Separate packet control, transport capacity and downstream service.
Try these before opening the explanations. This is a self-assessment, not a stored score or certification.
No. The NAK carries no accepted payload and does not advance the expected toggle. This bulk handshake must not be used as an isochronous retry model.
It does not fit that stated planning bound. Even a fitting average budget cannot exclude burst overflow or a long downstream memory-service pause. The reserve is an exercise assumption, not a universal MIPI efficiency.
Ready to continue when you can produce:
USB transfer type/speed and accepted data, not merely token attempts.
MAC/pixel payload accounting plus packet overhead and finite-buffer assumptions.
First define where the counter is generated and whether it covers receive frames, code/block errors or another layer. Correlate lane/PCS status, framing, clocks, resets, configuration, traffic and physical measurements; one CRC/FCS counter does not uniquely locate the corruption.
Common trap: Declaring an SI or driver root cause from one counter.
Sweep supported settings at stated traffic and operating corners while recording BER/error counts and margin. The response can be multi-dimensional and need not form a symmetric U-shaped curve, so do not assume the geometric centre or maximum setting is optimal.
Common trap: Inferring a universal curve shape from one one-dimensional sweep.