Mail Archives: pgcc/1999/06/23/15:23:37.1
On Tue, Jun 22, 1999 at 03:11:18AM -0500, Jon wrote:
> CFLAGS = -Wall -mpentium -mcpu=pentiumpro -march=pentiumpro -O4
> -malign-loops=2 -malign-jumps=2 -malign-functions=2 -malign-double
> --fast-math -pg
>
> WHY is pgcc so much slower? I'm happy to give over my code in order for
> anyone(Marc?) to look at this. pgcc is about 30% slower. Shouldn't it
> be the other way around?? :) That's one reason why I was also very
> confused.
Well, for one thing you are using profiling, don't do that if you want
performance ;) The other thing is that pgcc is sometimes slower, yes.
Usually its suboptimal use of the flags (you have to invest more work in
compiler flags with pgcc indeed..), and sometimes its just pgcc generating
slower code.
> I see another message about the k6 stuff actually being slower too.
> Seems pgcc has a problem of sorts doing the opposite it intends!
Then you definitely shouldn't use it ;) pgcc does fine in most cases.
Profiling is not such a case, for example. If egc sis better with
profiling use that instead!
--
-----==- |
----==-- _ |
---==---(_)__ __ ____ __ Marc Lehmann +--
--==---/ / _ \/ // /\ \/ / pcg AT goof DOT com |e|
-=====/_/_//_/\_,_/ /_/\_\ XX11-RIPE --+
The choice of a GNU generation |
|
- Raw text -