Pitfalls and red flags

Pitfalls & Red Flags for CXL Attach Modes focuses on cache/memory latency, coherency traffic, attach-mode bandwidth. The goal is to connect the observable symptom to protocol mechanism, ownership, and regression risk.

• A channel looks idle, but upstream credit or ready behavior is the actual blocker.

• The spec allows behavior that the local scoreboard assumed was illegal.

• A bridge changes width, ID, burst shape, or ordering attributes silently.

• Reset releases one side of the interface earlier than the other.

• The fix improves a directed test but regresses real mixed traffic.

Layer responsibility check

LAYER RESPONSIBILITY — CXL Attach Modes

layer          owns                         common failure
-----------    --------------------------   -----------------------
software       intent, ordering needs       wrong assumption
transaction    id/addr/len/attributes       ordering / outstanding
link/channel   handshake, credits, retry    backpressure / deadlock
physical       clock/reset/lanes/PHY        timing / training / SI
observability  waveform/log/counter         missing evidence

Protocol deep dive

PCIe is reliable packet delivery over unreliable links; debug flows PHY -> DLL -> TLP -> firmware.

Concept diagram

PCIe DEBUG TOP-DOWN

L0 link healthy?  -> credits OK?  -> TLP completes?  -> driver happy?

Skip a layer and you will mis-own the bug.

Metric graph

LINK DEGRADE EXAMPLE

target x4 Gen4  ---- ---- ---- ----
actual   x4 Gen4  ---- ---- ---- ----   (eval board)
actual   x1 Gen3  -                   (product board)

Package/SI often shows up as width downgrade, not hard fail.

Metrics and artifacts to collect

• link width/speed

• replay count

• completion timeout

• AER error log

• LTSSM history

Mini case study

Endpoint enumerated but DMA timed out: completion credits exhausted because a switch port was misconfigured in firmware, not because the endpoint was broken.

Debug branches

• If degrade at width/speed, PHY/SI before driver.

• If replay storm, link layer before transaction layer.

• If CXL coherency bug, separate .io vs .cache vs .mem traffic.

Senior review question

Ask: what is the first transaction that deviates, and which spec rule does it test?

Key takeaways

• Connect every protocol claim to a transaction identity and measurable metric.

• Store the artifact (waveform, log, counter) next to every signoff decision.

Common pitfalls

• Debugging timeouts without finding the first bad transaction.

• Quoting peak bus width without payload efficiency and retry overhead.

• Treating VIP compliance as a substitute for system integration replay.

Principal review addendum

Re-read CXL Attach Modes against one concrete product workload, not a synthetic directed test.

CXL.io, CXL.cache, and CXL.mem layer coherent and memory semantics on PCIe infrastructure.