diff options
Diffstat (limited to 'gdb/sparc-tdep.c')
| -rw-r--r-- | gdb/sparc-tdep.c | 16 |
1 files changed, 14 insertions, 2 deletions
diff --git a/gdb/sparc-tdep.c b/gdb/sparc-tdep.c index 3fe4dfc..16e68a1 100644 --- a/gdb/sparc-tdep.c +++ b/gdb/sparc-tdep.c @@ -431,8 +431,20 @@ sparc_frame_chain (struct frame_info *frame) { /* Value that will cause FRAME_CHAIN_VALID to not worry about the chain value. If it really is zero, we detect it later in - sparc_init_prev_frame. */ - return (CORE_ADDR) 1; + sparc_init_prev_frame. + + Note: kevinb/2003-02-18: The constant 1 used to be returned + here, but, after some recent changes to frame_chain_valid(), + this value is no longer suitable for causing frame_chain_valid() + to "not worry about the chain value." The constant ~0 (i.e, + 0xfff...) causes the failing test in frame_chain_valid() to + succeed thus preserving the "not worry" property. I had considered + using something like ``get_frame_base (frame) + 1''. However, I think + a constant value is better, because when debugging this problem, + I knew that something funny was going on as soon as I saw the + constant 1 being used as the frame chain elsewhere in GDB. */ + + return ~ (CORE_ADDR) 0; } CORE_ADDR |