Messages in this thread | | | Date | Tue, 01 Dec 2009 14:47:12 +0900 | From | Tejun Heo <> | Subject | Re: [PATCH] percpu: explain quick paths in pcpu_[de]populate_chunk() |
| |
Hello,
On 12/01/2009 02:40 PM, Cong Wang wrote: >> So, I don't know. The first iteration only loop looks a bit unusual >> for sure but it isn't something conceptually convoluted. > > Now this seems to be better. So with this change, we can do: > > pcpu_first_pop_region(chunk, rs, re, start, end); > if (rs < re && ...) > return; > > Right?
Yeap, but is that any better? Passing lvalue loop parameters to loop constructs is customary. For almost all other cases, it's not, so
pcpu_first_pop_region(chunk, &rs, &re, start, end)
would be better but then we have two similar looking interfaces which take different types of parameters. Also, you probably can drop rs < re test there but for me it just seems easier to simply check the first iteration. If you think it's something worth changing and it looks better afterwards, please send a patch.
Thanks.
-- tejun
| |