Messages in this thread | | | From | Denis Vlasenko <> | Subject | Re: [Patch 17/23] mask v2 = [6/7] nodemask_t_ia64_changes | Date | Fri, 9 Apr 2004 23:04:18 +0300 |
| |
On Friday 09 April 2004 20:53, Paul Jackson wrote: > > > It may well make sense for the O(1) scheduler to be inlining this. > > > > Why? > > I was thinking that perhaps this call was in a certain critical > performance path of the O(1) scheduler.
Even if it is used in some time critical place, adding call overhead ~350 byte function will barely be noticeable on speed, but *will* be noticeable on size.
> Turned out it wasn't - see further Nick Piggin's followups to this > same thread. > > My latest bitmap/cpumask patch moves this out of line, for ia64. > The other arch's that use this large find_next_bit() code might > want to move it out too.
We have far too many large inlines to kill them one by one. Nedd to automate that. -- vda
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |