X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f X-Recipient: geda-user AT delorie DOT com X-SourceIP: 95.97.163.245 X-Authenticated-Sender: b DOT mykendevelopment AT upcmail DOT nl Message-ID: <52139050.1030601@iae.nl> Date: Tue, 20 Aug 2013 17:50:40 +0200 From: myken User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130625 Thunderbird/17.0.7 MIME-Version: 1.0 To: geda-user AT delorie DOT com Subject: [geda-user] gnetlist VHDL back-end behaviour, please explain. Content-Type: multipart/alternative; boundary="------------090703040806000305090106" Reply-To: geda-user AT delorie DOT com This is a multi-part message in MIME format. --------------090703040806000305090106 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Hello All, I'm trying to understand the VHDL back-end of gnetlist (and writing a tutorial in the process). If I include some symbols from the vhdl library and run netlist I get a vhdl file with a component and the correct definitions: (pin : in Std_Logic) if the "pintype" is defined as IN in the symbol. So far so good. However if I want pins to showup in my top-level ENTRY declaration I need to include the "ipad-1.sym" or "opad-1.sym" in my design. When I include the "ipad-1.sym" I get a symbol with a big OUT printed inside it. This OUT is the result of the pintype definition inside the symbol, its set to OUT (pintype=OUT). Now the fun begins, if I run gnetlist I get a (pin : in Std_Logic) definition in my ENTRY declaration. Even if pintype is set to OUT in the symbol the back-end insists on making it an input type port signal. Apparently the VHDL backend of gnetlist looks at the device attribute (set to IPAD) to determine the direction of the port signal and not the pintype attribute as it does with symbols. Can anyone explain to me why? What is the design philosophy behind this? Thanks and regards, Robert. --------------090703040806000305090106 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Hello All,

I'm trying to understand the VHDL back-end of gnetlist (and writing a tutorial in the process).
If I include some symbols from the vhdl library and run netlist I get a vhdl file with a component and the correct definitions:
(pin : in Std_Logic) if the "pintype" is defined as IN in the symbol.
So far so good.
However if I want pins to showup in my top-level ENTRY declaration I need to include the "ipad-1.sym" or "opad-1.sym" in my design.
When I include the "ipad-1.sym" I get a symbol with a big OUT printed inside it. This OUT is the result of the pintype definition inside the symbol, its set to OUT (pintype=OUT). Now the fun begins, if I run gnetlist I get a (pin : in Std_Logic) definition in my ENTRY declaration. Even if pintype is set to OUT in the symbol the back-end insists on making it an input type port signal.
Apparently the VHDL backend of gnetlist looks at the device attribute (set to IPAD) to determine the direction of the port signal and not the pintype attribute as it does with symbols.
Can anyone explain to me why? What is the design philosophy behind this?

Thanks and regards,
Robert.
--------------090703040806000305090106--