Apologies. I'd applied Reini's patch to my optimization branch. I think
it got lost when I switched to Dave's variant my original code (which
didn't have Reini's C90 fix). Sorry!
STATIC void
S_null_listop_in_list_context(pTHX_ OP *o)
{
+ OP *kid;
+
PERL_ARGS_ASSERT_NULL_LISTOP_IN_LIST_CONTEXT;
/* This is an OP_LIST in list context. That means we
* can ditch the OP_LIST and the OP_PUSHMARK within. */
- OP *kid = cLISTOPo->op_first;
+ kid = cLISTOPo->op_first;
/* Find the end of the chain of OPs executed within the OP_LIST. */
while (kid->op_next != o) {
assert(kid);