[PATCH] Fix postincrement before switch

[Richard Henderson]

On Thu, Mar 13, 2003 at 03:37:58PM -0500, Jakub Jelinek wrote:
> Or should emit_queue be done in expand_start_stmt_expr instead
> (either of the patches fixes the testcase, though I've
> just bootstrapped/regression tested the previously posted patch),
> or in both places?

On second thought, I think we need an emit_queue in 
try_casesi and try_tablejump, for when there's a post-inc
in the index_expr itself.


r~