A10 PCIe HIP "nreset_status" signal tobe deasserted randomly, even though npor&pin_pers remains high

[JET60200]

hello expert,

We uses Arria10 PCIe HIP core to build a EP device.   In general, the FPGA works fine when plugging inside x86 SERVER PCIE Slot at most cases . 

But for a part of x86 servers, we found A10 PCIe HIP “"nreset_status"” signal will be deasserted sometimes and randomly, while at that moment PCIe HIP's " npor & pin_pers " signals keep at High. According to " ug-01145_avmm-1_0.pdf " description,  "

apps_rstn, which is derived from npor
or pin_perstn.

"

Thus one word to say , in our exceptional case,  eventhough either  "npor" or " pin_perstn " keeps  to "high",    " apps_rstn " still be deasserted to LOW (0).  I want to find why it happens ?

Appreciate any helps. 

[KhaiChein_Y_Intel]

Hi,

Could you share the signaltap with signals below? What is the speed grade and lane number used?

pll_cal_busy

tx_cal_busy[n-1:0]

rx_cal_busy[n-1:0]

pin_perst

npor

reset_status

pld_clk_inuse

pll_locked

rx_is_lockedtoref[n-1:0]

rx_is_lockedtodata[n-1:0]

rx_std_signaldetect[n-1:0]

ltssmstate[4:0]

lane_act[3:0]

currentspeed[1:0]

rx_ready[n-1:0]

tx_ready[n-1:0]

rx_analogreset[n-1:0]

rx_digitalreset[n-1:0]

tx_analogreset[n-1:0]

tx_digitalreset[n-1:0]

tx_st*

rx_st*

Thanks

Best regards,

KhaiY

[JET60200]

Thanks @KhaiChein_Y_Intel for reply,

We checked PCIe HIP part library code, and found below code. To my understanding, this "output reset_status" signal dedrived from "reset_status_hip" signal, while the "reset_status_hip" comes from :

 twentynm_hssi_gen3_x8_pcie_hip  {  .reset_status                 (reset_status_hip),   }  ;

My puzzle is : what does this "hssi_gen3_x8_pcie_hip  {  .reset_status     )"  stand for ? 

  At what condition, this "hssi_gen3_x8_pcie_hip  {  .reset_status     )"  would be  deasserted ?

ThankS a lot !!

[ ==

;;  \qpro19_2_ip\board_pcie_a10_hip_0\altera_pcie_a10_hip_191\synth\altpcie_a10_hip_pipen1b.v 

   wire                    reset_status_hip;

   output                reset_status,

   twentynm_hssi_gen3_x8_pcie_hip  # (

      <params ...> 

      ) wys  (

               ...                  

               .reset_status                 (reset_status_hip),      // where does the [ reset_status_hip] come from  ??

               ...  

               // synthesis translate_on 

               .wake_oen                     (wakeoen              )

        );

   altpcie_sc_bitsync_node #(

    ) reset_status_altpcie_sc_bitsync(

      .clk       (pld_clk) ,// clock

      .rst_n     (npor_sync) ,// async reset

      .data_in   (reset_status_hip) ,// data in

      .data_out  (reset_status) // data out

   );

]

[KhaiChein_Y_Intel]

Hi,

Does the transceiver calibration successful? Can you check if the pin_perst of the Hard IP instance connects to the corresponding nPERST pin of the device correctly? Could you share the signal tap result?

These pins have the following locations:

• NPERSTL0: bottom left Hard IP and CvP blocks

• NPERSTL1: top left Hard IP block

• NPERSTR0: bottom right Hard IP block

• NPERSTR1: top right Hard IP block

Thanks

Best regards,

KhaiY

[KhaiChein_Y_Intel]

Hi,

We do not receive any response from you to the previous question/reply/answer that I have provided. This thread will be transitioned to community support. If you have a new question, feel free to open a new thread to get the support from Intel experts. Otherwise, the community users will continue to help you on this thread. Thank you.

Best regards,

KhaiY

[MikeH1]

I am seeing what sounds like same problem.

I have created post:  A10 PCIe Device not seen by Windows - Intel Community

that includes SigTap capture.