Re: [4.4] Strange performance regression?

[Tim Prince <n8tm at aol dot com>]

Ian Lance Taylor wrote:
> francesco biscani <bluescarni@gmail.com> writes:
>
> > I'm experiencing a strange behaviour with GCC 4.4.1. Basically I have
> > some C++ mathematical code which gets a ~x2 performance drop if I
> > *remove* the following debug line from the code:
>
>
> This message is not appropriate for the mailing list gcc@gcc.gnu.org.
> It is appropriate for gcc-help@gcc.gnu.org.  Please take any followups
> to gcc-help.  Thanks.
>
>
> In my experience, a performance drop in a tight loop when you remove a
> line of code means that your loop is extremely sensitive to cache line
> boundaries.  It can be difficult to find the optimal code other than
> by testing various command line options.  Options to particularly test
> are -falign-loops, -falign-labels, and -falign-jumps.

That seems useful advice.  The align- options could help the hot loops 
fit Loop Stream Detector criteria.  If you set -funroll-loops, you may 
exceed the loop size which fits LSD on older CPUs, but you would often 
make the LSD unnecessary.
> Also, be sure that you are using a -mtune option appropriate for the
> processor on which you are running.  E.g., you mention Core2, so you
> should be using -mtune=core2.
For the 64-bit compiler, the default may be better than core2, but for 
32-bit you should be using at least -march=pentium-m.  If you are using 
vectorizer, -mtune=barcelona could make a difference either way.
How are you controlling which threads run on which cache, in case there 
are cache sharing considerations?