Hmm, This seems to have been discussed in the thread beginning at http://www.eda-stds.org/sv-bc/hm/2067.html and in Mantis 49. The LRM text should state this explicitly. The reasoning is still not clear, though. Is it just a by-product of a DPI need? It means that a legal function header is not legal in a prototype declaration. Thanks, Shalom ________________________________ From: owner-sv-bc@server.eda-stds.org [mailto:owner-sv-bc@server.eda-stds.org] On Behalf Of Bresticker, Shalom Sent: Monday, June 26, 2006 12:05 PM To: sv-bc@server.verilog.org Subject: [sv-bc] import using a function prototype - why must data type be explicit? 20.6 describes using a function prototype in a modport. When you go into the syntax, you find: function_prototype ::= function function_data_type function_identifier ( [ tf_port_list ] ) function_data_type ::= data_type | void function_data_type_or_implicit ::= function_data_type | [ signing ] { packed_dimension } As far as I can see, this means that the data type of the function return value needs to be explicitly specified, whereas in a simple function declaration, it can be left off and then it will be implicitly logic. Why is this? Was it deliberate? Thanks, Shalom Shalom Bresticker Intel Jerusalem LAD DA +972 2 589-6852 +972 54 721-1033 I don't represent Intel
This archive was generated by hypermail 2.1.8 : Mon Jun 26 2006 - 05:02:26 PDT