Method and apparatus for a configurable packet routing, buffering and scheduling scheme to optimize throughput with deadlock prevention in SRIO-to-PCIe bridges
View Patent ↗A method and apparatus for a configurable packet routing, buffering and scheduling scheme to optimize throughput with deadlock prevention in SRIO-to-PCIe Bridges have been described. A single level enqueue method with dynamic buffering and dequeuing based on packet re-ordering is disclosed. Single level packet routing and scheduling to meet SRIO and PCIe rules to enqueue packets based on FType/TType is disclosed. Backpressure based on ingress watermarks for different packet types is disclosed. Use of a circular-reorder queue (CRQ) for both ingress and egress allows packet reordering and packet passing.
1. An apparatus comprising:
a serializer/deserializer (SERDES) interface having an input coupled to receive packets;
an ingress physical layer module (PLMi) directly coupled to said SERDES for receiving said packets;
an ingress transport layer module (TLMi) having a port mapping block, a routing block, a port width conversion (PWC) block, and transport processing block, said TLMi directly coupled to said PLMi;
an ingress packet buffer module (PBMi) having an enqueue block, a plurality of virtual output queues (VOQs), a plurality of data memories, and a dequeue block, said PBMi directly coupled to said TLMi; and
said dequeue block directly coupled to a circular-reorder queue (CRQ) said CRQ implemented in hardware and having a last free pointer to an open location in said CRQ, a next to send pointer in said CRQ a first free location pointer in said CRQ, a last packet enqueued pointer in said CRQ, and wherein said CRQ has an acknowledgement identification for each of said received packets.
2. The apparatus of claim 1 further comprising:
an endpoint fabric module (EFM) having a scheduler, a switch, an arbiter, and a port select, said EFM operatively coupled to said PBMi.
3. The apparatus of claim 2 further comprising:
an ingress user core gasket, said ingress user core gasket operatively coupled to said EFM;
an egress user core gasket, said egress user core gasket operatively coupled to said EFM;
a user core, said user core operatively coupled to said ingress user core gasket; and
said user core operatively coupled to said egress user core gasket.
4. The apparatus of claim 3 further comprising:
an ingress logical layer module (LLMi), said LLMi operatively coupled to said EFM, and said LLMi operatively coupled to said ingress user core gasket, and said LLMi operatively coupled to said egress user core gasket; and
an egress logical layer module (LLMe), said LLMe operatively coupled to said EFM, and said LLMe operatively coupled to said ingress user core gasket, and said LLMe operatively coupled to said egress user core gasket.
5. The apparatus of claim 4 further comprising:
an egress packet buffer module (PBMe), said PBMe having an enqueue block, a plurality of data memories, a plurality of said circular-reorder queues (CRQ), and a dequeue block, said PBMe directly coupled to said EFM.
6. The apparatus of claim 5 further comprising:
an egress transport layer module (TLMe) having a transport processing block, and a port width conversion (PWC) block, said TLMe operatively coupled to said PBMe.
7. The apparatus of claim 6 further comprising:
an egress physical layer module (PLMe), said PLMe operatively coupled to said TLMe;
said PLMe operatively coupled to said SERDES interface; and
said PLMe operatively coupled to said PLMi.
8. The apparatus of claim 7 wherein said PBMi said plurality of VOQs and said PBMi said plurality of data memories are user configurable based on a number of ports in use from said SERDES interface.
9. The apparatus of claim 8 wherein said PBMe said plurality of CRQs and said PBMe said plurality of data memories are user configurable based on a number of ports in use from said SERDES interface.