Re: [PATCH 0/2] Make chained function calls in expressions work

["Ulrich Weigand" <uweigand at de dot ibm dot com>]

Siva Chandra wrote:

> Restricting to just those values whose address is returned in a hidden
> param (per ABI) is OK, but an implementation detail made me to include
> all types of return values. There are two times when a return value's
> address is required:
> 
> 1. When a method is invoked on the return value: GDB has to evaluate
> the 'this' pointer. This is done in eval.c.

Ah, I knew there was something else I was missing :-)

In this case, I agree with the rest of your explanation.  Please add a
comment in the final version of the patch that explains why the extra
copy is necessary.

There's just one minor change I think would be good: can we allocate that
extra copy on the stack at least only for *class* types, then?  I'd prefer
to avoid this for the common case of routines just returning a scalar.

Thanks,
Ulrich

-- 
  Dr. Ulrich Weigand
  GNU/Linux compilers and toolchain
  Ulrich.Weigand@de.ibm.com