Patch for IA64 Linux build, PR build/12646

[Steve Ellcey <sje@cup.hp.com>]

I submitted a defect, PR build/12646, about gdb not building on IA64
Linux.  I have come up with this patch to fix it but I am not sure
if it is the right patch, I.e. is this the right stop_reason function
to put into the libunwind_frame_unwind structure.  Apparently this
file is only used when using the system unwind library and IA64 may
be the only platform that does that.  It does fix the build and when
I run the testsuite I get this summary on the gdb.base runs:

                === gdb Summary ===

# of expected passes            4995
# of unexpected failures        171
# of unexpected successes       1
# of expected failures          29
# of known failures             20
# of untested testcases         3
# of unsupported tests          4

I don't have a baseline for what failures to expect on IA64 Linux,
I don't run the gdb testsuite very often and my recollection is that
it is not a very clean run even before the build broke.

Can someone look this patch over and approve it if it looks OK?
I'd like to get the build problem fixed for 7.3 as well as on the
main line.

Steve Ellcey
sje@cup.hp.com


2011-04-07  Steve Ellcey  <sje@cup.hp.com>

	* libunwind-frame.c (libunwind_frame_unwind): Add
	default_frame_unwind_stop_reason value to initalization.


Index: libunwind-frame.c
===================================================================
RCS file: /cvs/src/src/gdb/libunwind-frame.c,v
retrieving revision 1.31
diff -u -c -r1.31 libunwind-frame.c
*** libunwind-frame.c	13 Jan 2011 16:23:22 -0000	1.31
--- libunwind-frame.c	7 Apr 2011 21:26:39 -0000
***************
*** 237,242 ****
--- 237,243 ----
  static const struct frame_unwind libunwind_frame_unwind =
  {
    NORMAL_FRAME,
+   default_frame_unwind_stop_reason,
    libunwind_frame_this_id,
    libunwind_frame_prev_register,
    NULL,