NAME

lofig - mbk logical figure

DESCRIPTION

The lofig is used to describe a hierarchical level of the logical view of a figure. The logical view is also called the netlist, for list of nets, because it represents the interconnections between elements of a cell.

The declarations needed to work on lofig are available in the header file "/labo/include/mlo.h".

The following C structure supports the description of the netlist representation :

typedef struct lofig {
	struct lofig	∗NEXT;
	struct chain	∗MODELCHAIN;
	struct locon	∗LOCON;
	struct losig	∗LOSIG;
	struct ptype	∗BKSIG;
	struct loins	∗LOINS;
	struct lotrs	∗LOTRS;
	struct lotrs	∗LOCAP;
	struct lotrs	∗LORES;
	struct lotrs	∗LOSELF;
	char	∗NAME;
	char	MODE;
	struct ptype	∗USER;
} lofig_list;

NEXT: Pointer to the next lofig loaded in ram.
NAME: Name of the figure. It identifies a figure (or model), so it should be unique in order to warranty consistency of cells libraries. In order to ensure this consistency, the NAME field must be filled with the disk file name while parsing, and must be used as file name for driving.
LOINS: Pointer to the head of the list of instances being used for the model description. See loins(3) for details.
LOCON: Pointer to the head of the list of connectors (terminals) of the model. See locon(3) for details.
LOSIG: Pointer to the head of the list of signal of the model. See losig(3) for details.
BKSIG: Pointer to a hash table that contains arrays of LOSIG. It is only an other way to represent the same elements that in the list, but allows a faster access. The Mbk Logical Functions warranty the consitency between the two representations.
LOTRS: Pointer to the list of transistors belonging to the figure. It is to be noticed that transistors are not instances of a particular type.
LOCAP: Pointer to the list of capacitors belonging to the figure.
LORES: Pointer to the list of resistors belonging to the figure.
LOSELF: Pointer to the list of inductors belonging to the figure.
MODE: Caracter indicating the status of the figure in memory. This field can take two values :

´A´: all the cell is loaded in ram.
´P´: only information concerning the model interface is present, that means connectors, and signals linked to external connectors. All other lists are empty.

MODELCHAIN: Pointer to a chain list, see chain(3) for details, of names. These are the names of the models that are at least instanciated once in the figure.
USER: Pointer to a ptype list, see ptype(3) for details, that is a general purpose pointer used to share informations on the model.

NAME

DESCRIPTION

SEE ALSO