Hello,


I encountered what appears to be a bug in UHD that I have not been able to 
reproduce. I am using a USRP x310, UHD 3.13 on a RHEL 7 OS. The issue appears 
to be an interaction with a bad network interface.



The hang:

After I call multi_usrp::make(<my_ip_addr>), it triggers a series of print 
statements in UHD. The output looks something like this:

[INFO][UHD] linux; GNU C++ version 4.8.5 (Red Hat 4.8.5-39); Boost_105300; 
UHD_3.13.0.3-0-unknown

[INFO][X300] x300 initialization sequence...

[INFO][X300] Maximum frame size: 8000 bytes.

And at this point it hangs indefinitely. I tried pinging the USRP, and it 
responds to pings. I tried running uhd_usrp_probe, and it hangs with the exact 
same printout as my application. Eventually for an unrelated reason, I ran 
dhclient, and the hanging uhd_usrp_probe in another program unfroze and 
completed successfully.



I have an x310 connected directly to a host PC's NIC (no switch in between). 
The application used a hard coded IP address to connect to the USRP. In theory 
my setup does not have a DNS server or DHCP client.  The reason I ran dhclient 
was to bring up a bad network interface when I reconnected the field setup to 
my lab LAN.



Is this a known behavior? It appears that UHD tried to scan through the 
available network configs even though it should use the static IP to connect to 
the USRP.



The kicker is that I haven't been able to replicate the problem. It persisted 
for two days with the field configuration (through reboot cycles), and then 
appears to have been resolved by running dhclient. I tried forcing a bad 
network config in the same way, but can't repeat it.



Thanks,

Richard



_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to