Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx> writes:
> On 07/17/2018 08:25 PM, Huang, Ying wrote:
>>> Seriously, though, does it hurt us to add a comment or two to say
>>> something like:
>>>
>>> /*
>>> * Should not even be attempting cluster allocations when
>>> * huge page swap is disabled. Warn and fail the allocation.
>>> */
>>> if (!IS_ENABLED(CONFIG_THP_SWAP)) {
>>> VM_WARN_ON_ONCE(1);
>>> return 0;
>>> }
>> I totally agree with you that we should add more comments for THP swap
>> to improve the code readability. As for this specific case,
>> VM_WARN_ON_ONCE() here is just to capture some programming error during
>> development. Do we really need comments here?
>
> If it's code in mainline, we need to know what it is doing.
>
> If it's not useful to have in mainline, then let's remove it.
OK. Will add the comments.
Best Regards,
Huang, Ying
