Size of generated programs (fwd)

Size of generated programs (fwd)

[Bill Walker]

Forwarded message:
> Precedence: bulk
> 
> I've had egcs up and running for a few days now, and have encountered no
> problems with the programs I've compiled (other than the ones I make for 
> myself).  One thing I've noticed is that the stripped executables are 
> significantly larger than those I get with gcc2.  I compile with
> 
> -O3 -ffast-math -funroll-loops -malign-*=0  -malign-double -march=pentium (or
> -m486).
> 
> Any idea why this happens?
> 
> David Ronis
> 
>

	I have had essentially the same experience, with C, FORTRAN,
and C++.  The increase is anywhere from about 2% to about 10%, after
stripping, and running through mcs -d.  OTOH, the egcs compiler seems to
compile things that I just couldn't compile before, so I consider it
a wash.

	My platform is OSR 5.0.4.

BW



-- 
Bill Walker Ph.D.
Chairman, Dept. of Computer Science
East Central University
Ada, Oklahoma 74820-6899

e-mail:  bw@cs.ecok.edu 
phone:   580 332 8000 ext. 594
FAX:     580 332 4616

Re: Size of generated programs (fwd)

[Elliot Lee]

On Mon, 8 Dec 1997, Bill Walker wrote:

> > I've had egcs up and running for a few days now, and have encountered no
> > problems with the programs I've compiled (other than the ones I make for 
> > myself).  One thing I've noticed is that the stripped executables are 
> > significantly larger than those I get with gcc2.  I compile with
> > 
> > -O3 -ffast-math -funroll-loops -malign-*=0  -malign-double -march=pentium (or
> > -m486).
> 
> 	I have had essentially the same experience, with C, FORTRAN,
> and C++.  The increase is anywhere from about 2% to about 10%, after
> stripping, and running through mcs -d.  OTOH, the egcs compiler seems to
> compile things that I just couldn't compile before, so I consider it
> a wash.

Add -fno-exceptions to the command line.

-- Elliot					http://www.redhat.com/
"They don't let my code go into shipping products," Gates said. "They
 haven't done that for eight years." (at the 1997 PDC)

Re: Size of generated programs (fwd)

[Dale Martin]

"Bill Walker" <bw@student.ecok.edu> writes:
> 	I have had essentially the same experience, with C, FORTRAN,
> and C++.  The increase is anywhere from about 2% to about 10%, after
> stripping, and running through mcs -d.  OTOH, the egcs compiler seems to
> compile things that I just couldn't compile before, so I consider it
> a wash.
> 
> 	My platform is OSR 5.0.4.
> 
> BW

As an interesting contrast, I have a program that I'm working that has an
11M binary when compiled with gcc-2.  With egcs, it's 4.5M.  (Stripped in
both cases) And egcs has found some small problems with the code, which is
C++.  Keep up the good work!

Later,
	Dale
-- 
+--------------------  finger for pgp public key  ---------------------+
| Dale E. Martin | University of Cincinnati Savant Research Laboratory |
| dmartin@ececs.uc.edu    |     http://www.ececs.uc.edu/~dmartin       |
+----------------------------------------------------------------------+

Re: Size of generated programs (fwd)

[Bill Walker]

Forwarded message:
> > 
> > 	I have had essentially the same experience, with C, FORTRAN,
> > and C++.  The increase is anywhere from about 2% to about 10%, after
> > stripping, and running through mcs -d.  OTOH, the egcs compiler seems to
> > compile things that I just couldn't compile before, so I consider it
> > a wash.
> 
> Add -fno-exceptions to the command line.
> 
	Sorry -- didn't make much difference at all.

BW


-- 
Bill Walker Ph.D.
Chairman, Dept. of Computer Science
East Central University
Ada, Oklahoma 74820-6899

e-mail:  bw@cs.ecok.edu 
phone:   580 332 8000 ext. 594
FAX:     580 332 4616