Andrew Pinski writes:
>
> On Jul 28, 2006, at 4:47 AM, Richard Guenther wrote:
>
> >
> > The memory requirement for PR12245 will nearly double.
>
> Saying it will double is not prove, please provide the memory usage
> dumps. If it does double then you should not be using x86 to optimize
> the memory usage and instead using powerpc-linux-gnu to do that work.
appended are the timing results for i486, powerpc and sparc, no memory
usage reports yet.
Daniel Jacobowitz writes:
> On Fri, Jul 28, 2006 at 11:44:12AM +0200, Jan Hubicka wrote:
> > Interesting, the major reason for disabling -m64 by default for 32bit
> > compilers was the fact that it enforces HOST_WIDE_INT to be 64bit
> > slowing down the whole compiler considerably. Are Debian's folks happy
> > to wait longer for compilation or has this overhead changed?
>
> They haven't done any measuring I know of, but we needed 64-bit
> compilers for a variety of reasons and this was much less disruptive
> than packaging a second copy of GCC.
"They"?, thought you still join the party ;) the s390, sparc,
powerpc architectures are all 32bit userland, but need at least a
64bit compiler building the kernel. Another reason is to have some
applications to run in 64bit userland.
The timing results are user time (using the time command). sparc
biarch currently doesn't build on the trunk, so no results yet.
2006-07-29 2006-07-21
4.1bi 4.1 4.2bi 4.2
PR8361 -O2 i486 19,69 18,82 95,58% 44,24 45,00 101,72%
PR8361 -O0 i486 5,80 5,11 88,10% 7,73 8,06 104,27%
2006-06-27
PR8361 -O2 powerpc 32,48 32,84 101,11% 76,10 75,78 99,58%
PR8361 -O0 powerpc 9,83 9,80 99,69% 13,69 13,48 98,47%
PR8361 -O2 sparc 46,20 45,49 98,46% n/a 129,38 n/a
PR8361 -O0 sparc 11,33 11,22 99,03% n/a 18,10 n/a
Matthias
