RFC: fix PR 12707

Tom Tromey tromey@redhat.com
Fri Jan 18 15:04:00 GMT 2013


>>>>> "Jan" == Jan Kratochvil <jan.kratochvil@redhat.com> writes:

Jan> On Fri, 18 Jan 2013 15:30:43 +0100, Tom Tromey wrote:
>> It works ok now precisely because the debug symbols don't include this
>> information.  Here it is without debuginfo:
>> 
>> (gdb) p GDB<char>::even_harder<int> ('a')
>> No symbol "GDB<char>" in current context.
>> (gdb) p 'int GDB<char>::even_harder<int>' ('a')
>> $1 = 97

Jan> This is bad but it is not a regression.  Also I find such expression less
Jan> essential to GDB than to place a breakpoint there.

Again, we are talking about different things.

That whole sub-thread was in the context of extending all symbols to
have the return type.  If we do that, then we will regress some
expressions.

The status quo is that some symbols have return types and some do not.
This causes bugs.  My argument is that uniformity is better.

Tom



More information about the Gdb-patches mailing list