[PATCH 1/3] Fix Unlikely(x) == y

[Adrian Bunk]

On Sun, Feb 17, 2008 at 10:50:03PM +1100, Michael Ellerman wrote:
> On Sat, 2008-02-16 at 10:39 -0800, Arjan van de Ven wrote:
>...
> > for mordern (last 10 years) x86 cpus... the cpu branchpredictor is better
> > than the coder in general. Same for most other architectures.
> > 
> > unlikely() creates bigger code as well.
> > 
> > Now... we're talking about your super duper hotpath function here right?
> > One where you care about 0.5 cycle speed improvement? (less on modern
> > systems ;)
> 
> The first patch was to platforms/ps3 code, which runs on the Cell, in
> particular the PPE ... which is not an x86 :)
> 
> eg:
> 
> [michael at schoenaich ~]$ cat branch.c
> #include <stdio.h>
> int main(void)
> {
>         int i, j;
> 
>         for (i = 0, j = 0; i < 1000000000; i++)
>                 if (i % 4 == 0)
>                         j++;
> 
>         printf("j = %d\n", j);
>         return 0;
> }
> [michael at schoenaich ~]$ ppu-gcc -Wall -O3 -o branch branch.c
> [michael at schoenaich ~]$ time ./branch
> real    0m5.172s
> 
> [michael at schoenaich ~]$ cat branch.c
> ..
>         for (i = 0, j = 0; i < 1000000000; i++)
>                 if (__builtin_expect(i % 4 == 0, 0))
>                         j++;
> ..
> [michael at schoenaich ~]$ ppu-gcc -Wall -O3 -o branch branch.c
> [michael at schoenaich ~]$ time ./branch
> real    0m3.762s
> 
> 
> Which looks as though unlikely() is helping. Admittedly we don't have a
> lot of kernel code that looks like that, but at least unlikely is doing
> what we want it to.


This means it generates faster code with a current gcc for your platform.

But a future gcc might e.g. replace the whole loop with a division
(gcc SVN head (that will soon become gcc 4.3) already does 
transformations like replacing loops with divisions [1]).

And your __builtin_expect() then might have unwanted effects on gcc.

Or the kernel code changes much but the likely/unlikely stays unchanged
although it becomes wrong.

If it is a real hotpath in the kernel where you have _measurable_ 
performance advantages from using likely/unlikely it's usage might be 
justified, but otherwise it shouldn't be used.


> cheers

cu
Adrian

[1] e.g. the while() loop in timespec_add_ns() in include/linux/time.h
    gets replaced by a division and a modulo (whether this 
    transformation is correct in this specific case is a different
    question, but that's the level of code transformation gcc already 
    does today)

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed