Re: GNU Linker Utility Assert

[Tom Rix]

Nick Clifton wrote:

> Hi Tom,
>
> > It seems to me that a better place to catch this problem would be
> > the assembler.
>
> True - but it would still be nice if the linker could provide a
> helpful error message instead of an assertion failure.
>

Yes.  I agree.
Changing the linker assertion is fine but the message needs to be
changed since this is more a problem with an invalid function descriptor
symbol than a relocation.    I will take care of it.

Symbol creation problems should be caught in the assembler.

To make sure the correct problem is fixed I need a testcase.

Tom

--
Tom Rix
GCC Engineer
address@hidden