This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Debugging data on Alpha
- To: Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>
- Subject: Re: Debugging data on Alpha
- From: Richard Henderson <rth at redhat dot com>
- Date: Fri, 29 Dec 2000 10:10:17 -0800
- Cc: bernds at redhat dot com, gcc at gcc dot gnu dot org
- References: <10012291242.AA02707@vlsi1.ultra.nyu.edu>
On Fri, Dec 29, 2000 at 07:42:17AM -0500, Richard Kenner wrote:
> Here's an example: look at the variables declared around line 6819
> of expr.c (right before the call to get_inner_reference).
I presume you mean line 6858 --
6856 int volatilep = 0;
6857 unsigned int alignment;
6858 tree tem = get_inner_reference (exp, &bitsize, &bitpos, &offset,
6859 &mode1, &unsignedp, &volatilep,
6860 &alignment);
struct foo { int f; } x;
int bar() { return x.f; }
Breakpoint 4, expand_expr (exp=0x2000009bb80, target=0x200000bd220,
tmode=SImode, modifier=EXPAND_NORMAL) at ../../../egcs-std/gcc/expr.c:6858
6858 tree tem = get_inner_reference (exp, &bitsize, &bitpos, &offset,
(gdb) p volatilep
$1 = 0
(gdb) n
6865 if (tem == exp)
(gdb) p tem
$2 = (union tree_node *) 0x200000a8e00
(gdb) p bitsize
$3 = 32
(gdb) p bitpos
$4 = 0
(gdb) p offset
$5 = (union tree_node *) 0x0
I've tried this with recent cvs gdb (using dwarf2) and 4.18 (using mdebug).
So. Is there something else that should be tried to reproduce the problem?
r~