add non-recursive, faster versions of enter/leave

libprof.h

@@ -66,6 +66,14 @@ extern uint32_t volatile *prof_tcnt;
 	if(!--prof.rec) prof.elapsed -= *prof_tcnt;	\
 }
 
+/* Variant of prof_enter()/prof_leave() for non-recursive contexts */
+#define prof_enter_norec(prof) {			\
+	prof.elapsed += *prof_tcnt;			\
+}
+#define prof_leave_norec(prof) {			\
+	prof.elapsed -= *prof_tcnt;			\
+}
+
 /* prof_exec(): Measure a single block of code
    This operation can be used when profiling is not required, and instead
    libprof is used to measure the performance of a single bit of code. Use it