You can define different kind of SITES, which are where and how macros can be placed. You define a SITE type like this:
Code:
SITE siteName
CLASS {PAD | CORE} ;
[SYMMETRY {X | Y | R90} ... ;] (will discuss this later in macro definition)
SIZE width BY height ;
END siteName
The tags should be obvious -- they define the class (in the core or in the IO / PAD region), what symmetry / rotation is allowed if any, and the size of each site. Usually your LEF file will have some SITE definitions, then your DEF file will have a ton of SITE instantiations for all the places standard cells and IO can go.
Then each MACRO definition in your LEF will say what kind of SITE the MACRO needs to sit in.
Does the "SITE" mean the placement point where the standard cells or macros can be placed ?
If so, I am very curious that why there are SIZE section in SITE attribute.
And it seems the CORE SITE is the minimal standard cells.
So why the MACRO SITE size is the same as the MACRO physical size ?
Not placement point, but type of place something can be placed. SITE definition is a type of place in the LEF. SITE instantiations in the DEF file create each place a MACRO of certain SITE types can sit.
owen_li said:
If so, I am very curious that why there are SIZE section in SITE attribute. And it seems the CORE SITE is the minimal standard cells.
So why the MACRO SITE size is the same as the MACRO physical size ?
Thanks !
Each SITE has a SIZE of course! It's the base area that the SITE occupies. A macro with SITE X sits on a SITE of type X, and the sizes should match or be integral multipliers of the base SITE SIZE.
I think if you experiment with LEF and DEF in a layout tool you'll understand this better than my bad English description.