[USRP-users] B210 USB transfer error

Marcus D. Leech mleech at ripnet.com
Wed Oct 1 14:16:20 EDT 2014


On 10/01/2014 01:08 PM, Garver, Paul W wrote:
> I'm using bus power since I, like the OP, had the same issue and 
> external power didn't make a difference in the initial 
> troubleshooting.  I think the drives could make a difference, 
> especially since the 2.5" and mSSD/mSATA are on different rails. Say 
> the USB 3.0 controller on the motherboard is running on the 3.3VDC 
> rail and the Samsung 840 is pulling enough current to max out the spec 
> on that rail.  But the 2.5" would be fine since it uses the +5VDC 
> rail. Unfortunately, without the spec (which we won't get from 
> Gigabyte) we won't know.  This could happen regardless of how the B200 
> is powered.
>
> For clarification, I'm using the B200 not B210.
>
> Just a thought -- thanks for your help and I hope we can resolve this 
> issue soon!
>
> PWG
The fact that different hard-disk hardware makes a difference indicates 
to me that this is a "systems" issue, rather than one relating directly
   to the B2xx hardware.

I can't remember whether you forwarded an lscpi -vv listing to the 
list.  That can help identify what the different bits of hardware might 
be that
   could be involved.


> ------------------------------------------------------------------------
> *From: *mleech at ripnet.com
> *To: *"Paul W Garver" <garverp at gatech.edu>
> *Sent: *Wednesday, October 1, 2014 12:53:27 PM
> *Subject: *Re: [USRP-users] B210 USB transfer error
>
> Not sure why/how changing disk drives can affect the performance of 
> the USB subsystem, particularly if you're using the external
>
>   power supply on the B210.
>
> Unless the machine has some kind of SATA-to-USB bridge internally, and 
> that controller is shared with the B210.
>
> On 2014-10-01 12:46, Garver, Paul W wrote:
>
>     I saw your comment about the issue not affecting the maint branch
>     and tried it.  Initially, it worked writing ~100GB on a Sandisk
>     Extreme II 2.5" SSD (successfully did this 4x), but when I
>     switched to a Samsung 840 Evo 1TB mSSD, I saw the issue again,
>     almost immediately.  This time, though, libusb was reporting
>     BAD_PACKET  (1).
>     Built maint branch commit id:
>     8995533832f685c4fae388c52e52b06c25a82a2b
>     UHD Lib reports: linux; GNU C++ version 4.8.2; Boost_105400;
>     UHD_003.007.002-0-g89955338
>     The Samsung 840 Evo power spec is 1.8A @ 3.3VDC.  I wonder if that
>     is an issue -- unfortunately on these small form factor PCs I am
>     using (Gigabyte BRIX Pro), there is no spec and the vendors are
>     usually super secretive about giving you any real technical info.
>      It also may be relevant to note that I believe the 2.5" SSDs are
>     using the +5VDC rail, but the mSATAs use +3.3VDC.
>     PWG
>
>     ------------------------------------------------------------------------
>     *From: *"Marcus D. Leech" <mleech at ripnet.com>
>     *To: *"Paul W Garver" <garverp at gatech.edu>
>     *Cc: *usrp-users at lists.ettus.com
>     *Sent: *Monday, September 22, 2014 12:46:01 PM
>     *Subject: *Re: [USRP-users] B210 USB transfer error
>
>     On 09/22/2014 12:44 PM, Garver, Paul W wrote:
>
>         I've updated to UHD 3.7.2 with the 3.14.0 kernel.  I installed
>         from source and downloaded the 3.7.2 images as well.  After
>         writing ~130GB @ 25MSPS I received the same USB transfer
>         status 5 error.  Also, you asked about sample rate  -- I tried
>         a few different values and didn't notice any difference.  I
>         haven't tried USB2.0 - I can do this for debugging purposes
>         but it will not be a solution for me as I my application
>         requires 25MSPS.  I really don't have a good idea if this is a
>         hardware or software issue.
>         Thanks,
>         PWG
>
>     Thanks for the experiments, Paul.
>
>     Ettus R&D is currently working this issue.
>
>
>         ------------------------------------------------------------------------
>         *From: *"Marcus D. Leech" <mleech at ripnet.com>
>         *To: *"Paul W Garver" <garverp at gatech.edu>
>         *Cc: *"Tony Arkles" <t.arkles at gmail.com>,
>         usrp-users at lists.ettus.com
>         *Sent: *Friday, September 19, 2014 5:04:55 PM
>         *Subject: *Re: [USRP-users] B210 USB transfer error
>
>             I tried the 3.14.0 kernel on Ubuntu 14.04 for the BRIX
>             (UHD 3.7.1) and still continue to see the same error
>             message as the OP.
>             PWG
>
>         Could you also update to UHD 3.7.2 ?
>
>
>
>             ------------------------------------------------------------------------
>             *From: *"Tony Arkles" <t.arkles at gmail.com>
>             *To: *mleech at ripnet.com
>             *Cc: *"Paul W Garver" <garverp at gatech.edu>,
>             usrp-users at lists.ettus.com
>             *Sent: *Friday, September 19, 2014 2:43:30 PM
>             *Subject: *Re: [USRP-users] B210 USB transfer error
>
>             I was having a similar problem on Ubuntu 14.04. Exact same
>             error message. It wasn't really sample rate dependent, I
>             was seeing it at 4MS/s, which worked fine over usb2.
>             I ended up upgrading to a 3.14 kernel (using ubuntu .debs
>             from a newer version) and it cleared up. There were notes
>             in the change log for the kernel saying that there were
>             fixes for the Intel USB3 chipset.
>
>             On Sep 19, 2014, at 11:45 AM, "Marcus D. Leech via
>             USRP-users" <usrp-users at lists.ettus.com
>             <mailto:usrp-users at lists.ettus.com>> wrote:
>
>                 I'd like to understand if this is sample-rate
>                 dependent.  Does it happen more frequently at higher
>                 samples rates, or is it just dependent on
>
>                   runtime?  What about when on USB-2.0?
>
>                 A goodly chunk of the R&D team at Ettus are on the
>                 road until Monday.
>
>                 On 2014-09-19 13:29, Garver, Paul W via USRP-users wrote:
>
>                     I can also report I have had this exact same issue with a B200, although I typically run rx_samples_to_file.  I've also tried external power to no avail.  I've also run the B200 on the haswell Intel NUC which also has the series 8/C210 USB 3.0 controller.  I haven't seen this issue yet on the NUC but I don't have a lot of runtime on it yet.
>
>                     Could you please post your results to the mailing list or can Ettus support contact me offline?  I'm very interested in the resolution to this issue.
>
>                     System Info:
>                     Gigabyte Brix GB-BXi7-4770R
>                     Ubuntu 14.04 64 bit
>                     Intel Corporation 8 Series/C220 Series Chipset Family USB xHCI (rev 05)
>
>                     PWG
>
>                     Date: Fri, 19 Sep 2014 02:08:04 +0000
>                     From: "Nowlan, Sean" <Sean.Nowlan at gtri.gatech.edu  <mailto:Sean.Nowlan at gtri.gatech.edu>>
>                     To: "USRP-users at lists.ettus.com  <mailto:USRP-users at lists.ettus.com>" <USRP-users at lists.ettus.com  <mailto:USRP-users at lists.ettus.com>>
>                     Subject: [USRP-users] B210 USB transfer error
>                     Message-ID:
>                     	<625d92335cfe4d91923faf730b3e2a97 at apatlisdmail01.core.gtri.org  <mailto:625d92335cfe4d91923faf730b3e2a97 at apatlisdmail01.core.gtri.org>>
>                     Content-Type: text/plain; charset="us-ascii"
>
>                     Hi all -
>
>
>                     I've seen this pop up now and again on this list. Basically while running an RX operation (osmocom_fft or uhd_fft) the B210 will sporadically disconnect from the USB 3.0 hub and then it will immediately reconnect. I can see this in dmesg and /var/log/syslog. It was suggested to me that this may be a power brown out issue since I have a TCXO GPSDO module mounted on board, but I observe the same error even while connected to wall power through the standard 6V, 3A power supply. Here is the console output I see.
>
>
>                     UHD Error:
>                          The receive packet handler caught an exception.
>                          RuntimeError: usb rx6 transfer status: 5
>                     UHD source block got error code 0xf
>                     terminate called after throwing an instance of 'uhd::runtime_error'
>                        what():  RuntimeError: usb tx4 submit failed: LIBUSB_ERROR_NO_DEVICE
>                     Aborted (core dumped)
>
>                     System info:
>                     Intel(R) Core(TM) i5-3210M CPU @ 2.50GHz
>                     USB controller: Intel Corporation 7 Series/C210 Series Chipset Family USB xHCI Host Controller (rev 04)
>                     Ubuntu 14.04 LTS
>                     linux; GNU C++ version 4.8.2; Boost_105400; UHD_003.007.002-94-ge56809a0?
>
>                     Sean
>
>                     _______________________________________________
>                     USRP-users mailing list
>                     USRP-users at lists.ettus.com  <mailto:USRP-users at lists.ettus.com>
>                     http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>
>                 _______________________________________________
>                 USRP-users mailing list
>                 USRP-users at lists.ettus.com
>                 <mailto:USRP-users at lists.ettus.com>
>                 http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>
>
>
>         -- 
>         Marcus Leech
>         Principal Investigator
>         Shirleys Bay Radio Astronomy Consortium
>         http://www.sbrac.org
>
>
>
>     -- 
>     Marcus Leech
>     Principal Investigator
>     Shirleys Bay Radio Astronomy Consortium
>     http://www.sbrac.org
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/attachments/20141001/cba0b0af/attachment-0002.html>


More information about the USRP-users mailing list