Re: [PATCH v4 4/4] KEYS: Avoid false positive ENOMEM error on key read

[Waiman Long]

On 3/18/20 4:27 AM, David Howells wrote:
> Waiman Long <longman@xxxxxxxxxx> wrote:
>
>> +static inline void __kvzfree(const void *addr, size_t len)
>> +{
>> +	if (addr) {
>> +		memset((void *)addr, 0, len);
>> +		kvfree(addr);
>> +	}
>> +}
> I wonder if that would be better as "kvfree(memset(...))" as memset() will
> return the address parameter.  If memset is not inline, it avoids the need for
> the compiler to save the parameter.
>
> David

Doing this is micro-optimization. As the keys subsystem is that
performance critical, do we need to do that to save a cycle or two while
making the code a bit harder to read?

Cheers,
Longman