Re: [cdt-dev] Memory usage

[turnham@xxxxxxxxxx]

Hi Andrew,

Yes this is a known issue...The quick answer to your question is that Yes
there is a way to control how much memory the parser uses.  If you
"right-click" on your project and choose "Properties", one of the available
options is "Parse Quality".  By default this is set to the Maximum, but
moving the slider to the left will decrease the amount of work the parser
does and the information it stores.  Of course changing the parse quality
will affect the depth of information provided.  (You'll have to remove your
project's parse information and re-parse for the change to take effect).

At the "highest" setting the parser does (or tries to do) a relatively
complete job of parsing right down to the statement\identifier level and
creates references such as "variable uses", "function callees\callers",
etc.  At the "lowest" setting the parser will only show the highest level
objects (classes, functions, etc).

Improvement of the parser memory consumption is definitely a work-item for
us (me).

Let me know if changing the Parse Quality allows the parse to finish
parsing your project.

Thanks...

============================
  Jeff Turnham
  Linux Application Development
  IBM Toronto Lab, D2-328
  (905) 413-4319
============================


                                                                                                                    
                    aph@xxxxxxxxxx                                                                                  
                    Sent by:              To:     cdt-dev@xxxxxxxxxxx                                               
                    cdt-dev-admin@e       cc:                                                                       
                    clipse.org            Subject:     [cdt-dev] Memory usage                                       
                                                                                                                    
                                                                                                                    
                    04/12/2002                                                                                      
                    01:54 PM                                                                                        
                    Please respond                                                                                  
                    to cdt-dev                                                                                      
                                                                                                                    
                                                                                                                    



Firstly, sorry if this is a FAQ or covered somewhere in the docs.

I've been trying to use the CDT on some fairly large projects.
What I've found it that the parser's memory requirements are quite
alarming.

For example, when trying to parse gcc I ran out of memory after about
330 files.  (The average gcc source file is about 16k.)  I then expanded
the max Java heap to 256M with

    setenv IBM_JAVA_OPTIONS '-Xmx268435456'

and this time managed to parse 2990 files before running out of
memory.  However, many projects have more source than this and I don't
know that it's reasonable to require users to purchase multiple
gigabytes of memory if they want to use Eclipse on large code bases.

So, is this a known issue?  Is there anything I can do to reduce
memory usage?  Is perhaps the parser getting confused by something in
the gcc source?

Thanks,
Andrew.

_______________________________________________
cdt-dev mailing list
cdt-dev@xxxxxxxxxxx
http://dev.eclipse.org/mailman/listinfo/cdt-dev