ndcucfg reboot fails occasionally

mbazdar · Nov 5th 2020

Hello Support Team,

From time to time, my efusA9X board gets stuck when I execute ndcucfg reboot. When this happens, on debug UART I have this output:

!NWUI: RefreshAvailableNetwork failed 0x80342002

!NWUI: RefreshWiFiInterface failed 0x80342002

NETUIQC: NWQueryInterfaceEx failed 0x80342002

[SDMAC] Offload: Setting device to pnp power state D3

NETUIQC: Error refreshing wireless info

After this I have no access to the board over ethernet and I have to reset it or restart power in order to recover it.

My kernel is V1.40 and ndcucfg is V69.

Thanks in advance for checking this out.

Best regards,

Miodrag Bazdar

mbazdar · Nov 5th 2020

Hello,

I've got different output on debug UART this time:

NDCUCFG V 69 started.Platform: efusA9X

!NWUI: RefreshAvailableNetwork failed 0x80342002

!NWUI: RefreshWiFiInterface failed 0x80342002

NETUIQC: NWQueryInterfaceEx failed 0x8034[SDMAC] Offload: Setting device to pnp power state D3 2002

NETUIQC: Error refreshing wireless infoException 'Raised Exception' (0xe06d7363): Thread-Id=04d90052(pth=a30920e8), Proc-Id=04d60052(pprc=a3091e40) 'winvnc.exe', VM-active=04d60052(pprc=a3091e40) 'winvnc.exe' PC=400510c4(coredll.dll+0x000410c4) RA=80257648(kernel.dll+0x0000e648) SP=0008e57c, BVA=0008e7dc

Best regards,

Miodrag Bazdar

fs-support_ZU · Nov 6th 2020

Hello,

in second boot log exceptions comes from winvnc maybe you start this tool to early?

mbazdar · Nov 6th 2020

Hello,

No, I don't think so. This winvnc exception is raised when VNC connection is made. Winvnc tool itself is started on start up. Here is how it happens: system is started. I copy NK.bin using FTP to \SDCard. Telnet to the device and start ndcucfg. Run reboot from ndcucfg. Sometimes I get output from above on debug UART, and machine seems to be frozen. I cannot access it anymore over network and it does not reboot. In the first example, there is no winvnc exception.

Best regards,

Miodrag Bazdar

mbazdar · Nov 9th 2020

Hello,

I've run into this problem again. Here is the log from debug UART this time:

NDCUCFG V 69 started. Platform: efusA9X

!NWUI: RefreshAvailableNetwork failed 0x80342002

!NWUI: RefreshWiFiInterface failed 0x80342002

NETUIQC: NWQueryInterfaceEx failed 0x80342002

NETUIQC: Error refreshing wireless info[SDMAC] TXRX: Pause Tx Schedular queues except managment queue as we are going detach this peer

[SDMAC] Offload: Setting device to pnp power state D3

!NWUI: RefreshAvailableNetwork failed 0x80342002

!NWUI: RefreshWiFiInterface failed 0x80342002

NETUIQC: NWQueryInterfaceEx failed 0x80342002

After running ndcucfg reboot, I've lost connection over ETHNETA1 and WiFi, but I still could connect using fixed IP address on ETHNETB1 so I might be able to provide you with logs and info necessary to debug this issue (if you tell me what could be useful for you).

PS The issue occurs relatively often, but I still haven't found the way to reproduce it each time.

Best regards,

Miodrag Bazdar

mbazdar · Nov 9th 2020

Hello,

New info: when ndcucfg reboot fails, I connect to the device using ETHNETB1 and execute IOCTL_HAL_REBOOT from command line application I have written. System reboots immediately!

So, my impression: for sure there is a problem with ndcucfg. It seems this problem leads to network devices using DHCP loose their connections. This seems to be related with the issue about losing IP connection I have also reported.

Best regards,

Miodrag Bazdar

fs-support_ZU · Nov 10th 2020

Hello,

>> New info: when ndcucfg reboot fails, I connect to the device using ETHNETB1 and execute IOCTL_HAL_REBOOT from command line application I have written. System reboots immediately!

<< Ok, but I await this also when you call NDCUCfg->reboot. Because reboot fails only sometimes as you said. NDCUCfg uses SetSystemPowerState(_T("reboot"), 0, 0) for trigger the reboot. The advantage of this call is that windows moduls got notifications before IOCTL_HAL_REBOOT is executed.

>> So, my impression: for sure there is a problem with ndcucfg. It seems this problem leads to network devices using DHCP loose their connections. This seems to be related with the issue about losing IP connection I have also reported.

<< You may be right. But I await it is a timing problem while (driver) initialization.

PS: when i understood you right: NDCUCfg->reboot problem means, that after reboot sometimes ETHNETA1 and WLAN are not available?

So i think i can write a test program which checks this in a endless loop!

mbazdar · Nov 10th 2020

Hello,

>> PS: when i understood you right: NDCUCfg->reboot problem means, that after reboot sometimes ETHNETA1 and WLAN are not available?

<< When NDCUCfg->reboot problem happens, ETHNETA1 and WLAN are present in the system, but are without IP address. I can see they are present when I connect to the device over ETHNETB1. When I execute NDCUCfg->reboot again over ETHNETB1, nothing happens. When I execute IOCTL_HAL_REBOOT over ETHNETB1 system reboots immediately, and after restart all network interfaces work again normally and system functions properly.

Best regards,

Miodrag Bazdar

fs-support_ZU · Nov 10th 2020

Hello,

thanks for this Information, then you may be right "SetSystemPowerState" may "stick" somewhere while notfy network module. My collegue just invatigate the issue with "DHCP loose their Connections" and "disable ETHNETA1".

mbazdar · Nov 10th 2020

Hello,

You are welcome. In the end, we have to be sure that losing IP on any interface is not fatal for any other interface. Moreover, any interface (WiFi or wired) should be able to recover automatically from losing connection. It may happen that all these issues: "NDCUCfg->reboot", "DHCP loose their Connections" and "disable ETHNETB1" have common source.

Best regards,

Miodrag Bazdar

Share