Re: [Patch] statistics infrastructure - update 9

[Martin Peschke]

Andi Kleen wrote:
> > Good question. Btw. - faster by what order of magnitude?
>
> pushf + popf is on K8 at least ~18 cycles, on P4 it is much more
> because they synchronize the pipeline there (hundreds of cycles)
>
> cpu local add would be a few cycles at best and doesn't have
> any impact on the pipeline
>
>
> > local_irq_save/restore seems to be fine for kernel/profile.c
> >
> >
> > Reason 1:
> > cpu_local_* uses __get_cpu_var, which conflicts with struct statistic
> > being embedded into struct xyz that is allocated whenever the client
> > needs it.
> >
> > I could try to use local_t in conjunction with local_add etc.
> > (as seen in include/linux/dmaengine.h in 2.6.17-mm6).
> > Does this also yield a performance gain worth consideration?
>
> Yes, but you would need preempt_disable() then. For non preemptible
> kernels (far majority) that would be already a big win.
>
>
> > So, removing local_irq_save/restore would require statistics to be
> > switched on and their buffers being available all the time. That is,
> > buffers holding counters etc. can't be allocated at run time - what
> > if allocation fails? (Should I leave this issue to clients?).
>
> Can't you use RCU for this?
>
>
> > Reason 4:
> > The alleged overhead of local_irq_save/restore (as compared
> > to atomic operations) 
>
> local_* doesn't need to be atomic. IT isn't on x86 at least.
> On some other architectures it can be, but i think it's just a SMOP
> of fixing them.
>
> -Andi

Thanks. I am seriously considering these techniques.

If I manage to use RCU for most of the read-mostly struct statistic
and to push any other locking issues down into individual statistic
disciplines (utilisation indicator, histogram and so on), I should be
able to use local_t for some disciplines (particularyl counter and
histogram) without needing other locking primitives, like
local_irq_save/restore currently currently found in the code.

Not a change that can be done this afternoon, though.
And it will require careful review.

Martin

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/