Re: GDB bug?

[Bjorn Helgaas <bjorn.helgaas@hp.com>]

On Monday 17 April 2006 14:42, Richard Harke wrote:
> I've been working on a rather frustating problem and
> decided to look at the assembly code for a routine using
> the GDB disassemble command. This showed
>         alloc    r16=ar.pfs,91,88,0
> Since this is an impossible combination, I have to conclude
> that GDB is at fault. (The routine actually takes two pointers
> as parameters)

I took a quick look through the GDB bug database, and didn't see
anything that looked like this problem.

Maybe you could open a problem report here:
    http://www.gnu.org/software/gdb/bugs/

In the report, please include at least:
    - the GDB version
    - the incorrect disassembly of the whole 16-byte bundle
    - a hex dump (e.g, x/x) of the bundle