[Bug sanitizer/61185] New: Wrong value in error message

[Bug sanitizer/61185] New: Wrong value in error message

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61185

            Bug ID: 61185
           Summary: Wrong value in error message
           Product: gcc
           Version: 4.10.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: sanitizer
          Assignee: unassigned at gcc dot gnu.org
          Reporter: mpolacek at gcc dot gnu.org
                CC: dodji at gcc dot gnu.org, dvyukov at gcc dot gnu.org,
                    jakub at gcc dot gnu.org, kcc at gcc dot gnu.org

int
main ()
{
  unsigned i = 32, j = 40;
  int u = 9;
  u <<= i - j;
}

with -fsanitize=shift says
s.c:6:5: runtime error: shift exponent -8 is negative
but it should say
s.c:6:5: runtime error: shift exponent 4294967288 is too large for 32-bit type
'int'

[Bug sanitizer/61185] Wrong value in error message

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61185

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.10.0

--- Comment #1 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
And the reason is that we convert the (i - j) to int here:

10501           /* Convert the non vector shift-count to an integer, regardless
10502              of size of value being shifted.  */
10503           if (TREE_CODE (TREE_TYPE (op1)) != VECTOR_TYPE
10504               && TYPE_MAIN_VARIANT (TREE_TYPE (op1)) !=
integer_type_node)
10505             op1 = convert (integer_type_node, op1);

[Bug sanitizer/61185] Wrong value in error message

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61185

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
IMHO not a bug.  The compiler really uses (int) (i - j) as the shiftcount, and
that is negative.

[Bug sanitizer/61185] Wrong value in error message

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61185

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
But, if you want to use the source type before this conversion, make sure to
also
test stuff like:
int
main ()
{
  unsigned long long l = 0xffffffff00000001ULL;
  volatile int u = 9;
  u <<= l;
  return 0;
}
where the shift count after (int) conversion is in range, while before the
conversion it is not (ILP32 and LP64 targets only here).

[Bug sanitizer/61185] Wrong value in error message

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61185

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

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

--- Comment #4 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Ok, I guess it's not worth it.  Closing.