Re: [Tinycc-devel] ASM Output?

[Jim Peters]

Dave Dodge wrote:
> When I run the tests on a Celeron, tcc and gcc are fairly close.  But
> as these numbers show, there appears to be some sort of pathological
> behavior with gcc's output and the P4.  Even if I tell gcc to tune the
> code to the P4, it doesn't get any better:
> 
> Looking at the assembly from gcc, nothing jumps out as an obvious
> problem; but since my attempts to do things like speed up code with
> SSE intrinsics always backfires horribly I probably don't really know
> what to look for :-)

Perhaps you are seeing some weird cache-thrashing behaviour on your P4
caused by some unlucky combination of addresses in the GCC-generated
code?  I have seen unusual situations like this mentioned elsewhere
(e.g. there was one example in the FFTW.org docs, although I can't
find it right now.)

You are doing almost nothing except input/output (and the input/output
should dominate).  I assume you've tried the code without the mapping,
i.e. doing pure byte-by-byte input/output, to see if that is where the
problem lies ... ?

Jim

-- 
 Jim Peters                  (_)/=\~/_(_)                 address@hidden
                          (_)  /=\  ~/_  (_)
 Uazú                  (_)    /=\    ~/_    (_)                http://
 Brighton, UK       (_) ____ /=\ ____ ~/_ ____ (_)            uazu.net