[BUG] Regression in 2.14.90 (relative to 2.13.90)

[Carlo Wood]

On Thu, Nov 13, 2003 at 05:09:13PM -0800, H. J. Lu wrote:
> It is a problem since gdb may pick up the wrong .debug_line info
> with PC == 0 if it sees it first. Maybe gdb should ignore .debug_line
> info with PC == 0.

It doesn't??

Then that is a bug in gdb - and this change in binutils
is to work around that bug in gdb then?

I think that the solution is evident then:
change binutils back to produce a PC = 0 for relocs
that do not exist - and have gdb ignore DIE's with
a PC of 0.

-- 
Carlo Wood <carlo@alinoe.com>