Bob, True, when A_to_D or D_to_A are used, the types are implicitly revealed. However, if we have none of those converters, the type is still not known. An example for that is shown in our favorite Figure 12 with node "nd1". If the connections to nd1 would come from [External Circuits] only, and their ports would all be user defined, we wouldn't know the type for them. I think this shouldn't be too hard to fix. One way would be to add a type column to the [Node Declaration] keyword. This would tell the tool what types it should expect in case all connections to it are with user defined port names. We could make this column required for those situations and optional for other situations. Arpad ========================================================= -----Original Message----- From: Bob Ross [mailto:bob@teraspeed.com] Sent: Wednesday, March 26, 2008 4:57 PM To: Muranyi, Arpad Cc: ibis@server.eda.org; ibis-users@server.eda.org Subject: Re: [IBIS] Port type with [External Circuit] Arpad: A clarifcation per suggestion 1. is reasonable. You can make proposals regarding declaring the port type for user-defined ports. I think this will take some work and require revisiting a lot of old issues since a method already exists for converting ports using D_to_A and A_to_D. So, some implicit declaration method is already available. Bob Muranyi, Arpad wrote: > Hello IBIS fans, > > We discovered a potential problem with the port types when using > [External Circuit]s. > > The "Ports under [Model]s" section of the spec discusses the naming > convention of the reserved port names used with [External Model]. > > | The first letter of the port name designates it as either digital ("D") or > | analog ("A"). Reserved ports 1 through 13 listed above are assumed or > | implied under the native IBIS [Model] keyword. > > Later, the "Ports under [External Model]s" section describes the port name > rules for [External Circuit]s: > > | The [External Circuit] keyword allows the user to define any number of ports > | and port functions on a circuit. The [Circuit Call] keyword instantiates > | [External Circuit]s and connects their ports to specific die nodes (this can > | include pads). In this way, the ports of an [External Circuit] declaration > | become specific component die nodes. Note that, if reserved digital port > | names are used with an [External Circuit], those ports will be connected > | automatically as defined in the port list above (under [External Circuit], > | reserved analog port names do not retain particular meanings). > >>From these sections of the spec we know how to deal with the ports of > [External Model]s, i.e. it is known what their type is, but there is a > problem with [External Circuit]s. > > Other than the usage of the reserved digital port names, there is no way > of knowing what the port type is for a user defined port name, because we > do not have a mechanism to declare port types in IBIS. Unfortunately the > [Node Declaration] keyword doesn't do more than list the die nodes and/or > pads, so this keyword would not be useful to find out what the type of a > port may be that is connected to it. > > This is a problem, because some tools need to know the type of a port in > order to know how to connect things. Looking into the [External Circuit] > to find out the port type may not be as easy as it sounds either if only > compiled (or encrypted) models are available. > > I would like to make two suggestions: > > 1) If the "first letter rule" was only intended to apply to the reserved > port names (to be honest, this is not 100 % clear to me from reading the > spec.) we need to write a clarification BIRD to clear this up. > > 2) If the "first letter rule" doesn't apply to user defined port names, > I would propose that we invent a port type declaration mechanism to declare > the user defined port types. > > Any comments or questions are appreciated. > > Thanks, > > Arpad -- Bob Ross Teraspeed Consulting Group LLC Teraspeed Labs 121 North River Drive 13610 SW Harness Lane Narragansett, RI 02882 Beaverton, OR 97008 401-284-1827 503-430-1065 http://www.teraspeed.com 503-246-8048 Direct bob@teraspeed.com Teraspeed is a registered service mark of Teraspeed Consulting Group LLC -- This message has been scanned for viruses anddangerous content by MailScanner, and isbelieved to be clean. -------------------------------------------------------------------- |For help or to subscribe/unsubscribe, e-mail majordomo@eda-stds.org |with the appropriate command message(s) in the body: | | help | subscribe ibis <optional e-mail address, if different> | subscribe ibis-users <optional e-mail address, if different> | unsubscribe ibis <optional e-mail address, if different> | unsubscribe ibis-users <optional e-mail address, if different> | |or e-mail a request to ibis-request@eda-stds.org. | |IBIS reflector archives exist under: | | http://www.eda-stds.org/pub/ibis/email_archive/ Recent | http://www.eda-stds.org/pub/ibis/users_archive/ Recent | http://www.eda-stds.org/pub/ibis/email/ E-mail since 1993Received on Wed Mar 26 16:17:41 2008
This archive was generated by hypermail 2.1.8 : Wed Mar 26 2008 - 16:17:53 PDT