Re: [PATCH net-next v1 3/8] pds_core: devlink health: use retained error fmsg API

[Przemek Kitszel <przemyslaw.kitszel@intel.com>]

On 10/10/23 18:53, Nelson, Shannon wrote:
> On 10/10/2023 3:43 AM, Przemek Kitszel wrote:
>>
>> Drop unneeded error checking.
>>
>> devlink_fmsg_*() family of functions is now retaining errors,
>> so there is no need to check for them after each call.
>>
>> Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
>> Signed-off-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
>> ---
>> add/remove: 0/0 grow/shrink: 0/1 up/down: 0/-57 (-57)
>> ---
>>   drivers/net/ethernet/amd/pds_core/devlink.c | 27 ++++++---------------
>>   1 file changed, 7 insertions(+), 20 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/amd/pds_core/devlink.c 
>> b/drivers/net/ethernet/amd/pds_core/devlink.c
>> index d9607033bbf2..fcb407bdb25e 100644
>> --- a/drivers/net/ethernet/amd/pds_core/devlink.c
>> +++ b/drivers/net/ethernet/amd/pds_core/devlink.c
>> @@ -154,33 +154,20 @@ int pdsc_fw_reporter_diagnose(struct 
>> devlink_health_reporter *reporter,
>>                                struct netlink_ext_ack *extack)
>>   {
>>          struct pdsc *pdsc = devlink_health_reporter_priv(reporter);
>> -       int err;
>>
>>          mutex_lock(&pdsc->config_lock);
>> -
>>          if (test_bit(PDSC_S_FW_DEAD, &pdsc->state))
>> -               err = devlink_fmsg_string_pair_put(fmsg, "Status", 
>> "dead");
>> +               devlink_fmsg_string_pair_put(fmsg, "Status", "dead");
>>          else if (!pdsc_is_fw_good(pdsc))
>> -               err = devlink_fmsg_string_pair_put(fmsg, "Status", 
>> "unhealthy");
>> +               devlink_fmsg_string_pair_put(fmsg, "Status", 
>> "unhealthy");
>>          else
>> -               err = devlink_fmsg_string_pair_put(fmsg, "Status", 
>> "healthy");
>> -
>> +               devlink_fmsg_string_pair_put(fmsg, "Status", "healthy");
>>          mutex_unlock(&pdsc->config_lock);
>>
>> -       if (err)
>> -               return err;
>> -
>> -       err = devlink_fmsg_u32_pair_put(fmsg, "State",
>> -                                       pdsc->fw_status &
>> -                                               
>> ~PDS_CORE_FW_STS_F_GENERATION);
>> -       if (err)
>> -               return err;
>> -
>> -       err = devlink_fmsg_u32_pair_put(fmsg, "Generation",
>> -                                       pdsc->fw_generation >> 4);
>> -       if (err)
>> -               return err;
>> -
>> +       devlink_fmsg_u32_pair_put(fmsg, "State",
>> +                                 pdsc->fw_status & 
>> ~PDS_CORE_FW_STS_F_GENERATION);
>> +       devlink_fmsg_u32_pair_put(fmsg, "Generation", 
>> pdsc->fw_generation >> 4);
>>          return devlink_fmsg_u32_pair_put(fmsg, "Recoveries",
>>                                           pdsc->fw_recoveries);
>> +
> 
> Please don't add an extra blank line here.

sure, thanks!

> 
>>   }
> 
> Generally, I like this cleanup.  I did have a similar question about 
> return values as Jiri's comment - how would this do with some code 
> checkers that might complain about ignoring all those return values?

yeah, going full void makes things easier at the end

> 
> Going to full void functions would fix that.  You can add some temporary 
> "scaffolding" code, an intermediate layer to help in the driver 
> conversion, which would then get removed at the end once all the drivers 
> are converted.
> 
> This do_something/check_error/do_something/check_err pattern happens in 
> other APIs in the kernel - see the devlink_info_* APIs, for example: do 
> you foresee changing other interfaces in a similar way?

in principle, it's a good idea; personally, I will be fixing those that
I somehow encounter (like here, I plan to add some devlink health report
for intel ice driver)

> 
> sln
> 
>> -- 
>> 2.40.1
>>
>