Copy the sign bit when truncating a float tarval.

[r24403]

diff --git a/ir/tv/fltcalc.c b/ir/tv/fltcalc.c
index 0f162da..eedee4d 100644 (file)
--- a/ir/tv/fltcalc.c
+++ b/ir/tv/fltcalc.c
@@ -811,7 +811,10 @@ static void _trunc(const fp_value *a, fp_value *result) {
        /* and the mask and return the result */
        sc_and(_mant(a), temp, _mant(result));
 
-       if (a != result) memcpy(_exp(result), _exp(a), value_size);
+       if (a != result) {
+               memcpy(_exp(result), _exp(a), value_size);
+               result->sign = a->sign;
+       }
 }
 
 /********