Rx block lock in E-Tile Hard IP for Ethernet

[Ajas]

Hi,

I am doing simulation using E-Tile Hard IP for Ethernet with 25G x 4channels . Tx  serial lines are looped back to Rx. 

Rx_block_lock signal is getting asserted for the first time reset is released . (i_sl_csr_rst_n, i_sl_tx_rst_n, i_sl_rx_rst_n and i_reconfig_reset together)  and also other status signals such as rx_pcs_ready, ehip_ready, tx_lanes_stable. But after that, if i_sl_tx_rst_n and i_sl_rx_rst_n are asserted and de-asserted, rx_block_lock is not getting asserted for all 4 channels. only for channel 1 it is getting asserted.

As per my understanding, if the clock, reset and serial connection is proper, rx_block_lock will be asserted, as it is IP's logic. So I am not sure what is causing the above behavior.

Please provide any suggestion to overcome this issue. any kind of setting for simulation to be done or similar.

Thanks,

Aj.

[SengKok_L_Intel]

I have performed a simulation by asserting the sl_tx_rst_n & sl_rx_rst_n signals for 10 clock cycles after the rx_block_lock signals are locked. 

Please refer to the attached simulation waveform, after triggered the sl_tx_rst_n & sl_rx_rst_n signals, the rx_block_lock becomes 0, however, after some time, it is able to lock again. 

Therefore, I don't see the problem that you encountered, perhaps, you may need to run the simulation slightly longer to see if this is able to lock. 

[Ajas]

thank you for the response.

is it with PTP configuration or only Ethernet?

[SengKok_L_Intel]

I was running with Ethernet without 1588 enabled.

[Ajas]

Thank you for the reply.

I am running with PTP configuration. As soon as the tx/rx_rstn de-asserted, local_fault signal is also asserted. Could you please suggest in which scenario in simulation, local_fault would assert, incase of loopback?

[SengKok_L_Intel]

After asserted tx/rxrstn, does the local_fault de-asserted? Or it stays asserted?

[Ajas]

It gets asserted after de-assertion of reset. and stays asserted.

[SengKok_L_Intel]

Below is what I observed in the simulation:

After asserted tx/rx rstn again, the IP will take some time to block lock, and then asserted the *_ready signals.

During this period, the signal may not stable, and the local_fault signal can be asserted. However, after the *_ready signals are asserted, the local_fault signal will stay low.

This is NOT expected if the local_fault signal always stays high after de-asserted the tx/rx rstn. 

[SengKok_L_Intel]

In your simulation, you may check the local_fault signals again after the *_ready signals are asserted.

[Ajas]

Thank you for the suggestion.

Could you please suggest any reason for local_fault to be asserted? as it stays high even after the ready signals(ehip_ready, tx_lanes_stable) are asserted?

[SengKok_L_Intel]

It could be due to the block boundary is not aligned yet. You can refer to *_pcs_ready signal.

[SengKok_L_Intel]

If further support is needed in this thread, please post a response within 15 days. After 15 days, this thread will be transitioned to community support. The community users will be able to help you with your follow-up questions.