diff --git a/qemu/include/fpu/softfloat.h b/qemu/include/fpu/softfloat.h
index ab9ae929..989a9072 100644
--- a/qemu/include/fpu/softfloat.h
+++ b/qemu/include/fpu/softfloat.h
@@ -252,6 +252,14 @@ static inline int float16_is_any_nan(float16 a)
     return ((float16_val(a) & ~0x8000) > 0x7c00);
 }
 
+static inline float16 float16_abs(float16 a)
+{
+    /* Note that abs does *not* handle NaN specially, nor does
+     * it flush denormal inputs to zero.
+     */
+    return make_float16(float16_val(a) & 0x7fff);
+}
+
 /*----------------------------------------------------------------------------
 | The pattern for a default generated half-precision NaN.
 *----------------------------------------------------------------------------*/