This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Stepanov results with development gcc
- To: dberlin at redhat dot com (Daniel Berlin)
- Subject: Re: Stepanov results with development gcc
- From: Joe Buck <jbuck at racerx dot synopsys dot com>
- Date: Fri, 8 Dec 2000 12:28:27 -0800 (PST)
- Cc: dje at watson dot ibm dot com (David Edelsohn), lucier at math dot purdue dot edu (Brad Lucier), gcc at gcc dot gnu dot org
Daniel Berlin writes:
> Yes.
> In fact, on a gnupro 99r1 based branch compiler (BeOS) on x86, which is the
> same time frame, we have better than perfect.
> It's like .7 or .8.
No, any report with a value < 1 is IMPERFECT, and large deviations
from 1 indicate a bug.
It means that we are misoptimizing test 0, a very simple loop.
All the loops should run at the same rate. If any loop runs slower,
it means that the compiler is putting out pessimal code for that loop.
A perfect compiler will produce 1.0 for every test.