public inbox for gcc-help@gcc.gnu.org
 help / color / mirror / Atom feed
* x86 implementation for GCC development
@ 2001-11-07  6:46 Florian Weimer
  2001-11-07  8:11 ` guerby
  0 siblings, 1 reply; 4+ messages in thread
From: Florian Weimer @ 2001-11-07  6:46 UTC (permalink / raw)
  To: gcc-help

To cut down compilation times (bootstrapping the Ada and C front ends
currently requires well over three hours over here), I think I've got
to acquire a new piece of hardware.  Ordinary benchmarks from credible
parties are usually run on Microsoft Windows, and involve software
compiled by dozens of compilers, but not GCC.

So my question: Which of the two available x86 implementations should
I choose, in order to minimize bootstrapping time?  FPU performance
and so on does not matter.

Price vs. performance is probably not so much an issue.  However,
non-x86 architectures are a bit too expensive. :-/

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: x86 implementation for GCC development
  2001-11-07  6:46 x86 implementation for GCC development Florian Weimer
@ 2001-11-07  8:11 ` guerby
  2001-11-07  8:47   ` guerby
  2002-01-20 10:37   ` Florian Weimer
  0 siblings, 2 replies; 4+ messages in thread
From: guerby @ 2001-11-07  8:11 UTC (permalink / raw)
  To: fw; +Cc: gcc-help

You can find a lot of kernel compile time information on Athlon vs
Pentium, eg:

<http://www.linuxhardware.org/features/01/07/13/145249.shtml>

IIRC in most tests I've seen, Athlon is better at compiling the kernel
and cheaper as well.

For reference, on my P3 1GHz laptop, i686-pc-linux-gnu bootstrap for
"c,ada" takes 45 minutes, "ada,c,c++,f77" takes 1 hour,
gnatlib_and_tools is an additional 5 minutes. For "c,c++,f77", make
check is 45 minutes, and ACATS is 35 minutes.

Hope this helps,

-- 
Laurent Guerby <guerby@acm.org>

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: x86 implementation for GCC development
  2001-11-07  8:11 ` guerby
@ 2001-11-07  8:47   ` guerby
  2002-01-20 10:37   ` Florian Weimer
  1 sibling, 0 replies; 4+ messages in thread
From: guerby @ 2001-11-07  8:47 UTC (permalink / raw)
  To: guerby; +Cc: fw, gcc-help

For the record, on P3 1GHz, bootstraping "c" takes 25 minutes, as does
make check.

-- 
Laurent Guerby <guerby@acm.org>

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: x86 implementation for GCC development
  2001-11-07  8:11 ` guerby
  2001-11-07  8:47   ` guerby
@ 2002-01-20 10:37   ` Florian Weimer
  1 sibling, 0 replies; 4+ messages in thread
From: Florian Weimer @ 2002-01-20 10:37 UTC (permalink / raw)
  To: guerby; +Cc: gcc-help

<guerby@acm.org> writes:

> For reference, on my P3 1GHz laptop, i686-pc-linux-gnu bootstrap for
> "c,ada" takes 45 minutes, "ada,c,c++,f77" takes 1 hour,
> gnatlib_and_tools is an additional 5 minutes. For "c,c++,f77", make
> check is 45 minutes, and ACATS is 35 minutes.

After looking at the test results for Athlon XP processors and some
personal tests on a Dual 1 GHz Pentium III box (slightly under 40
minutes for Ada + C + GNAT tools), I decided to buy an Athlon-based
System (with 1.6 GHz clock speed), which compiles the C and Ada front
ends and the Ada library and tools in about 25 minutes, which is quite
impressive compared to over three long hours on my old system. :-)

(BTW: I'm typing this on the new system, please forgive any errors in
the mail setup.)

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2002-01-20 18:37 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2001-11-07  6:46 x86 implementation for GCC development Florian Weimer
2001-11-07  8:11 ` guerby
2001-11-07  8:47   ` guerby
2002-01-20 10:37   ` Florian Weimer

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).