[Devel] Re: [PATCH] fdset's leakage

[Kirill Korotaev]

>>Your logic looks fine for me. Do we have already round_up_pow_of_two() function or
>>should we create it as something like:
>>unsinged long round_up_pow_of_two(unsigned long x)
>>{
>>  unsigned long res = 1 << BITS_PER_LONG;
> 
> 
> You'll get a zero here. Should be 1 << (BITS_PER_LONG - 1).
Good that so many people are watching when you even didn't write it yet :)))
Thanks!

Kirill