Author: Robert Hyatt
Date: 15:34:07 08/15/98
Go up one level in this thread
On August 14, 1998 at 12:57:34, fca wrote:
>On August 13, 1998 at 14:10:20, Ian Osgood wrote:
>
>>Robert Hyatt said:
>
>>>Crafty is
>>>going to be in the new SpecInt98 benchmark... it will be the first 64 bit
>>>integer program and the first non-numerical algorithm that will really stress-
>>>test compiler optimizers
>
>>So soon hundreds of compiler writers around the world will be tuning their
>>compilers to make Crafty run faster! I LIKE IT!
>
>How can you suggest that compiler authors would ever try to cheat by optimising
>their compilers for specific benchmarks... it would never occur to them! Or if
>they recognise a timing loop, of course they would never scale down the number
>of times the loop is traversed.
>
>;-))
>
>Kind regards
>
>Federation of Compiler Authors
I don't mind if they cheat or not, just so they get "long long" support up to
snuff. IE this is ugly in most compilers:
long long i;
if (i) { } or if (!i) { }
because most do a "load 32 bits, test for zero, jump if zero, load other 32
bits, test for zero, jump if zero, fall into the first block. I'd asked the
gcc guys a couple of years ago to change this to load first 32 bits, or with
second 32 bits, test for zero, jump if zero, so there is *one* jump, which is
better.
Ditto for other long long optimizations. They are bad at present. Maybe they
will get better.. :)
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.