Mario Domenech Goulart wrote:
> There's an interesting discussion in the OpenBSD mailing
> list about the use of inline.

There is no such thing as "best optimization". Some optimization 
techniques tend to show better performance on the average, but given a 
specific code, the same technique can fail dramatically.

Inlining makes the code faster if the called function is small 
(preferably smaller than a cache-line), and called only a few times. No 
inlining will result in faster code if the function is large (at least 
one complete cache-line is used), and that function has a good chance to 
be already in the cache when called.

Also inlined functions allow the block-reordering code to do interesting 
things. Block-reordering was introduced in gcc3, BSDs are using 
gcc-2.95, AFAIK.

So far the facts.

Now the manual says that for gcc the inline-limit is 600 bytes, compared 
to the 32-128 byte cache line size of the x86 CPUs. It could be worth to 
try lower numbers with inline-functions...


/Ervin

--
gentoo-performance@g.o mailing list