Hi!
> When some objects are allocated by one CPU but freed by another CPU we can
> consume lot of cycles doing divides in obj_to_index().
>
> (Typical load on a dual processor machine where network interrupts are
> handled
> by one particular CPU (allocating skbufs), and the other CPU is running the
> application (consuming and freeing skbufs))
>
> Here on one production server (dual-core AMD Opteron 285), I noticed this
> divide took 1.20 % of CPU_CLK_UNHALTED events in kernel. But Opteron are
> quite modern cpus and the divide is much more expensive on oldest
> architectures :
>
> On a 200 MHz sparcv9 machine, the division takes 64 cycles instead of 1 cycle
> for a multiply.
>
> Doing some math, we can use a reciprocal multiplication instead of a divide.
>
> If we want to compute V = (A / B) (A and B being u32 quantities)
> we can instead use :
>
> V = ((u64)A * RECIPROCAL(B)) >> 32 ;
>
> where RECIPROCAL(B) is precalculated to ((1LL << 32) + (B - 1)) / B
Well, I guess it should be gcc doing this optimalization, not we by
hand. And I believe gcc *is* smart enough to do it in some cases...
pavel
--
Thanks for all the (sleeping) penguins.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
