxxxx>
>>> Signed-off-by: Viresh Kumar <viresh.kumar@xxxxxxxxxx>
>>> ---
>>> Stephen, is the understanding correct that -ENODEV is the only error
>>> returned for missing clocks ?
>>>
>>> Dmitry: I hope this is on the lines of what you were looking for ?
>>
>> Viresh, thank you! This is not what I was looking for because clk core
>> doesn't return -ENODEV for a missing clock, but -ENOENT. The ENODEV
>> certainly should break drivers.
>
> My bad.
>
>> I was looking for this:
>>
>> diff --git a/drivers/opp/core.c b/drivers/opp/core.c
>> index 0305861fee1b..3dd9cdbc0e75 100644
>> --- a/drivers/opp/core.c
>> +++ b/drivers/opp/core.c
>> @@ -1264,7 +1264,7 @@ static struct opp_table
>> *_update_opp_table_clk(struct device *dev,
>> if (IS_ERR(opp_table->clk)) {
>> int ret = PTR_ERR(opp_table->clk);
>>
>> - if (ret == -EPROBE_DEFER) {
>> + if (ret != -ENOENT) {
>> dev_pm_opp_put_opp_table(opp_table);
>> return ERR_PTR(ret);
>> }
>
> You should be looking for this instead, isn't it ?
>
> diff --git a/drivers/opp/core.c b/drivers/opp/core.c
> index 049d45e70807..4bfcbe5b57af 100644
> --- a/drivers/opp/core.c
> +++ b/drivers/opp/core.c
> @@ -1268,7 +1268,7 @@ static struct opp_table *_update_opp_table_clk(struct device *dev,
> if (!ret)
> return opp_table;
>
> - if (ret == -ENODEV) {
> + if (ret == -ENOENT) {
> dev_dbg(dev, "%s: Couldn't find clock: %d\n", __func__, ret);
> return opp_table;
> }
>
>

This will work too.

Then also could be better to replace the "if (ret != -EPROBE_DEFER)"
with dev_err_probe(..).