Re: [patch] Fix C++ demangling of minsyms with symver

[Jan Kratochvil <jan dot kratochvil at redhat dot com>]

On Thu, 02 Jun 2011 20:03:45 +0200, Tom Tromey wrote:
> I wonder about the other untouched calls to cplus_demangle.

I think all the other calls are either for full symtab like
TYPE_FN_FIELD_PHYSNAME - which is no longer ever mangled after physname - or
for user entered mangled name which will (after the patchset being posted) be
found as a fallback in the mangled minsym tables.

Just for non-DWARF debug info readers the full symtab can still contain the
mangled names.  And for DWARF there will be needed some way to find the
mangled names anyway (regression PR symtab/12707).

So the fix is incomplete for non-DWARF readers but there would be needed first
some conclusion on how to fix symtab/12707.


Thanks,
Jan