From: Michael Walter <michael.walter@gmail.com>
To: Brian Hurt <bhurt@spnz.org>
Cc: skaller <skaller@users.sourceforge.net>, Jon <jdh30@cam.ac.uk>,
caml-list@yquem.inria.fr
Subject: Re: [Caml-list] The boon of static type checking
Date: Mon, 7 Feb 2005 01:16:05 -0500 [thread overview]
Message-ID: <877e9a17050206221653d14456@mail.gmail.com> (raw)
In-Reply-To: <Pine.LNX.4.44.0502062248540.5563-100000@localhost.localdomain>
On Sun, 6 Feb 2005 23:34:02 -0600 (CST), Brian Hurt <bhurt@spnz.org> wrote:
> Probably a bad idea, but I've got to jump in here.
>
> Full disclosure: I *hate* C++. Mainly because I've actually written real
> programs in it. The next time I have to use C++ in any sort of serious
> way I'm registering c++sucks.com and starting a website to catalog all the
> different ways C++ sucks. Feel free to stop reading at this point.
:-)
> ...
> > g++ seems to generate better
> > code than ocamlopt for similar simple problems
> > (see Alioth for quantitative evidence given silly
> > set of sample 'problems')
>
> Yep. And, conservatively, 10 times as much effort has gone into the gcc
> optimizer as the Ocaml optimizer. Possibly 100 times. For, according to
> Alioth, about a 10% improvement. It's only with gcc 3.x that C++ managed
> to beat Ocaml on performance.
More effort having gone into gcc and better performance of gcc are
arguments pro gcc, right? ;-)
> > IMHO the single major inefficiency in C++ is also a source
> > of efficiency -- lack of a garbage collector.
>
> It's a source of efficiency on the small scale- it's easy to write a 1,000
> line program with hand allocation. Rather harder to write a 10,000 line
> program, and a major bitch to write a 100,000 line program without garbage
> collection.
Personally I like it that in C++ you actually have the choice to use
appropriate garbage collection schemes when you desire to do (yep,
multiple kind of GCs for different subsystems/data/... is a win).
Makes it easier with > 1,000,000 line programs :-)
> > OTOH Ocaml has a massive inefficiency C++ does not,
> > namely boxing, and ocaml tools currently do not
> > seem to be very good at unboxing.
>
> This is also a big efficiency for Ocaml, for two reasons. First, it
> encourages code reuse. Second, Ocaml only needs one implementation of a
> function for all types. C++'s template system, to unbox everything,
> duplicates code like crazy- generally, a new implementation for every
> type.
Very true. Technically this would not be very difficult in C++,
though. There's a certain trade off you have to make for either way.
> > Again, C++ provides inline functions which
> > provide a way Ocaml does not have so easily
> > for optimising.
>
> Don't assume that inlining is optimization. Actually, it generally isn't.
> Having actually timed it on modern hardware, a function call costs like
> 2-3 clock cycles these days. Plus 1-2 clock cycles per argument. This is
> compared to the 10-30 clock cycles a mispredicted branch costs, the 20+
> clock cycles an L1 cache miss/L2 cache hit costs, and the 100-350+ clock
> cycles of an L2 cache miss/memory fetch.
Inlining for very small functions generally is an optimization.
> What little inlining is a good idea can be done by the compiler,
Yep, usually the "inline when appropriate" compiler switch + something
like__forceinline for very small functions (for paranoia reasons)
seems to be the way to go in my experience.
Michael
next prev parent reply other threads:[~2005-02-07 6:16 UTC|newest]
Thread overview: 169+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-02-02 21:31 Estimating the size of the ocaml community Yaron Minsky
2005-02-02 21:36 ` [Caml-list] " Christopher A. Watford
2005-02-02 21:54 ` Frédéric Gava
2005-02-03 3:58 ` skaller
2005-02-03 6:35 ` Erik de Castro Lopo
2005-02-03 16:29 ` Olivier Pérès
2005-02-03 18:06 ` Thomas Fischbacher
2005-02-03 18:34 ` Frédéric Gava
2005-02-03 21:16 ` Thomas Fischbacher
2005-02-03 21:58 ` Paul Snively
2005-02-03 22:42 ` Bardur Arantsson
2005-02-03 23:29 ` Thomas Fischbacher
2005-02-03 22:33 ` josh
2005-02-03 23:22 ` Thomas Fischbacher
2005-02-03 23:39 ` Richard Jones
2005-02-04 9:04 ` Frédéric Gava
2005-02-04 9:37 ` Richard Jones
2005-02-04 10:11 ` Olivier Andrieu
2005-02-04 11:14 ` Frédéric Gava
2005-02-04 12:15 ` Richard Jones
2005-02-04 12:46 ` Marcin 'Qrczak' Kowalczyk
2005-02-04 12:51 ` Gerd Stolpmann
2005-02-04 13:43 ` Richard W. M. Jones
2005-02-04 16:01 ` Gerd Stolpmann
2005-02-04 16:52 ` Oliver Bandel
2005-02-04 17:21 ` Frédéric Gava
2005-02-04 17:55 ` Oliver Bandel
2005-02-04 16:48 ` Oliver Bandel
2005-02-04 12:15 ` Olivier Andrieu
2005-02-04 16:42 ` Oliver Bandel
2005-02-04 10:58 ` Oliver Bandel
2005-02-04 17:27 ` Damien Doligez
2005-02-04 17:59 ` Oliver Bandel
2005-02-04 1:17 ` Michael Walter
2005-02-04 10:53 ` Oliver Bandel
2005-02-04 22:01 ` Thomas Fischbacher
2005-02-05 12:27 ` Oliver Bandel
2005-02-06 0:08 ` Thomas Fischbacher
2005-02-03 23:29 ` Richard Jones
2005-02-04 2:33 ` Jon Harrop
[not found] ` <877e9a170502031856175260c8@mail.gmail.com>
2005-02-04 2:56 ` Michael Walter
2005-02-04 10:26 ` [Caml-list] The boon of static type checking Jon Harrop
2005-02-04 17:02 ` Damien Doligez
2005-02-04 18:00 ` Jon Harrop
2005-02-04 20:38 ` Christophe TROESTLER
2005-02-04 21:42 ` Jon Harrop
2005-02-04 22:11 ` Christophe TROESTLER
2005-02-05 0:58 ` Jon Harrop
2005-02-05 1:52 ` Shivkumar Chandrasekaran
2005-02-07 18:47 ` Damien Doligez
2005-02-05 5:24 ` Jacques Garrigue
2005-02-04 21:52 ` Thomas Fischbacher
2005-02-04 22:27 ` Christophe TROESTLER
2005-02-05 10:00 ` Remi Vanicat
2005-02-06 11:18 ` Christophe TROESTLER
2005-02-04 22:55 ` Thomas Fischbacher
2005-02-06 0:02 ` Thomas Fischbacher
2005-02-06 0:56 ` Erik de Castro Lopo
2005-02-06 10:03 ` Thomas Fischbacher
2005-02-06 1:34 ` Richard Jones
2005-02-06 2:30 ` Erik de Castro Lopo
2005-02-06 9:54 ` Marcin 'Qrczak' Kowalczyk
2005-02-06 10:05 ` Thomas Fischbacher
2005-02-05 21:48 ` Michael Walter
2005-02-06 10:22 ` Radu Grigore
2005-02-06 12:16 ` Gerd Stolpmann
2005-02-06 14:59 ` skaller
2005-02-06 22:30 ` Radu Grigore
2005-02-07 3:15 ` Erik de Castro Lopo
2005-02-06 17:28 ` Jon
2005-02-06 22:26 ` Radu Grigore
2005-02-07 2:51 ` skaller
2005-02-07 1:54 ` skaller
2005-02-07 5:34 ` Brian Hurt
2005-02-07 6:16 ` Michael Walter [this message]
2005-02-07 14:58 ` Igor Pechtchanski
2005-02-12 15:22 ` Brian Hurt
2005-02-12 16:11 ` Thomas Fischbacher
2005-02-12 18:47 ` Brian Hurt
2005-02-12 21:58 ` Thomas Fischbacher
2005-02-12 17:06 ` skaller
2005-02-12 22:57 ` Michael Walter
2005-02-13 1:12 ` Thomas Fischbacher
2005-02-13 1:51 ` Tony Edgin
2005-02-13 2:12 ` Thomas Fischbacher
2005-02-13 10:26 ` Daniel Heck
2005-02-13 18:28 ` Thomas Fischbacher
2005-02-13 20:52 ` Michael Walter
2005-02-13 21:42 ` Thomas Fischbacher
2005-02-13 22:51 ` Michael Walter
2005-02-13 23:59 ` Thomas Fischbacher
2005-02-14 0:11 ` Michael Walter
2005-02-14 0:42 ` Thomas Fischbacher
2005-02-14 1:11 ` Michael Walter
2005-02-14 1:46 ` Michael Vanier
2005-02-14 1:57 ` Michael Walter
2005-02-14 14:19 ` Stefan Monnier
2005-02-14 14:36 ` [Caml-list] " Thomas Fischbacher
2005-02-14 1:19 ` [Caml-list] " Michael Walter
2005-02-14 17:29 ` Martin Berger
2005-02-14 18:44 ` skaller
2005-02-14 19:17 ` Thomas Fischbacher
2005-02-14 2:22 ` skaller
2005-02-14 8:04 ` Paul Snively
2005-02-14 9:33 ` Thomas Fischbacher
2005-02-14 9:39 ` Thomas Fischbacher
2005-02-14 2:10 ` skaller
2005-02-13 2:27 ` Brian Hurt
2005-02-13 2:34 ` Michael Walter
2005-02-07 10:57 ` Ville-Pertti Keinonen
2005-02-07 16:58 ` skaller
2005-02-07 17:24 ` Ville-Pertti Keinonen
2005-02-07 17:56 ` Paul Snively
2005-02-07 17:59 ` skaller
2005-02-07 17:30 ` skaller
2005-02-07 13:07 ` Marcin 'Qrczak' Kowalczyk
2005-02-12 15:42 ` Brian Hurt
2005-02-07 17:42 ` Ken Rose
2005-02-07 2:23 ` skaller
2005-02-04 9:29 ` [Caml-list] Estimating the size of the ocaml community Thomas Fischbacher
2005-02-04 10:26 ` Andreas Rossberg
2005-02-04 17:54 ` Thomas Fischbacher
2005-02-04 15:43 ` Oliver Bandel
2005-02-04 19:54 ` Christophe TROESTLER
2005-02-04 20:20 ` Karl Zilles
2005-02-04 22:07 ` Thomas Fischbacher
2005-02-04 9:41 ` Richard Jones
2005-02-04 10:03 ` Thomas Fischbacher
2005-02-04 16:00 ` Oliver Bandel
2005-02-04 17:32 ` sejourne_kevin
2005-02-04 18:46 ` Oliver Bandel
2005-02-05 1:49 ` skaller
2005-02-04 8:55 ` Ville-Pertti Keinonen
2005-02-04 9:36 ` Thomas Fischbacher
2005-02-04 10:30 ` Oliver Bandel
2005-02-04 22:02 ` Thomas Fischbacher
2005-02-05 13:14 ` Oliver Bandel
2005-02-05 16:37 ` Why can't types and exceptions be nested (was: Re: [Caml-list] Estimating the size of the ocaml community) Richard Jones
2005-02-05 17:04 ` Basile STARYNKEVITCH
2005-02-05 19:26 ` Oliver Bandel
2005-02-06 2:56 ` skaller
2005-02-04 21:55 ` [Caml-list] Estimating the size of the ocaml community Basile STARYNKEVITCH
2005-02-03 19:04 ` ronniec95
2005-02-03 20:06 ` skaller
2005-02-03 20:50 ` chris.danx
2005-02-03 21:14 ` Jon Harrop
2005-02-03 21:34 ` chris.danx
2005-02-03 22:07 ` Bardur Arantsson
2005-02-03 21:47 ` Nicolas Cannasse
2005-02-04 3:52 ` skaller
2005-02-04 16:12 ` Oliver Bandel
2005-02-05 2:04 ` skaller
2005-02-03 20:35 ` chris.danx
2005-02-03 8:36 ` sejourne_kevin
2005-02-03 8:39 ` Matthieu Brucher
2005-02-03 16:23 ` Olivier Pérès
2005-02-03 10:10 ` Stefano Zacchiroli
2005-02-03 16:44 ` Vincenzo Ciancia
2005-02-02 22:10 ` [Caml-list] " Kenneth Knowles
2005-02-02 22:40 ` Michael Jeffrey Tucker
2005-02-02 22:52 ` Richard Jones
2005-02-02 23:42 ` Nicolas Cannasse
2005-02-03 6:53 ` Evan Martin
2005-02-03 6:57 ` Eric Stokes
2005-02-03 20:53 ` chris.danx
2005-02-03 23:29 ` Sylvain LE GALL
2005-02-03 23:38 ` sejourne_kevin
2005-02-07 8:49 ` Sven Luther
2005-02-07 9:23 ` Johann Spies
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=877e9a17050206221653d14456@mail.gmail.com \
--to=michael.walter@gmail.com \
--cc=bhurt@spnz.org \
--cc=caml-list@yquem.inria.fr \
--cc=jdh30@cam.ac.uk \
--cc=skaller@users.sourceforge.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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 NNTP newsgroup(s).