Hardware Based Packet Classification for High Speed Internet Routers

7.3 Packing Approach 89
7.3.2 Shadow Packing
In strict partitioning, we viewed columns as the primary dimension of TCAM chips
and sought to pack tables into xed width columns. In shadow packing, on the other
hand, we view rows as the primary dimension of TCAM chips and seek to pack tables
within xed height rows. We consider shadow packing because of the following
two observations.
First, with strict partitioning, when tables of varying width are allocated to the
same column, the number of bits assigned to each table t is equal to h(t) × w(t ′ )
where t ′ is the widest table assigned to that column. This leads to many unused bits
if tables of different widths are assigned to the same column. On the other hand,
horizontally packed tables can be placed next to each other as keeping the vertical
boundaries across multiple tables is unnecessary. Of course, there may be wasted
bits if tables of different heights are packed in the same column. We will allow
tables to be stacked in the same row if they t within the row boundaries.
Second, with strict partitioning, the table ID’s between tables in different columns
cannot be shared. Thus, the number of bits used for table IDs grows essentially linearly
with the number of columns. On the other hand, horizontally aligned tables
in the same row can potentially share some “row ID” bits in their table IDs; these
tables would be distinguished by their horizontal offsets.
Based on the above two observations, we design the shadow packing scheme
that achieves more space efciency by packing tables horizontally and allowing
multiple tables to share bits in their IDs. We rst dene the concept of shadowing
for table ID inheritance and the concept of shadow packing trees for representing
table ID inheritance. Then, we present the shadow packing algorithm and discuss
the procedure for processing packets with shadow packing.
7.3.2.1 Shadowing
In Figure 7.9(a), table t 0 shadows tables t 00 and t 01 .Wedene the concept of shadowing
as follows:
Denition 7.2 (Shadowing Relationship). For a table t stored in a TCAM, we
use VBegin(t) and VEnd(t) to denote the vertical indexes of the TCAM entries
where the table begins and ends respectively, and use HBegin(t) and HEnd(t) to
denote horizontal indexes of the TCAM bit columns where the table begins and
ends respectively. For any two tables t 1 and t 2 where [VBegin(t 2 ),VEnd(t 2 )] ⊆
[VBegin(t 1 ),VEnd(t 1 )] and HEnd(t 1 ) < HBegin(t 2 ),wesayt 1 shadows t 2 .
When table t 1 shadows t 2 , the ID of t 1 can be reused as part of t 2 ’s ID. Suppose
table t shadows tables t 1 ,⋅⋅⋅,t m , because t’s ID denes the vertical TCAM region
[Begin(t), End(t)], each t i (1 ≤ i ≤ m) can use t’s ID to distinguish t i from tables
outside [Begin(t),End(t)] vertically, and use ⌈logm⌉ bits to distinguish t i from tables
inside [Begin(t), End(t)] vertically. Horizontally, table t and each table t i can be
distinguished by properly setting the GMR of the TCAM.