RFA: Please review the new C++ demangler patch

[Gabriel Dos Reis]

Phil Edwards <phil@jaj.com> writes:

| On Sat, Jun 28, 2003 at 12:50:05AM +0200, Gabriel Dos Reis wrote:
| > Phil Edwards <phil@jaj.com> writes:
| > 
| > | On Fri, Jun 27, 2003 at 11:36:35PM +0200, Gabriel Dos Reis wrote:
| > | > |  :-)  Why fold
| > | > | a completely unused library into the executable?
| > | > 
| > | > static link against libstdc++-v3, does not mean the whole libstdc++-v3
| > | > (especially the unused bits) is pulled into the executable.  Does it?
| > | > The contention seems to be the dependence on <string> and <vector>.
| > | 
| > | Depends on how smart the linker is, and how fine-grained the library's
| > | object files are.
| > | 
| > | Anyhow, this is getting off-topic.
| > 
| > We cannot get anywhere with handwaving.
| 
| What more would you like me to say?  Static linking doesn't affect the
| demangler issues, so I felt this discussion is getting off-topic.

Some people have been complaining that they don't want a demangler
depending on libstdc++.so at runtime.  If the dependency on
libstdc++.so at runtime time is the issue, then I do not agree with
your statment above.  Or else, the issue is not the dependency on
libstdc++.so at runtime. 

-- Gaby