Intel NETWORK PROCESSOR IXP2800 - Manual (Page 382)

382 Hardware Reference Manual
Intel
®
IXP2800 Network Processor
Performance Monitor Unit
acknowledge signal (CAP_CSR_RD_RDY). When the data is returned, CAP puts the read data
into the Push Data FIFO, arbitrates for the S_Push_Bus, and then the Push/Pull Arbiter pushes the
data to the destination identified in PP_ID.
11.2.3 Configuration Registers
Because the CHAP unit resides on the APB, the offset associated with each of these registers is
relative to the Memory Base Address that the configuration software sets in the PMUADR register.
Each counter has one command, one event, one status, and one data register associated with it.
Each counter is “packaged” with these four registers in a “counter block”. Each implementation
selects the number of counters it will implement, and therefore how many counter blocks (or slices)
it will have. These registers are numbered 0 through N - 1 where N represents the number of
counters - 1. See Figure 140 .
11.3 Performance Measurements
There are several measurements that can be made on each of the hardware blocks. These
measurements together would enable improvements in hardware and software implementation and
architectural issues. Table 152 describes the different blocks and their associated performance
measurement events.
Figure 140. Conceptual Diagram of Counter Array
Register
Interface
Counter
Block 0
Counter
Command
Register
Events
Register
Status
Register
Data
Register
Counter
Block 1
Counter
Command
Register
Events
Register
Status
Register
Data
Register
Counter
Block 2
Counter
Command
Register
Events
Register
Status
Register
Data
Register
Event
Signals