Re: [PATCH] thermal: adding check if the thermal firmware is running

[Jonathan Bither <jonbither@gmail.com>]

On 5/17/24 13:25, Guilherme Giácomo Simões wrote:
> Em sex., 17 de mai. de 2024 às 13:57, Jonathan Bither
> <jonbither@gmail.com> escreveu:
>>
>> On 5/17/24 10:16, Guilherme Giacomo Simoes wrote:
>>> In the dmesg is showing the message "failed to read out thermal zone"
>>> as if the temperature read is failed by don't find the thermal zone.
>>>
>>> After researching and debugging, I see that this specific error is
>>> occurrenced because the thermal try read the temperature when is started,
>>> but the firmware is not running yet.
>>>
>>> For more legibiliti i change the tt.c for return EAGAIN when this was occurrence.
>>> After this change, in my computer I compile and install kernel in /boot
>>> and in my dmesg the message "failed to read out thermal zone" is not show
>>> any more.
>>>
>>> I would like to thanks for Rafael Wysocki <refael.j.wysocki@intel.com> for
>>> your suggestions in mu first patch that results in this another patch.
>>> ---
>>>    drivers/net/wireless/intel/iwlwifi/mvm/tt.c | 10 ++++++++--
>>>    1 file changed, 8 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/tt.c b/drivers/net/wireless/intel/iwlwifi/mvm/tt.c
>>> index 8083c4b2ab6b..68ab9966330c 100644
>>> --- a/drivers/net/wireless/intel/iwlwifi/mvm/tt.c
>>> +++ b/drivers/net/wireless/intel/iwlwifi/mvm/tt.c
>>> @@ -620,8 +620,14 @@ static int iwl_mvm_tzone_get_temp(struct thermal_zone_device *device,
>>>
>>>        mutex_lock(&mvm->mutex);
>>>
>>> -     if (!iwl_mvm_firmware_running(mvm) ||
>>> -         mvm->fwrt.cur_fw_img != IWL_UCODE_REGULAR) {
>>> +     const int res = iwl_mvm_firmware_running(mvm);
>>> +
>>> +     if (!res) {
>>> +             ret = -EAGAIN;
>>> +             goto out;
>>> +     }
>>> +
>> You could skip using the res variable and move the mutex lock here and
>> simplify the above a bit. Ex:
>>
>>           int temp;
>>
>> -       mutex_lock(&mvm->mutex);
>> +       if (!iwl_mvm_firmware_running(mvm))
>> +               return -EAGAIN;
>>
>> -       if (!iwl_mvm_firmware_running(mvm) ||
>> -           mvm->fwrt.cur_fw_img != IWL_UCODE_REGULAR) {
>> +       mutex_lock(&mvm->mutex);
>> +       if (mvm->fwrt.cur_fw_img != IWL_UCODE_REGULAR) {
>>                   ret = -ENODATA;
>>                   goto out;
>>           }
>>
>>> +     if (mvm->fwrt.cur_fw_img != IWL_UCODE_REGULAR) {
>>>                ret = -ENODATA;
>>>                goto out;
>>>        }
> Hey Jonathan, Thank you for your suggestion.
> I sended a v2 patch of this patch
> https://patchwork.kernel.org/project/linux-wireless/patch/20240517171311.3705-1-trintaeoitogc@gmail.com/
>
> If you want, you can send this suggestion in this patch v2.
Hey Guilherme, no worries.
>
> Thanks.