public inbox for gcc@gcc.gnu.org
 help / color / mirror / Atom feed
From: Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at>
To: Andreas Jaeger <aj@suse.de>
Cc: Geoff Keating <geoffk@redhat.com>, <gcc@gcc.gnu.org>
Subject: Re: Compiler for Red Hat Linux 8
Date: Thu, 19 Jul 2001 12:23:00 -0000	[thread overview]
Message-ID: <Pine.BSF.4.33.0107192106360.54441-100000@deneb.dbai.tuwien.ac.at> (raw)
In-Reply-To: <u8g0btru8q.fsf@gromit.moeb>

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 8940 bytes --]

On Thu, 19 Jul 2001, Andreas Jaeger wrote:
>> Here at DBAI, for example, we simply cannot use GCC 3.0 because it's
>> an order of magnitude slower while generating larger and also slower
>> binaries.  I don't know, though, what this means for KDE and other
>> code you need to compile as part of your GNU/Linux distribution.
> So, your advice for distributors would be to wait for GCC 3.1 - or use
> a compiler taken from the GCC 3.1 branch ... [...]
> ... unless somebody improves the GCC 3.0 branch significantly?

Yes.  Or, use GCC 3.0 for all languages except C++ and ship another, newer
version for C++.

Unfortunately, such another, newer version does not exist yet (the current
CVS head is even worse than GCC 3.0), so you might need to use GCC 2.95
for C++ sources and your users requiring C++.

(Which is, of course, *very* weird, considering that improved C++ support
is one of the most important features of GCC 3.0.)

> Btw. what are exactly the problems you're facing with GCC 3.0?

Compile times (and memory usage) are going through the roof due to
the tree-based inliner and the new libstdc++, and the generated
binaries are larger and often (sensibly) slower.

Today I performed benchmarks using a computational logic programming
system called DLV ( http://www.dbai.tuwien.ac.at/proj/dlv ) which uses
template and STL heavy C++ code on one of our ia32 boxes.

It's just one binary, but the benchmarks exercise completely different
modules and thus types of code. And the results are, well, depressing:

Specifically, compare compile times and -O0 and -O3 of 2.95 versus 3.0!

(I'd really be interested to see how build times compare for KDE, Gnome
or other heavy C++ code, ideally also benchmarks.)

Gerald

	   	  GCC 2.95			   GCC 3.0
	Compile time    Binary size      Compile time	Binary Size
-O0		6:19	3915128	 		 8:20	4159780
-O1		4:20	4203480			11:40	4829732
-O2		5:56	4209368			14:09	4862532
-O3		5:47	4221464			32:04	6166052

   2.95 -O0 |2.95 -O1|2.95 -O2|2.95 -O3| 3.0 -O0 | 3.0 -O1| 3.0 -O2| 3.0 -O3|
============================================================================+
STRATCOMP1-ALL                         |                                    |
      14.07 |   8.20 |   8.28 |   8.09 |  115.09 | 110.63 |  94.47 |  94.35 |
STRATCOMP-770.2-Q                      |                                    |
       4.73 |   1.43 |   1.44 |   1.30 |    6.12 |   1.64 |   1.46 |   1.44 |
2QBF1       |                          |                                    |
      84.64 |  27.01 |  25.75 |  25.78 |  132.42 |  37.78 |  35.30 |  40.93 |
PRIMEIMPL2  |                          |                                    |
     118.51 |  20.01 |  19.35 |  18.87 |  190.20 |  27.21 |  24.54 |  28.85 |
ANCESTOR    |                          |                                    |
      48.66 |  13.89 |  13.79 |  13.18 |   58.68 |  12.67 |  12.47 |  11.72 |
3COL-SIMPLEX1                          |                                    |
      44.60 |  13.62 |  13.49 |  12.05 |   53.87 |  12.11 |  12.61 |  11.91 |
3COL-LADDER1|                          |                                    |
     218.88 |  62.88 |  59.50 |  57.33 |  306.52 |  73.05 |  71.62 |  80.45 |
3COL-N-LADDER1                         |                                    |
     110.45 |  24.72 |  23.09 |  24.36 |  178.58 |  27.58 |  25.78 |  28.44 |
3COL-RANDOM1|                          |                                    |
     106.49 |  20.74 |  20.24 |  19.50 |  173.43 |  26.04 |  23.82 |  26.58 |
HP-RANDOM1  |                          |                                    |
      50.18 |  16.62 |  16.64 |  14.82 |   82.80 |  16.35 |  16.65 |  16.28 |
HAMCYCLE-FREE                          |                                    |
      11.18 |   2.70 |   2.63 |   2.10 |   20.91 |   3.20 |   2.96 |   2.49 |
DECOMP2     |                          |                                    |
     159.62 |  35.21 |  35.41 |  33.76 |  196.10 |  29.71 |  29.81 |  34.68 |
BW-P4-Esra-a|                          |                                    |
     359.78 | 110.74 | 109.56 | 108.27 |  557.31 | 118.59 | 115.80 | 122.23 |
BW-P5-nopush|                          |                                    |
      31.00 |   8.82 |   8.73 |   8.48 |   46.51 |   9.12 |   8.90 |   9.39 |
BW-P5-pushbin                          |                                    |
      32.26 |   8.20 |   7.99 |   7.82 |   49.46 |   8.68 |   8.21 |   8.81 |
BW-P5-nopushbin                        |                                    |
      11.30 |   2.91 |   2.90 |   2.71 |   16.68 |   2.92 |   2.92 |   3.05 |
3SAT-1      |                          |                                    |
     363.47 |  63.99 |  61.34 |  63.08 |  577.92 |  81.84 |  74.44 |  83.96 |
3SAT-1-CONSTRAINT                      |                                    |
     207.35 |  36.67 |  32.39 |  32.81 |  343.00 |  43.70 |  39.51 |  46.89 |
HANOI-Towers|                          |                                    |
      25.71 |   6.80 |   6.74 |   6.25 |   31.30 |   6.18 |   5.85 |   6.59 |
RAMSEY      |                          |                                    |
      56.77 |  16.00 |  15.55 |  14.00 |   76.73 |  16.35 |  16.71 |  14.99 |
CRISTAL     |                          |                                    |
      54.62 |  16.57 |  16.09 |  14.95 |   66.37 |  17.54 |  16.29 |  15.37 |
HANOI-K     |                          |                                    |
     361.15 |  64.51 |  59.92 |  58.67 |  563.88 |  77.92 |  71.98 |  80.24 |
21-QUEENS   |                          |                                    |
      88.28 |  19.63 |  18.92 |  18.89 |  175.51 |  23.67 |  22.79 |  24.31 |
MSTDir[V=13,A=40]                      |                                    |
     133.41 |  45.45 |  43.67 |  42.18 |  195.41 |  37.55 |  35.80 |  35.59 |
MSTDir[V=15,A=40]                      |                                    |
     133.48 |  46.41 |  44.56 |  43.25 |  194.97 |  37.75 |  36.06 |  35.05 |
MSTUndir[V=13,A=40]                    |                                    |
      77.61 |  23.47 |  22.87 |  21.37 |  122.09 |  20.18 |  19.24 |  19.16 |
MSTUndir[V=15,A=40]                    |                                    |
    1216.05 | 386.56 | 368.97 | 352.86 | 1911.62 | 328.32 | 314.92 | 302.76 |
TIMETABLING |                          |                                    |
      93.56 |  26.44 |  25.94 |  24.52 |  125.14 |  26.55 |  24.88 |  25.10 |
============================================================================+

STRATCOMP1-ALL:
STRATCOMP, random instance, |companies| = 60, |products| = 180,
all sets

STRATCOMP-770.2-Q:
STRATCOMP, random instance, |companies| = |products| = 770,
first model, with query

2QBF1:
2QBF, 1000 all-quantified, 20 existentially-quantified variables
10000 clauses, 5CNF

PRIMEIMPL2:
Prime Implicants with 180 variables and 774 clauses (all models [246])

ANCESTOR:
Double ancestor board of size 14 (only grounding)

3COL-SIMPLEX1:
3COL simplex graph, |edges| = 1980, |nodes| = 1035, one model

3COL-LADDER1:
3COL ladder graph, |edges| = 2998, |nodes| = 2000, one model

3COL-N-LADDER1:
3COL, propositional Niemelä style
ladder graph, |edges| = 2998, |nodes| = 2000, one model

3COL-RANDOM1:
3COL random graph, |edges| = 1100, |nodes| = 500, one model

HP-RANDOM1:
Hamiltonian Path on a random graph, |edges| = 700, |nodes| = 85, one model
generated with Stanford Graph Base random_graph(85,700,0,0,0,0,0,1,1,33)
undirected graph, represented as a directed one

HAMCYCLE-FREE:
Hamiltonian Cycle with a free guess.
n = 60, one model

DECOMP2:
query decomposition (k=3), one model

BW-P4-Esra-a:
blocksworld problem P4 with Esra's bw_domain_a program

BW-P5-nopush:
blocksworld problem P5 with Axel's C-translation without push

BW-P5-pushbin:
blocksworld problem P5 with Axel's pushed and binarised C-translation

BW-P5-nopushbin:
blocksworld problem P5 with Axel's binarised C-translation without push

3SAT-1:
3SAT with 280 variables and 1204 clauses, randomly generated, one model

3SAT-1-CONSTRAINT:
3SAT with 280 variables and 1204 clauses, randomly generated, one model, constraint encoding

HANOI-Towers:
"Towers of Hanoi" with 3 stacks, 4 disks, and 15 steps.

RAMSEY:
"Ramsey(3,6) != 17"

CRISTAL:
"Deductive database use as done by Christoph Koch in CERN"

HANOI-K:
"Towers of Hanoi" in K with 3 stacks, 4 disks, and 15 steps.

21-QUEENS:
"N-Queens with 21 queens"

MSTDir[V=13,A=40]:
min spanning tree [prim], directed graph with 13 vertices and 40 arcs

MSTDir[V=15,A=40]:
min spanning tree [prim], directed graph with 15 vertices and 40 arcs

MSTUndir[V=13,A=40]:
min spanning tree [prim], undirected graph with 13 vertices and 40 arcs

MSTUndir[V=15,A=40]:
min spanning tree [prim], undirected graph with 15 vertices and 40 arcs

TIMETABLING:
A timetable problem of the first year of the faculty of Science
of University of Calabria for 1 class, one model

  reply	other threads:[~2001-07-19 12:23 UTC|newest]

Thread overview: 55+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-07-17 13:04 Geoff Keating
2001-07-17 15:52 ` Joe Buck
2001-07-17 17:48   ` Per Bothner
2001-07-18  8:55     ` Joseph S. Myers
2001-07-17 18:24 ` Craig Rodrigues
2001-07-18  2:41 ` Andreas Jaeger
2001-07-18  9:03   ` H . J . Lu
2001-07-18 12:01     ` Joe Buck
2001-07-18 12:46       ` H . J . Lu
2001-07-18 13:22         ` Joe Buck
2001-07-18 13:31           ` H . J . Lu
2001-07-18 14:28             ` David Edelsohn
2001-07-18 15:03               ` Joern Rennecke
2001-07-18 15:12                 ` David Edelsohn
2001-07-18 15:24                   ` Joe Buck
2001-07-18 17:05                     ` H . J . Lu
2001-07-19  4:56                     ` Toon Moene
2001-07-18 15:41                 ` Joseph S. Myers
2001-07-18 16:23                   ` H . J . Lu
2001-07-18 12:18     ` Sergey Ostrovsky
2001-07-18 15:19       ` Ken Whaley
2001-07-18 15:30         ` Toon Moene
2001-07-18 15:59           ` Ken Whaley
2001-07-18 16:08             ` Toon Moene
2001-07-18 13:30   ` Gerald Pfeifer
2001-07-19  5:17     ` Andreas Jaeger
2001-07-19 12:23       ` Gerald Pfeifer [this message]
2001-07-18 19:07   ` LinuxVN
2001-07-18 13:44 ` Toon Moene
2001-07-17 17:37 mike stump
2001-07-17 20:00 Benjamin Kosnik
2001-07-18 13:21 Benjamin Kosnik
2001-07-18 14:33 Geoff Keating
2001-07-18 14:41 dewar
2001-07-18 15:29 ` Geoff Keating
2001-07-18 17:50   ` Joe Buck
2001-07-18 18:59 ` Michael Eager
2001-07-18 19:26   ` Justin Guyett
2001-07-19  9:05     ` Mark Mitchell
2001-07-19 19:28   ` akbar A.
2001-07-18 22:10 ` Per Bothner
2001-07-18 22:19   ` Joe Buck
2001-07-18 22:38     ` Per Bothner
2001-07-18 23:00       ` Alex Rosenberg
2001-07-19 14:05       ` Jonathan Larmour
2001-07-18 20:02 dewar
2001-07-19  0:29 Bernard Dautrevaux
2001-07-19  1:16 ` Toon Moene
2001-07-19  1:36 Bernard Dautrevaux
2001-07-19  2:40 ` Joseph S. Myers
2001-07-19  3:02 ` Roman Zippel
2001-07-19  3:12 ` Russ Allbery
2001-07-19  4:33 dewar
2001-07-19 10:49 dewar
2001-07-19 23:16 Bernard Dautrevaux

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=Pine.BSF.4.33.0107192106360.54441-100000@deneb.dbai.tuwien.ac.at \
    --to=pfeifer@dbai.tuwien.ac.at \
    --cc=aj@suse.de \
    --cc=gcc@gcc.gnu.org \
    --cc=geoffk@redhat.com \
    /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 read-only IMAP folder(s) and NNTP newsgroup(s).