LayGO® Device Library

Much of the work associated with building and using LayGO® devices is routine and can be automated to a considerable extent. The LayGO® device library demonstrates one approach to implementing the ability to conveniently handle multiprotocol devices and can be seen as an example of a LayGO® implementation. The C-source code of the Device Library is included in the LayGO® Toolkit distribution

The approach is based on using a device database. The database associates a symbolic device name with an ordered list of protocol services. (The services typically represent a major device, beginning with the physical layer, but the lowest layer might be a virtual circuit. If the lowest layer is a virtual circuit, the major device to which it belongs must be opened and connected first.) The order of the list determines the order in which the protocol services will be stacked.

Once the device is constructed, it can be treated as a single connection, despite the fact that it might make use of multiple CIDs (Connection IDentifiers) internally. A device can be opened, connected, detached, attached, disconnected and closed just as individual CIDs are. However, stacking and unstacking are done automatically.

If the top layer in a device is not multiplexed, it is the layer through which data is read and written. Once the device is connected, dev_GetReadWriteCid() will return the CID on which data transfer can take place using normal LayGO® API functions.

If the top layer is multiplexed, it is expected that lgo_Open() will be used to obtain minor device CIDs through which data transfer will take place. Alternatively, devices can be defined which consist solely of a virtual circuit or which have a virtual circuit as the lowest layer.

All the Library's functions are listed below. For detailed descriptions and examples, download the LayGO® documentation.