[Bug gdb/30344] inconsistent variable values for source-level and instruction-level stepping

["vries at gcc dot gnu.org" <sourceware-bugzilla@sourceware.org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=30344

Tom de Vries <vries at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |NOTABUG
             Status|UNCONFIRMED                 |RESOLVED

--- Comment #1 from Tom de Vries <vries at gcc dot gnu.org> ---
(In reply to Yibiao Yang from comment #0)
> /**********************************
> We can found that, when stepping to source code of line 8, these local
> variables has different values. 
> **********************************/

But you haven't stepped to the same location (to be precise, the same pc) in
both cases.

In one case, using step, you've stepped to the start of line 8.

In the other case using stepi, you've stepped until hitting the first insn that
is part of line 8.

You should continue stepi-ing until you hit the same location, at which point
you'll see the same values:
...
$ gdb -q a.out
Reading symbols from a.out...
(gdb) start
Temporary breakpoint 1 at 0x400430: file small.c, line 6.
Starting program: /data/vries/gdb/a.out 

Temporary breakpoint 1, main (argc=1, argv=0x7fffffffdce8) at small.c:6
6         char c[] = "%i";
(gdb) si
0x0000000000400434      6         char c[] = "%i";
(gdb) 
0x0000000000400439      7         unsigned char uc[] = "%i";
(gdb) 
0x000000000040043e      9         const unsigned char cuc[] = "%i";
(gdb) 
0x0000000000400443      8         const char cc[] = "%i";
(gdb) 
0x0000000000400448      6         char c[] = "%i";
(gdb) 
0x000000000040044d      11        __builtin_sprintf(buf, (char *)c, 1);
(gdb) 
0x0000000000400452      11        __builtin_sprintf(buf, (char *)c, 1);
(gdb) 
0x0000000000400454      7         unsigned char uc[] = "%i";
(gdb) 
0x0000000000400459      11        __builtin_sprintf(buf, (char *)c, 1);
(gdb) 
0x000000000040045e      9         const unsigned char cuc[] = "%i";
(gdb) 
0x0000000000400463      11        __builtin_sprintf(buf, (char *)c, 1);
(gdb) 
0x0000000000400468      8         const char cc[] = "%i";
(gdb) si
0x000000000040046d      6         char c[] = "%i";
(gdb) 
7         unsigned char uc[] = "%i";
(gdb) 
8         const char cc[] = "%i";
(gdb) info locals
buf = "\000\000\000\000\000\000\000\000\000"
c = "%i"
uc = "%i"
cc = "%i\377"
cuc = "%i\177"
(gdb) print /x $pc
$1 = 0x400477
...

Using print /x $pc, we can verify that it is indeed the same location:
...
$ gdb -q a.out
Reading symbols from a.out...
(gdb) start
Temporary breakpoint 1 at 0x400430: file small.c, line 6.
Starting program: /data/vries/gdb/a.out 

Temporary breakpoint 1, main (argc=1, argv=0x7fffffffdce8) at small.c:6
6         char c[] = "%i";
(gdb) s
7         unsigned char uc[] = "%i";
(gdb) 
8         const char cc[] = "%i";
(gdb) p /x $pc
$1 = 0x400477
...

AFAICT, there's no bug in gcc and no bug in gdb.

-- 
You are receiving this mail because:
You are on the CC list for the bug.