Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: About compiler optimizations

Author: Robert Hyatt

Date: 14:52:43 12/19/02

Go up one level in this thread


On December 19, 2002 at 16:07:47, Matt Taylor wrote:

>On December 19, 2002 at 13:17:47, Gian-Carlo Pascutto wrote:
>
>>On December 19, 2002 at 13:11:12, Robert Hyatt wrote:
>>
>>>Not my experience.  It compiles all of Crafty just fine, and the profile-based
>>>optimizations seem to work well also.
>>
>>Crafty is hardly a refernce. You are in SPEC. Everybody who writes
>>a compiler, especially if they have hardware interests, is going to
>>make damn sure you get compiled correctly. But that doesn't help
>>the rest of us.
>>
>>>As I said, it If you do unsafe aliasing and such, it might
>>>have a problem, but nobody here writes that kind of code...
>>
>>I don't.
>>
>>An internal compiler error is a sure bug anyway, no matter
>>what code it is compiling.
>
>Silly user. The crashes must be your fault! That's a feature, not a bug! ;)
>With feedback like that for the Intel C compiler, perhaps there would really be
>a market for an optimizer with similar performance that actually worked.
>
>I haven't used Intel C much, but what I did compile never gave me issues.
>
>-Matt


I have seen _many_ compiler problems.  More so than I have seen with Intel,
for example.  GCC for solaris won't compile crafty and produce a working
executable.  HP's compiler was broken in long long stuff until I helped the
compiler guys track it down.  The list goes on...

But for me, and everyone _here_ intel's compiler is working flawlessly and
producing faster code than gcc by a wide margin...



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.