commit f6ffc404f26e80905d72215d5c688abfbd2d1a33
author Gabe Black <gabeblack@google.com> Fri Nov 02 17:43:29 2018 -0700
committer Gabe Black <gabeblack@google.com> Mon Nov 05 23:10:16 2018 +0000
tree d55ee11d5f9f64994046dd071ed0086834c08dae
parent 45abd002b6072469faf98008304c33961d1d9e6b

systemc: Explicitly keep the sc_port bind alongside sc_in's version.

The sc_in bind function was hiding the one from sc_port by changing
the const-ness of its parameter. This change explicitly exposes the
underlying sc_port version, and keeps it alongside the new sc_in
version.

This seems mildly dangerous and undesirable because now there are two
very similar functions which would both need to be overridden in order
to get new behavior, but I don't think it's any more dangerous and
undesirable than as (perhaps unintentionally) specified in the
standard.

Change-Id: Ib42a1f8e70bc97abeeeb8d614e71c4019b3a2323
Reviewed-on: https://gem5-review.googlesource.com/c/13880
Reviewed-by: Gabe Black <gabeblack@google.com>
Maintainer: Gabe Black <gabeblack@google.com>

src/systemc/ext/channel/sc_in.hh

diff --git a/src/systemc/ext/channel/sc_in.hh b/src/systemc/ext/channel/sc_in.hh
index 3fea8a8..85d089d 100644
--- a/src/systemc/ext/channel/sc_in.hh
+++ b/src/systemc/ext/channel/sc_in.hh
@@ -225,6 +225,8 @@
         _negFinder(*this, &sc_signal_in_if<bool>::negedge_event)
     {}
 
+    using sc_port<sc_signal_in_if<bool>, 1>::bind;
+
     virtual void
     bind(const sc_signal_in_if<bool> &i)
     {
@@ -390,6 +392,8 @@
         _negFinder(*this, &sc_signal_in_if<sc_dt::sc_logic>::negedge_event)
     {}
 
+    using sc_port<sc_signal_in_if<sc_dt::sc_logic>, 1>::bind;
+
     virtual void
     bind(const sc_signal_in_if<sc_dt::sc_logic> &i)
     {