Use unambiguous inline assembly to load a float variable. GNU as

silently converts 'fld' to 'flds', without taking the actual variable
type into account (!), but clang's integrated assembler rightfully
complains about it.

Discussed with:	cperciva

sys/i386/isa/npx.c

@@ -938,7 +938,7 @@ fpu_clean_state(void)
 	 * the x87 stack, but we don't care since we're about to call
 	 * fxrstor() anyway.
 	 */
-	__asm __volatile("ffree %%st(7); fld %0" : : "m" (dummy_variable));
+	__asm __volatile("ffree %%st(7); flds %0" : : "m" (dummy_variable));
 }
 #endif /* CPU_ENABLE_SSE */