On Oct 22, 2009, at 10:57, ext Laurent Desnogues wrote:
> On Wed, Oct 21, 2009 at 1:01 PM, wrote:
>> Current code is broken at least on gcc 4.2, the result of a
>> comparison
>> "-1 >= sizeof(type) * 8" results true and causes wrong code path to
>> be
>> taken. The fix utilizes abs() functi
On Wed, Oct 21, 2009 at 1:01 PM, wrote:
> Current code is broken at least on gcc 4.2, the result of a comparison
> "-1 >= sizeof(type) * 8" results true and causes wrong code path to be
> taken. The fix utilizes abs() function where applicable and otherwise
> adds a test to ensure both arguments
Current code is broken at least on gcc 4.2, the result of a comparison
"-1 >= sizeof(type) * 8" results true and causes wrong code path to be
taken. The fix utilizes abs() function where applicable and otherwise
adds a test to ensure both arguments are positive before making the
aforementioned comp
