Re: Add prefetch switch stack hook in scheduler function

[Nick Piggin <nickpiggin@yahoo.com.au>]

Ingo Molnar wrote:
> * Nick Piggin <nickpiggin@yahoo.com.au> wrote:
> 
> 

>>>such as?
>>
>>Not sure. thread_info? Maybe next->timestamp or some other fields in 
>>next, something in next->mm?
> 
> 
> next->thread_info we could and should prefetch - but from the generic 
> scheduler code (see the patch i just sent).
> 

Right. We're always testing the TIF_NEED_RESCHED field after the
switch.

> i'm not sure what you mean by prefetching next->timestamp, it's an 
> inline field to 'next', in the first cacheline of it, which we've 
> already used so it's present. (If you mean the value of next->timestamp, 
> that has no address meaning at all so would lead to unpredictable 
> results on some arches.)
> 

No, I meant the cacheline holding the field of course. I guess I
could have looked for a field further down, but even so, ->timestamp
might be 96 bytes into the structure on a 64-bit arch, which may or
may not be the first cacheline... but you get the idea.

> next->mm we might want to prefetch, but it's probably not worth it 
> because we are referencing it too soon, in context_switch(). (while the 
> kernel stack itself wont be referenced until the full context-switch is 
> done) But might be worth trying - but even then, it should be done from 
> the generic code, like the thread_info and kernel-stack prefetching.
> 
> 
>>I didn't really have a concrete example, but in the interests of being 
>>future proof...
> 
> 
> i'd like to keep generic bits in generic code, and only move things to 
> per-arch include files if absolutely necessary. next->mm is generic.
> 

Yeah, then a specific field _within_ next->mm or thread_info may
want to be fetched. In short, I don't see any argument why we
shouldn't call the function prefetch_task().

Secondly, I don't really like your prefetch(kernel_stack()) function
because it doesn't really give architectures enough control over
exactly what cachelines they get in memory.

prefetching and memory access patterns of all this stuff are fairly
architecture specific. I see nothing wrong with having a prefetch_task()
call. (Although I agree things like thread_info->flags and next->mm can
be done in generic code).

Nick

-- 
SUSE Labs, Novell Inc.

Send instant messages to your online friends http://au.messenger.yahoo.com