FYI, I'm back on the C++ camp now.

Tom Tromey
Mon Aug 12 18:55:00 GMT 2013

>>>>> "Matt" == Matt Rice <> writes:

Matt> I still haven't mustered up motivations for this... where the main
Matt> issue is that of nested structures,
Matt> both c, and c++ support nested structures but incompatibly, the c
Matt> compatible subset of c++ does not.
Matt> this in means de-nesting struct foo {  struct bar { } } like
Matt> interfaces, which I would probably advocate skipping and never fully
Matt> compiling with -Wc++-compat rather than denesting all of the
Matt> structures and putting them back together (in the case of a move to
Matt> c++)...

Matt> and given the ugliness with denesting all of gdb's structures I
Matt> figured someone would object to that anyways were we to stay at
Matt> -Wc++-compat as I don't really think it helps the code clarity.

FWIW I think denesting them is fine.
Usually if they have a tag it means they're referred to by other code.
So treating them globally makes sense -- it is how the code already
works, it is just written in a goofy way.


More information about the Gdb mailing list