Flexible debug observation point insertion in pipeline designs

1. An apparatus with flexible debug observation-points (DOPs), the apparatus comprising:

a pipeline comprising a plurality of stages;

a plurality of DOPs embedded in the plurality of stages;

a DOP memory provided at an end of the pipeline, wherein the plurality of DOPs are configured to allow shifting data through the plurality of DOPs to the DOP memory; and

a finite state machine (FSM) configured to perform DOP collection by triggering reading and transferring data stored in the plurality of DOPs to the DOP memory, wherein transferring the data comprises shifting the data through the plurality of DOPs to the DOP memory,

wherein the FSM is further configured to set a done-bit when the data from the plurality of DOPs has been transferred to the DOP memory,

wherein the done-bit is detectable by software and the transferred data is accessible by the software when the done-bit is set, and

wherein a DOP of the plurality DOPs comprises a flop-based storage and is identifiable by a unique DOP-identification (ID) associated with the DOP of the plurality DOPs.

2. The apparatus of claim 1 , wherein a stage of plurality of stages includes one or more embedded DOPs.

3. The apparatus of claim 1 , wherein the DOP of the plurality of DOPs is configured to latch data in response to the triggering the by the FSM and the latched data is readable by using a processor command, wherein the processor command is addressable to a DOP-ID associated with a DOP of the plurality DOPs that is enabled to respond to the processor command by making the latched data accessible to a processor bus.

4. The apparatus of claim 1 , wherein the DOP memory comprises a first-in-first-out (FIFO), wherein the FIFO is configured to make the stored data associated with the DOP of the plurality DOPs accessible to software.

5. The apparatus of claim 1 , wherein the FSM is further configured to send commands to read at least some DOPs of the plurality of DOPs sequentially.

6. The apparatus of claim 1 , wherein a stage of the plurality of stages comprises at least one of a logic circuitry, a hash table, a direct table, or a ternary content-addressable memory (TCAM).

7. The apparatus of claim 1 , wherein the DOP of the plurality DOPs includes self-trigger detect logic circuitry, wherein the self-trigger detect logic circuitry is configured to detect an input data matching a predefined criteria and, in response to detecting the input data matching the predefined criteria, to start a self-triggered mode of operation.

8. The apparatus of claim 7 , wherein a respective DOP is configured to capture, in the self-triggered mode of operation, the input data and to create a trigger for downstream DOPS.

9. A method for insertion of flexible debug observation-points (DOPs), the method comprising:

providing a pipeline comprising a plurality of stages;

embedding a plurality DOPs in the plurality of stages;

providing a DOP memory at an end of the pipeline, wherein the plurality of DOPs are configured to allow shifting of data through the plurality of DOPs to the DOP memory;

configuring a finite state machine (FSM) to perform DOP collection by triggering reading and transferring data stored in the plurality of DOPs to the DOP memory wherein transferring the data comprises shifting the data through the plurality of DOPs to the DOP memory;

and

configuring the FSM to set a done-bit when the data from the plurality of DOPs has been transferred to the DOP memory,

wherein the done-bit is detectable by software and the transferred data is accessible by the software when the done-bit is set, and

wherein the DOP of the plurality DOPs comprises a flop-based storage and is identifiable by a unique DOP-identification (ID) associated with that DOP.

10. The method of claim 9 , wherein embedding the plurality DOPs in the plurality of stages comprise embedding one or more DOPs of the plurality DOPs in a stage of plurality of stages.

11. The method of claim 9 , wherein the DOP of the plurality of DOPs is configured to latch data in response to the triggering by the FSM, wherein the latched data is configured to be readable by a processor command, wherein the processor command is addressable to a DOP-ID associated with a DOP of the plurality of DOPs that is enabled to respond to the processor command by making the latched data accessible to a processor bus.

12. The method of claim 9 , wherein providing the DOP memory comprises providing a first-in-first-out (FIFO), and the method comprises configuring the FIFO to make stored data accessible to software.

13. The method of claim 9 , further comprising configuring the FSM to: send commands to read at least some DOPs of the plurality of DOPs sequentially.

14. The method of claim 9 , wherein a stage of the plurality of stages comprises at least one of a logic circuitry, a hash table, a direct table, or a ternary content-addressable memory (TCAM).

15. The method of claim 9 , wherein the DOP of the plurality DOPs include self-trigger detect logic circuitry, wherein the method further comprises:

configuring the self-trigger detect logic circuitry to detect an input data matching a predefined criteria; and

in response to detecting the input data matching the predefined criteria, and starting a self-triggered mode of operation.

16. The method of claim 15 , further comprising a respective DOP to capture, in the self-triggered mode of operation, the input data and to create a trigger for downstream DOPS.

17. A communication system comprising:

a switch configured to receive input data at one or more input ports and to transmit output data from one or more output ports;

a pipeline comprising a plurality of stages configured to process the input data and to provide the output data;

a plurality of DOPs embedded in the plurality of stages;

a DOP memory provided at an end of the pipeline, wherein the plurality of DOPs are configured to allow shifting data through the plurality of DOPs to the DOP memory; and

a finite state machine (FSM) configured to perform DOP collection by triggering reading and transferring data stored in the plurality of DOPs to the DOP memory, wherein transferring the data comprises shifting the data through the plurality of DOPs to the DOP memory,

wherein the FSM is further configured to set a done-bit when the data from the plurality of DOPs has been transferred to the DOP memory,

wherein the done-bit is detectable by software and the transferred data is accessible by the software when the done-bit is set, and

wherein the DOP of the plurality DOPs comprises a flop-based storage and is identifiable by a unique DOP-identification (ID) associated with that DOP.

18. The system of claim 17 , wherein:

a stage of plurality of stages includes one or more embedded DOPs,

at least one DOP of the plurality of DOPs is configured to latch data in response to the triggering by the FSM,

the latched data is readable by using a processor command, and

the processor command is addressable to a DOP-ID associated with a DOP of the plurality DOPs that is enabled to respond to the processor command by making the latched data accessible to a processor bus.