Re: Suggestion for improving C++ parser memory usage

[Gabriel Dos Reis <gdr at integrable-solutions dot net>]

Robert Dewar <dewar@adacore.com> writes:

| Mike Hearn wrote:
| > On Wed, 29 Dec 2004 20:47:44 -0500, Robert Dewar wrote:
| >
| >>Possibly so, but what is the gain, a little bit of memory saved at
| >>compile time? Seems dubious ..
| 
| > Not when you have "only" 256mb of RAM and compiling a large C++ file can
| > flush half your desktop out to swap. I'd love to see reductions in g++
| > memory usage :)
| 
| I just don't think the optimization you suggest will have any significant
| effect in reducing memory used. Do you have data that says otherwise?

I tend to think that formulation of the question is silly, especially
in the context of this discussion.  Let me explain.  

Roger made a suggestion/question

  http://gcc.gnu.org/ml/gcc/2004-12/msg01199.html

of possible ways of reducing the (insane) memory usage in the C++
front-end.  You jumped in with your "seems dubious..." without
actually providing any concrete data.  And when Mike suggested that
any "little bit of memory" counts, you're asking for "data".


It appeared obvious to me that, if Roger was going to send actual
patch, he would also provide data points.  So at this point of
pre-patch, my understanding is that you're merely emitting negative
feedbacks than actually helping.

Did you follow gcc-patches, you would have noticed that bunch of
patches are being applied to GCC (and the C++ front-end in particular)
in the name of < 1% improvement. 
But, I guess, this discussion is not going to be based on facts but on
personal bias...

-- Gaby