Re: [PATCH/i386newframe] amd64newframe

[Mark Kettenis <kettenis@chello.nl>]

   Cc: Mark Kettenis <kettenis@chello.nl>, gdb-patches@sources.redhat.com
   From: Andreas Jaeger <aj@suse.de>
   Date: Mon, 19 May 2003 09:09:33 +0200
   User-Agent: Gnus/5.1003 (Gnus v5.10.3) XEmacs/21.4 (Portable Code, linux)
   Content-Type: text/plain; charset=us-ascii

   Michal Ludvig <mludvig@suse.cz> writes:

   > Mark Kettenis told me that:
   >>    (gdb)    warning: Unmapped DWARF Register #16 encountered
   >>    6       return 0;
   >>    (gdb)    warning: Unmapped DWARF Register #16 encountered
   >>    7       }
   >> Ah, that must be the new DWARF CFI frame unwinder.  The return
   >> address
   >> RA isn't mapped.  I could change the DWARF register mapping in
   >> x86-64-tdep.c, but this really is a problem with the DWARF CFI frame
   >> unwinder since the compiler is free to choose any number it wants for
   >> the return address column.  Should be fixed now.
   >
   > IMHO compiler shouldn't choose whatever it wants - RA column is
   > defined in the Amd64 ABI (sec.3.6).

Ah, but the DWARF specification allows the compiler to choose the
number of the RA column.  Even if the amd64 ABI says the RA is in
column 16, there might be other platforms where the column number
changes from frame to frame.  The current code is more correct than
the old code regardless of what we choose to do on amd64.

   > Anyway thanks for moving amd64 target to the new unwinder!

   Yes, thanks!  Michal, can you help testing, please?

Yes, that would be great.  As soon as it performs better than the old
stuff, I'd like to move things over to mainline.

Mark