This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: Some benchmarks comparing gcc-3.4.3 and 4.0.0 on PowerPC G5

From: Bradley Lucier <lucier at math dot purdue dot edu>
To: Steven Bosscher <stevenb at suse dot de>
Cc: Bradley Lucier <lucier at math dot purdue dot edu>, gcc at gcc dot gnu dot org
Date: Sun, 2 Jan 2005 12:36:07 -0500
Subject: Re: Some benchmarks comparing gcc-3.4.3 and 4.0.0 on PowerPC G5
References: <C09506A8-5C8F-11D9-A1A3-003065BA681E@math.purdue.edu> <200501021147.09910.stevenb@suse.de>

On Jan 2, 2005, at 5:47 AM, Steven Bosscher wrote:

3) You say "each compiler was built without checking".  Do you mean
   built as in "make" instead of "make bootstrap"?

make bootstrap.

About your other comments:

I agree that more thought about which options should be enabled at -O1 may be helpful.

I didn't run tests with -O2 because -O2 results in uniformly slower run times with 3.4.3.

Re:

2) You should probably compare columns with matching flags.  Comparing
   "4.0.0 (no options)" with "3.4.3 -mcpu=G4 -mtune=G5" is not really
   useful IMHO.

Well, it's interesting to me that tweaking options that are supposed to choose more accurately the instruction set and/or the scheduling for a specific processor does not uniformly help run times.

Brad

References:
- Some benchmarks comparing gcc-3.4.3 and 4.0.0 on PowerPC G5
  - From: Bradley Lucier
- Re: Some benchmarks comparing gcc-3.4.3 and 4.0.0 on PowerPC G5
  - From: Steven Bosscher

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]