This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Tremendous increase in compile times for 3.4.1 with -mcpu=G5
- From: Dale Johannesen <dalej at apple dot com>
- To: David Edelsohn <dje at watson dot ibm dot com>
- Cc: Bradley Lucier <lucier at math dot purdue dot edu>, Matt Austern <austern at apple dot com>, gcc at gcc dot gnu dot org, Dale Johannesen <dalej at apple dot com>, Michael Matz <matz at suse dot de>
- Date: Sat, 14 Aug 2004 13:45:07 -0700
- Subject: Re: Tremendous increase in compile times for 3.4.1 with -mcpu=G5
- References: <C7AB5C7A-EE24-11D8-8F7A-003065BA681E@math.purdue.edu> <200408142032.i7EKWCD29304@makai.watson.ibm.com>
On Aug 14, 2004, at 1:32 PM, David Edelsohn wrote:
Bradley Lucier writes:
Brad> I've been trying to digest this. Is this a problem that gcc has
with
Brad> 64-bit ppc architecture, i.e., one that will also show up on aix
and
Brad> linux64?
-mcpu=G5 is not a pure 64-bit PowerPC mode, it is a mixed mode:
64-bit computation within a function, but 32-bit ABI. The conversion
back
and forth at function boundaries causes the register allocator fits.
I believe Sparc V8plus mode is the only thing similar. That might well
have
the same problem.