pgcc-1.0.1 with and without haifa

[Paul Koning]

>>>>> "Ronald" == Ronald Wahl <Ronald.Wahl@informatik.tu-chemnitz.de> writes:

 Ronald> Hi, I just nbenched pgcc-1.0.1 with and without the haifa
 Ronald> scheduler. I noticed that with the haifa scheduler I get
 Ronald> faster integer but slower floating point code than without it
 Ronald> ( on a K6 ). The following optimization(s) will increase
 Ronald> fp-performance on a K6:
...
 Ronald> * Replacing FDIV with FMUL where possible.

I think that is a win on just about every machine architecture ever
created, for integer as well as float.  It may make sense as a general
optimization rule.

	paul