[RFA] Fix crash of convenience vars with typedefs.

[Tom Tromey]

>>>>> "Ulrich" == Ulrich Weigand <uweigand@de.ibm.com> writes:

Ulrich> Shouldn't the result of value_of_internalvar still have the original
Ulrich> typedef name instead of the type it resolves to?

This would be nice -- the wide character code relies on type names now,
for instance -- but most of gdb already eagerly strips typedefs.  E.g.,
look at value_cast, where the resulting value is not actually of the
type you pass in, but rather the type after stripping typedefs.

Tom