[RFC] legacy_virtual_frame_pointer: SP_REGNUM might be a pseudo-register

[RFC] legacy_virtual_frame_pointer: SP_REGNUM might be a pseudo-register

[Kevin Buettner]

Any comments on the patch below?

I found this change necessary to avoid internal errors in some gdb.trace
tests when running against the mips simulator.

	* arch-utils.c (legacy_virtual_frame_pointer): SP_REGNUM might
	be a pseudo-register; revise test accordingly.

Index: arch-utils.c
===================================================================
RCS file: /cvs/src/src/gdb/arch-utils.c,v
retrieving revision 1.145
diff -u -p -r1.145 arch-utils.c
--- arch-utils.c	28 Apr 2007 21:52:38 -0000	1.145
+++ arch-utils.c	7 May 2007 21:39:03 -0000
@@ -198,7 +198,7 @@ legacy_virtual_frame_pointer (CORE_ADDR 
      with things like Dwarf2's CFI.  */
   if (DEPRECATED_FP_REGNUM >= 0 && DEPRECATED_FP_REGNUM < NUM_REGS)
     *frame_regnum = DEPRECATED_FP_REGNUM;
-  else if (SP_REGNUM >= 0 && SP_REGNUM < NUM_REGS)
+  else if (SP_REGNUM >= 0 && SP_REGNUM < NUM_REGS + NUM_PSEUDO_REGS)
     *frame_regnum = SP_REGNUM;
   else
     /* Should this be an internal error?  I guess so, it is reflecting

Re: [RFC] legacy_virtual_frame_pointer: SP_REGNUM might be a pseudo-register

[Ulrich Weigand]

Kevin Buettner wrote:

> 	* arch-utils.c (legacy_virtual_frame_pointer): SP_REGNUM might
> 	be a pseudo-register; revise test accordingly.

I understand those register numbers will be passed on to the tracepoint
stub -- this will probably not understand GDB's pseudo registers ...

Bye,
Ulrich

-- 
  Dr. Ulrich Weigand
  GNU Toolchain for Linux on System z and Cell BE
  Ulrich.Weigand@de.ibm.com

Re: [RFC] legacy_virtual_frame_pointer: SP_REGNUM might be a pseudo-register

[Daniel Jacobowitz]

On Mon, May 07, 2007 at 02:48:24PM -0700, Kevin Buettner wrote:
> Any comments on the patch below?
> 
> I found this change necessary to avoid internal errors in some gdb.trace
> tests when running against the mips simulator.
> 
> 	* arch-utils.c (legacy_virtual_frame_pointer): SP_REGNUM might
> 	be a pseudo-register; revise test accordingly.

In addition to what Ulrich said, I'm planning to make SP_REGNUM into a
real register on MIPS; last week I did a first set of patches in this
direction and I hope to continue.

-- 
Daniel Jacobowitz
CodeSourcery

Re: [RFC] legacy_virtual_frame_pointer: SP_REGNUM might be a pseudo-register

[Kevin Buettner]

On Mon, 7 May 2007 19:31:39 -0400
Daniel Jacobowitz <drow@false.org> wrote:

> On Mon, May 07, 2007 at 02:48:24PM -0700, Kevin Buettner wrote:
> > Any comments on the patch below?
> > 
> > I found this change necessary to avoid internal errors in some gdb.trace
> > tests when running against the mips simulator.
> > 
> > 	* arch-utils.c (legacy_virtual_frame_pointer): SP_REGNUM might
> > 	be a pseudo-register; revise test accordingly.
> 
> In addition to what Ulrich said, I'm planning to make SP_REGNUM into a
> real register on MIPS; last week I did a first set of patches in this
> direction and I hope to continue.

Okay, patch withdrawn.

Thanks,

Kevin