Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.lang.python > #57249
| Newsgroups | comp.lang.python |
|---|---|
| Date | 2013-10-22 02:32 -0700 |
| References | (1 earlier) <mailman.1295.1382326510.18130.python-list@python.org> <5264dbbe$0$30000$c3e8da3$5496439d@news.astraweb.com> <mailman.1308.1382349341.18130.python-list@python.org> <5265bba8$0$29981$c3e8da3$5496439d@news.astraweb.com> <mailman.1345.1382433285.18130.python-list@python.org> |
| Message-ID | <ff24436e-1dfd-4061-9f2d-8a7d8c308d0c@googlegroups.com> (permalink) |
| Subject | Re: Python Front-end to GCC |
| From | Philip Herron <herron.philip@googlemail.com> |
On Tuesday, 22 October 2013 10:14:16 UTC+1, Oscar Benjamin wrote:
> On 22 October 2013 00:41, Steven D'Aprano
>
> >>> On the contrary, you have that backwards. An optimizing JIT compiler
>
> >>> can often produce much more efficient, heavily optimized code than a
>
> >>> static AOT compiler, and at the very least they can optimize different
>
> >>> things than a static compiler can. This is why very few people think
>
> >>> that, in the long run, Nuitka can be as fast as PyPy, and why PyPy's
>
> >>> ultimate aim to be "faster than C" is not moonbeams:
>
> >>
>
> >> That may be true but both the examples below are spurious at best. A
>
> >> decent AOT compiler would reduce both programs to the NULL program as
>
> >> noted by haypo:
>
> >> http://morepypy.blogspot.co.uk/2011/02/pypy-faster-than-c-on-carefully-
>
> > crafted.html?showComment=1297205903746#c2530451800553246683
>
> >
>
> > Are you suggesting that gcc is not a decent compiler?
>
>
>
> No.
>
>
>
> > If "optimize away
>
> > to the null program" is such an obvious thing to do, why doesn't the most
>
> > popular C compiler in the [FOSS] world do it?
>
>
>
> It does if you pass the appropriate optimisation setting (as shown in
>
> haypo's comment). I should have been clearer.
>
>
>
> gcc compiles programs in two phases: compilation and linking.
>
> Compilation creates the object files x.o and y.o from x.c and y.c.
>
> Linking creates the output binary a.exe from x.o and y.o. The -O3
>
> optimisation setting used in the blog post enables optimisation in the
>
> compilation phase. However each .c file is compiled independently so
>
> because the add() function is defined in x.c and called in y.c the
>
> compiler is unable to inline it. It also can't remove it as dead code
>
> because although it knows that the return value isn't used it doesn't
>
> know if the call has side effects.
>
>
>
> You might think it's silly that gcc can't optimise across source files
>
> and if so you're right because actually it can if you enable link time
>
> optimisation with the -flto flag as described by haypo. So if I do
>
> that with the code from the blog post I get (using mingw gcc 4.7.2 on
>
> Windows):
>
>
>
> $ cat x.c
>
> double add(double a, double b)
>
> {
>
> return a + b;
>
> }
>
> $ cat y.c
>
> double add(double a, double b);
>
>
>
> int main()
>
> {
>
> int i = 0;
>
> double a = 0;
>
> while (i < 1000000000) {
>
> a += 1.0;
>
> add(a, a);
>
> i++;
>
> }
>
> }
>
> $ gcc -O3 -flto x.c y.c
>
> $ time ./a.exe
>
>
>
> real 0m0.063s
>
> user 0m0.015s
>
> sys 0m0.000s
>
> $ time ./a.exe # warm cache
>
>
>
> real 0m0.016s
>
> user 0m0.015s
>
> sys 0m0.015s
>
>
>
> So gcc can optimise this all the way to the null program which takes
>
> 15ms to run (that's 600 times faster than pypy).
>
>
>
> Note that even if pypy could optimise it all the way to the null
>
> program it would still be 10 times slower than C's null program:
>
>
>
> $ touch null.py
>
> $ time pypy null.py
>
>
>
> real 0m0.188s
>
> user 0m0.076s
>
> sys 0m0.046s
>
> $ time pypy null.py # warm cache
>
>
>
> real 0m0.157s
>
> user 0m0.060s
>
> sys 0m0.030s
>
>
>
> > [...]
>
> >> So the pypy version takes twice as long to run this. That's impressive
>
> >> but it's not "faster than C".
>
>
>
> (Actually if I enable -flts with that example the C version runs 6-7
>
> times faster due to inlining.)
>
>
>
> > Nobody is saying that PyPy is *generally* capable of making any arbitrary
>
> > piece of code run as fast as hand-written C code. You'll notice that the
>
> > PyPy posts are described as *carefully crafted* examples.
>
>
>
> They are more than carefully crafted. They are useless and misleading.
>
> It's reasonable to contrive of a simple CPU-intensive programming
>
> problem for benchmarking. But the program should do *something* even
>
> if it is contrived. Both programs here consist *entirely* of dead
>
> code. Yes it's reasonable for the pypy devs to test things like this
>
> during development. No it's not reasonable to showcase this as an
>
> example of the potential for pypy to speed up any useful computation.
>
>
>
> > I believe that, realistically, PyPy has potential to bring Python into
>
> > Java and .Net territories, namely to run typical benchmarks within an
>
> > order of magnitude of C speeds on the same benchmarks. C is a very hard
>
> > target to beat, because vanilla C code does *so little* compared to other
>
> > languages: no garbage collection, no runtime dynamism, very little
>
> > polymorphism. So benchmarking simple algorithms plays to C's strengths,
>
> > while ignoring C's weaknesses.
>
>
>
> As I said I don't want to criticise PyPy. I've just started using it
>
> and I it is impressive. However both of those blog posts are
>
> misleading. Not only that but the authors must know exactly why they
>
> are misleading. Because of that I will take any other claims with a
>
> big pinch of salt in future.
>
>
>
>
>
> Oscar
You sir deserve a medal! I think alot of people are taking these sorts of benchmarks completely out of context and its great to see such a well rounded statement.
I applaud you so much! I've been sort of banging my head against the wall to describe what you just did as succinctly as that and couldn't.
Back to comp.lang.python | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
Python Front-end to GCC Philip Herron <herron.philip@googlemail.com> - 2013-10-20 10:56 -0700
Re: Python Front-end to GCC victorgarcianet@gmail.com - 2013-10-20 15:10 -0700
Re: Python Front-end to GCC John Nagle <nagle@animats.com> - 2013-10-22 23:48 -0700
Re: Python Front-end to GCC Philip Herron <herron.philip@googlemail.com> - 2013-10-23 00:25 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-23 09:42 +0100
Re: Python Front-end to GCC John Nagle <nagle@animats.com> - 2013-10-23 13:51 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-20 20:35 -0700
Re: Python Front-end to GCC Steven D'Aprano <steve@pearwood.info> - 2013-10-21 07:46 +0000
Re: Python Front-end to GCC Oscar Benjamin <oscar.j.benjamin@gmail.com> - 2013-10-21 10:55 +0100
Re: Python Front-end to GCC Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2013-10-21 23:41 +0000
Re: Python Front-end to GCC Oscar Benjamin <oscar.j.benjamin@gmail.com> - 2013-10-22 10:14 +0100
Re: Python Front-end to GCC Philip Herron <herron.philip@googlemail.com> - 2013-10-22 02:32 -0700
Re: Python Front-end to GCC Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2013-10-22 12:00 +0000
Re: Python Front-end to GCC Chris Angelico <rosuav@gmail.com> - 2013-10-22 23:20 +1100
Re: Python Front-end to GCC Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2013-10-22 17:27 +0000
Re: Python Front-end to GCC Chris Angelico <rosuav@gmail.com> - 2013-10-23 08:43 +1100
Re: Python Front-end to GCC Dave Angel <davea@davea.name> - 2013-10-22 14:04 +0000
Re: Python Front-end to GCC Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2013-10-22 15:22 +0000
Re: Python Front-end to GCC Grant Edwards <invalid@invalid.invalid> - 2013-10-22 15:39 +0000
Re: Python Front-end to GCC Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2013-10-22 16:40 +0000
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-22 17:50 +0100
Re: Python Front-end to GCC Chris Kaynor <ckaynor@zindagigames.com> - 2013-10-22 09:52 -0700
Re: Python Front-end to GCC Frank Miles <fpm@u.washington.edu> - 2013-10-22 16:53 +0000
Re: Python Front-end to GCC Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2013-10-22 17:23 +0000
Re: Python Front-end to GCC Chris Kaynor <ckaynor@zindagigames.com> - 2013-10-22 10:35 -0700
Re: Python Front-end to GCC Neil Cerutti <neilc@norwich.edu> - 2013-10-22 17:37 +0000
Re: Python Front-end to GCC Oscar Benjamin <oscar.j.benjamin@gmail.com> - 2013-10-22 18:37 +0100
Re: Python Front-end to GCC MRAB <python@mrabarnett.plus.com> - 2013-10-22 18:42 +0100
Re: Python Front-end to GCC Grant Edwards <invalid@invalid.invalid> - 2013-10-22 18:49 +0000
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 04:40 +0100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-25 04:55 -0700
Re: Python Front-end to GCC Dan Sommers <dan@tombstonezero.net> - 2013-10-25 12:55 +0000
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 15:18 +0100
Re: Python Front-end to GCC Ned Batchelder <ned@nedbatchelder.com> - 2013-10-25 10:35 -0400
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-25 11:26 -0700
Re: Python Front-end to GCC Grant Edwards <invalid@invalid.invalid> - 2013-10-25 19:06 +0000
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 19:40 +0100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-25 11:45 -0700
Re: Python Front-end to GCC rusi <rustompmody@gmail.com> - 2013-10-25 11:59 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-25 12:09 -0700
Re: Python Front-end to GCC rusi <rustompmody@gmail.com> - 2013-10-25 12:15 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 20:02 +0100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-25 12:18 -0700
Re: Python Front-end to GCC John Nagle <nagle@animats.com> - 2013-10-26 14:31 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-26 15:10 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-27 15:14 -0700
Re: Python Front-end to GCC rusi <rustompmody@gmail.com> - 2013-10-27 19:15 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-28 08:44 +0000
Re: Python Front-end to GCC rusi <rustompmody@gmail.com> - 2013-10-28 02:31 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 20:36 +0100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-25 12:49 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 21:14 +0100
Re: Python Front-end to GCC Tim Delaney <timothy.c.delaney@gmail.com> - 2013-10-26 07:11 +1100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-25 13:29 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 21:36 +0100
Re: Python Front-end to GCC Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2013-10-26 02:42 +0000
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 21:44 +0100
Re: Python Front-end to GCC Tim Delaney <timothy.c.delaney@gmail.com> - 2013-10-26 07:48 +1100
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 21:56 +0100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-25 14:02 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 22:11 +0100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-25 14:37 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-25 22:56 +0100
Re: Python Front-end to GCC Chris Angelico <rosuav@gmail.com> - 2013-10-26 13:36 +1100
Re: Python Front-end to GCC Neil Cerutti <neilc@norwich.edu> - 2013-10-22 17:15 +0000
Re: Python Front-end to GCC Grant Edwards <invalid@invalid.invalid> - 2013-10-22 18:58 +0000
Re: Python Front-end to GCC Grant Edwards <invalid@invalid.invalid> - 2013-10-22 20:26 +0000
Re: Python Front-end to GCC Grant Edwards <invalid@invalid.invalid> - 2013-10-22 15:36 +0000
Re: Python Front-end to GCC Oscar Benjamin <oscar.j.benjamin@gmail.com> - 2013-10-22 15:15 +0100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-21 13:14 -0700
Re: Python Front-end to GCC Ned Batchelder <ned@nedbatchelder.com> - 2013-10-21 16:29 -0400
Re: Python Front-end to GCC rusi <rustompmody@gmail.com> - 2013-10-21 20:40 -0700
Re: Python Front-end to GCC Philip Herron <herron.philip@googlemail.com> - 2013-10-21 04:08 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-21 13:26 -0700
Re: Python Front-end to GCC Philip Herron <herron.philip@googlemail.com> - 2013-10-21 14:03 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-21 16:04 -0700
Re: Python Front-end to GCC Piet van Oostrum <piet@vanoostrum.org> - 2013-10-21 23:45 -0400
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-21 21:24 -0700
Re: Python Front-end to GCC Steven D'Aprano <steve@pearwood.info> - 2013-10-22 05:25 +0000
Re: Python Front-end to GCC Dave Angel <davea@davea.name> - 2013-10-22 04:39 +0000
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-22 08:04 -0700
Re: Python Front-end to GCC Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2013-10-22 17:09 +0000
Re: Python Front-end to GCC Piet van Oostrum <piet@vanoostrum.org> - 2013-10-22 13:20 -0400
Re: Python Front-end to GCC Ned Batchelder <ned@nedbatchelder.com> - 2013-10-22 11:46 -0400
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-22 16:52 +0100
Re: Python Front-end to GCC Benjamin Kaplan <benjamin.kaplan@case.edu> - 2013-10-22 09:03 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-22 10:50 -0700
Re: Python Front-end to GCC rusi <rustompmody@gmail.com> - 2013-10-22 11:11 -0700
Re: Python Front-end to GCC Neil Cerutti <neilc@norwich.edu> - 2013-10-22 18:18 +0000
Re: Python Front-end to GCC Piet van Oostrum <piet@vanoostrum.org> - 2013-10-22 15:20 -0400
Re: Python Front-end to GCC Neil Cerutti <neilc@norwich.edu> - 2013-10-22 19:27 +0000
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-22 20:38 +0100
Re: Python Front-end to GCC Neil Cerutti <neilc@norwich.edu> - 2013-10-22 20:00 +0000
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-22 20:32 +0100
Re: Python Front-end to GCC Skip Montanaro <skip@pobox.com> - 2013-10-22 13:08 -0500
Re: Python Front-end to GCC MRAB <python@mrabarnett.plus.com> - 2013-10-22 19:16 +0100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-22 11:16 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-22 11:22 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-22 11:28 -0700
Re: Python Front-end to GCC Piet van Oostrum <piet@vanoostrum.org> - 2013-10-22 18:11 -0400
Re: Python Front-end to GCC Chris Angelico <rosuav@gmail.com> - 2013-10-23 17:28 +1100
Re: Python Front-end to GCC Grant Edwards <invalid@invalid.invalid> - 2013-10-22 22:47 +0000
Re: Python Front-end to GCC Ned Batchelder <ned@nedbatchelder.com> - 2013-10-22 14:23 -0400
Re: Python Front-end to GCC rusi <rustompmody@gmail.com> - 2013-10-22 11:40 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-22 19:58 +0100
Re: Python Front-end to GCC Ned Batchelder <ned@nedbatchelder.com> - 2013-10-22 14:40 -0400
Re: Python Front-end to GCC alex23 <wuwei23@gmail.com> - 2013-10-23 11:36 +1000
Re: Python Front-end to GCC rusi <rustompmody@gmail.com> - 2013-10-22 21:04 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-23 07:06 +0100
Re: Python Front-end to GCC MRAB <python@mrabarnett.plus.com> - 2013-10-22 19:47 +0100
Re: Python Front-end to GCC Ned Batchelder <ned@nedbatchelder.com> - 2013-10-22 13:56 -0400
Re: Python Front-end to GCC Michael Torrie <torriem@gmail.com> - 2013-10-22 22:05 -0600
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-23 07:13 +0100
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-26 14:25 -0700
Re: Python Front-end to GCC Mark Janssen <dreamingforward@gmail.com> - 2013-10-26 14:33 -0700
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-26 22:38 +0100
Re: Python Front-end to GCC Mark Lawrence <breamoreboy@yahoo.co.uk> - 2013-10-26 22:35 +0100
Re: Python Front-end to GCC Antoine Pitrou <solipsis@pitrou.net> - 2013-10-22 08:55 +0000
Re: Python Front-end to GCC Philip Herron <herron.philip@googlemail.com> - 2013-10-22 02:08 -0700
Re: Python Front-end to GCC Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2013-10-22 10:10 +0000
Re: Python Front-end to GCC Antoine Pitrou <solipsis@pitrou.net> - 2013-10-22 15:51 +0000
Re: Python Front-end to GCC Stefan Behnel <stefan_ml@behnel.de> - 2013-10-24 08:47 +0200
Re: Python Front-end to GCC xDog Walker <thudfoo@gmail.com> - 2013-10-25 14:49 -0700
csiph-web