why pcie1 devices are not listed in lspci

[nizhixian]

When we plug our wifi module into the S32G274ardb2 development board, lspci shows the following.
Network controller is our device

The follows are our own development board

The two figures above are our circuit diagrams.

//setenv hwconfig "pcie0:mode=pcie,clock=ext;pcie1:mode=rc&sgmii,clock=int,fmhz=100,xpcs_mode=0"
//setenv hwconfig "serdes1:mode=pcie,clock=ext,fmhz=100;pcie1:mode=rc"
//setenv hwconfig "pcie1:mode=rc&sgmii,clock=ext,fmhz=100,xpcs_mode=0"
//setenv hwconfig "pcie0:mode=pcie,clock=ext"
//setenv hwconfig "pcie1:mode=pcie,clock=ext,fmhz=100"

These are our debug log.

 We can't see the devices information in lspci -v.

please assit us in solving this issue,thank you

with best regards

[Daniel-Aguirre]

Hi,

Can you let us know with which BSP version are you getting this outcome? Also, can you help us elaborate more on regards of your problem? Can you provide us with full boot logs?

We may not be understanding fully what the problem under your setup is, we see both images showing information when using the lspci -v command, while you say that you are not getting the information, but we see information on both images. If you can help us providing more information, we can provide better support from our side. We do apologize for the inconvenience.

Please, let us know.

[nizhixian]

Our BSP version is 38.0. We want to mount our WIFI module on the PCIe slot of the S32G274a. On your official development board s32g274ardb2, after we insert our WIFI module into the PCIe0 slot, we can view our device information through the lspci -v.

However, when we insert the WIFI chip into the PCIe1 slot on our own designed development board, we cannot view the device information of our WIFI module through lspci -v. 

 The changed of the circuit diagrams was shown above

[Daniel-Aguirre]

Hi,

Thanks for your feedback. We understand that you said that you are using BSP38.0, but seeing the logs we see that you were using BSP35.0, can you confirm which version you are currently using?

For what we can also understand, under the NXP platform you cannot reproduce the behavior, hence we assume the behavior might be related to HW/SITA. As you have shown previously, there are additional capacitors to the lines. 

Is this only happening with the WiFi module? Or does any other device has the same behavior?

Please, let us know.

[nizhixian]

I'm sorry ,our information expression was wrong.Our BSP version is 35.0.
Could you tell me what HW/SITA stand for? And regarding the issue of capacitors, could you please provide more helpful information?
And in our new development board the pcie0 is working properly.However, we are not sure if the problem of pcie1 is with the hardware of software.Please give us more help，thank you
With Best regards.

[Daniel-Aguirre]

Hi,

Thanks for your feedback. HW refers to Hardware and SITA refers to Signal Integrity and Time Analysis, which at the end we are mentioning that there could be a problem with your board, since you are able to use our board no problem.

As for more information, since your board is custom, through this channel is not common for us to help on these situations. For that, we can recommend contacting your local NXP FAE/representative, since they should be able to support your custom build. We do apologize.

The general recommendations from our internal team are the following:

"In general, if the link is unstable, the problem may well be caused by bad hardware connection. So, customer may have a careful check on the hardware connections. Then, on the same board, change the hwconfig settings and let only SerDes_0 work in PCIe mode and SerDes_1 in SGMII mode, check if the WiFi card could be stably probed by S32G PCIe_0 or not.

As far as I know, the new BSP version, for example BSP36 or later, enhanced a few features of PCIe and also updated part of the PCIe drivers. Therefore, if the problem remains, please use the newer BSP version (BSP36 or later) to do the same test, and check if the problem disappears or not."

Please, let us know.

[nizhixian]

My debugging records are now upoloaded as attachments, and you can see that the PCIe1 always failed to link up.