Re: Optimization techniques and undefined behavior

On 01/05/19 12:40, Bart wrote:
> for (int i=0; i<1000000000; ++i) {
> y=x*3/3;
> z+=y;
> ++x;
> }


Given that gcc can optimise y=x*3/3 to y=x
I would have expected the whole loop to be recognised
as computing the sum of an arithmetic sequence
and therefore optimised away completely.


This blog post:


https://xebia.com/blog/gcc-compiler-optimizations-dissection-of-a-benchmark/


discusses a program with the loop:


for( a = 0; a < arg1; a++ )
{
      sum += a;
}


and claims that gcc with -O3 will optimise
the loop to take only costant time.


But this is not the case on my machine with gcc 4.9.2


--
Martin


Dr Martin Ward | Email: martin@gkc.org.uk | http://www.gkc.org.uk
G.K.Chesterton site: http://www.gkc.org.uk/gkc | Erdos number: 4

Return to the comp.compilers page.
Search the comp.compilers archives again.