[USRP-users] USRP B210 RX Overflow

Younes KHADRAOUI younes.khadraoui at telecom-bretagne.eu
Tue Apr 28 13:17:26 EDT 2015


Hello Martin, 

Actually, there is a mistake in the title. It is more an underflow according to the software developers. The PC is not fast enough and the USRP card finds the RX thread busy. 

Thank you, 
Younes

----- Mail original -----
De: "Martin Braun" <martin.braun at ettus.com>
À: "Younes KHADRAOUI" <younes.khadraoui at telecom-bretagne.eu>, "Marcus Müller" <marcus.mueller at ettus.com>
Cc: usrp-users at lists.ettus.com
Envoyé: Mardi 28 Avril 2015 19:07:22
Objet: Re: [USRP-users] USRP B210 RX Overflow

To add to what Marcus said,

UHD doesn't crash when it receives an overrun, it just prints an 'O'. If
your application is crashing, it's likely not because of the overrun.

Cheers,
Martin

On 28.04.2015 07:11, Younes KHADRAOUI via USRP-users wrote:
> Thank you Marcus, 
> 
> Actually, I'm using a plateform already tested by its developers (OpenAir Interface). I was in contact with the developers which ran the plateform with a quite similar configuration and it works without errors. Maybe the problem really comes from the USB controller. 
> 
> If anyone can confirm that no issue was detected using the following USB controller:
> 
> 00:14.0 USB controller: Intel Corporation Wellsburg USB xHCI Host Controller (rev 05)
> 00:1a.0 USB controller: Intel Corporation Wellsburg USB Enhanced Host Controller #2 (rev 05)
> 00:1d.0 USB controller: Intel Corporation Wellsburg USB Enhanced Host Controller #1 (rev 05)
> 
> 
> Thank you again for your help,
> Younes
> 
> ----- Mail original -----
> De: "Marcus Müller" <marcus.mueller at ettus.com>
> À: "Younes KHADRAOUI" <younes.khadraoui at telecom-bretagne.eu>
> Cc: usrp-users at lists.ettus.com
> Envoyé: Mardi 28 Avril 2015 16:01:51
> Objet: Re: [USRP-users] USRP B210 RX Overflow
> 
> Ok, I've skimmed through the output of the benchmark to superficially:
> The benchmark runs totally ok, so I guess the application you're using
> is either blocking for some reason, or really just saturating your CPU.
> So you'd have to look into your application, find out where it calls
> recv()/send(), and how much CPU it's consuming.
> 
> Greetings,
> Marcus
> 
> On 04/28/2015 03:53 PM, Younes KHADRAOUI wrote:
>> Marcus, 
>>
>> I tried also the USB 2 ports, this time the overflow occurs quicker :s
>>
>> Regards,
>> Younes
>>
>> ----- Mail original -----
>> De: "Marcus Müller" <marcus.mueller at ettus.com>
>> À: "Younes KHADRAOUI" <younes.khadraoui at telecom-bretagne.eu>
>> Cc: usrp-users at lists.ettus.com
>> Envoyé: Mardi 28 Avril 2015 15:50:01
>> Objet: Re: [USRP-users] USRP B210 RX Overflow
>>
>> Hi Younes,
>>
>> I don't know the Wellsburg chipset; but this means you have two USB2
>> controllers and one USB3 controller. Can you try one of the black USB ports?
>>
>> Greetings,
>> Marcus
>>
>> On 04/28/2015 03:46 PM, Younes KHADRAOUI wrote:
>>> Hello Marcus, 
>>>
>>> Thank you for your answer. Here is the result of lspci|grep -i usb:
>>>
>>> 00:14.0 USB controller: Intel Corporation Wellsburg USB xHCI Host Controller (rev 05)
>>> 00:1a.0 USB controller: Intel Corporation Wellsburg USB Enhanced Host Controller #2 (rev 05)
>>> 00:1d.0 USB controller: Intel Corporation Wellsburg USB Enhanced Host Controller #1 (rev 05)
>>>
>>> It does not look like the ones cited bu Ettus ! what do you think ? 
>>>
>>> Regards,
>>>
>>> ----- Mail original -----
>>> De: "Marcus Müller via USRP-users" <usrp-users at lists.ettus.com>
>>> À: usrp-users at lists.ettus.com
>>> Envoyé: Mardi 28 Avril 2015 15:38:30
>>> Objet: Re: [USRP-users] USRP B210 RX Overflow
>>>
>>> Hi Younes,
>>>
>>> that is actually a rate that you should be able to do with USB2.
>>> Thus, it's probably really your USB3 controller behaving massively
>>> erratically -- you can find out what model it is by executing
>>>
>>> lspci|grep -i usb
>>>
>>> Greetings,
>>> Marcus
>>>
>>> On 04/28/2015 12:46 PM, Younes KHADRAOUI via USRP-users wrote:
>>>> Hello usrp community, 
>>>>
>>>> I'm trying to run some software with a USRP B210 card. I'm using a Dell T5810 with a Processor Intel Xeon E5-1650 v3 (3,5GHz, six cores, 15Mo cache memory, Turbo, HT). The USRP is connected to a USB 3.0 port. I'm using the last UHD release (UHD_003.009.git-144-g407e3086). I also disabled all power management features in the BIOS and in the kernel.
>>>>
>>>> The software runs correctly for a while (few seconds) but the UHD does not reply fast enough and the software crashes and prints a "RX overflow" message. Here is the benchmark:
>>>>
>>>> $ sudo /usr/lib64/uhd/examples/benchmark_rate --tx_rate 7.68e6 --rx_rate 7.68e6 --args "master_clock_rate=30.72e6"
>>>>
>>>> linux; GNU C++ version 4.8.2; Boost_105400; UHD_003.009.git-144-g407e3086
>>>>
>>>>
>>>> Creating the usrp device with: master_clock_rate=30.72e6...
>>>> -- Operating over USB 3.
>>>> -- Initialize CODEC control...
>>>> -- Initialize Radio control...
>>>> -- Performing register loopback test... pass
>>>> -- Performing register loopback test... pass
>>>> -- Performing CODEC loopback test... pass
>>>> -- Performing CODEC loopback test... pass
>>>> -- Asking for clock rate 30.720000 MHz... 
>>>> -- Actually got clock rate 30.720000 MHz.
>>>> -- Performing timer loopback test... pass
>>>> -- Performing timer loopback test... pass
>>>>
>>>> UHD Warning:
>>>>     The requested decimation is odd; the user should expect CIC rolloff.
>>>>     Select an even decimation to ensure that a halfband filter is enabled.
>>>>     decimation = dsp_rate/samp_rate -> 123 = (30.720000 MHz)/(0.250000 MHz)
>>>>
>>>> UHD Warning:
>>>>     The requested interpolation is odd; the user should expect CIC rolloff.
>>>>     Select an even interpolation to ensure that a halfband filter is enabled.
>>>>     interpolation = dsp_rate/samp_rate -> 123 = (30.720000 MHz)/(0.250000 MHz)
>>>>
>>>> UHD Warning:
>>>>     The requested decimation is odd; the user should expect CIC rolloff.
>>>>     Select an even decimation to ensure that a halfband filter is enabled.
>>>>     decimation = dsp_rate/samp_rate -> 123 = (30.720000 MHz)/(0.250000 MHz)
>>>>
>>>> UHD Warning:
>>>>     The requested interpolation is odd; the user should expect CIC rolloff.
>>>>     Select an even interpolation to ensure that a halfband filter is enabled.
>>>>     interpolation = dsp_rate/samp_rate -> 123 = (30.720000 MHz)/(0.250000 MHz)
>>>> Using Device: Single USRP:
>>>>   Device: B-Series Device
>>>>   Mboard 0: B210
>>>>   RX Channel: 0
>>>>     RX DSP: 0
>>>>     RX Dboard: A
>>>>     RX Subdev: FE-RX2
>>>>   RX Channel: 1
>>>>     RX DSP: 1
>>>>     RX Dboard: A
>>>>     RX Subdev: FE-RX1
>>>>   TX Channel: 0
>>>>     TX DSP: 0
>>>>     TX Dboard: A
>>>>     TX Subdev: FE-TX2
>>>>   TX Channel: 1
>>>>     TX DSP: 1
>>>>     TX Dboard: A
>>>>     TX Subdev: FE-TX1
>>>>
>>>> Testing receive rate 7.680000 Msps on 1 channels
>>>> Testing transmit rate 7.680000 Msps on 1 channels
>>>>
>>>> Benchmark rate summary:
>>>>   Num received samples:    76801256
>>>>   Num dropped samples:     0
>>>>   Num overflows detected:  0
>>>>   Num transmitted samples: 76866664
>>>>   Num sequence errors:     0
>>>>   Num underflows detected: 0
>>>>
>>>>
>>>> Done!
>>>>
>>>> I also checked that my usb host controller is not one listed here http://www.ettus.com/kb/detail/usrp-b200-and-b210-usb-30-streaming-rate-benchmarks. Here is the result of lsusb: 
>>>>
>>>> Bus 002 Device 002: ID 8087:8002 Intel Corp. 
>>>> Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
>>>> Bus 001 Device 002: ID 8087:800a Intel Corp. 
>>>> Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
>>>> Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
>>>> Bus 003 Device 003: ID 413c:2107 Dell Computer Corp. 
>>>> Bus 003 Device 002: ID 046d:c077 Logitech, Inc. 
>>>> Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
>>>>
>>>> As the software is running, I also checked the CPU frequency and they were all at the maximum. 
>>>>
>>>> I'm stuck here for several weeks now. Any suggestion about what can be the reason of the overflow ?
>>>>
>>>> Thank you in advance,
>>>>
>>> _______________________________________________
>>> USRP-users mailing list
>>> USRP-users at lists.ettus.com
>>> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>>>
>>
> 
> 


-- 
Younes Khadraoui
PhD Student, Network Security and Multimedia Department.
IRISA/Telecom Bretagne.




More information about the USRP-users mailing list