See Marcus Muellier's reply of earlier.
You need to configure your VM ethernet in bridge, rather than NAT
configuration. VMWare NAT re-orders UDP datagrams, which is fatal for a
number of applications.
On 2015-07-09 10:55, George Hadley wrote:
> UHD version is 3.008.004.
>
> Looking at th
On Thu, 2015-07-09 at 09:01 -0400, Philip Balister wrote:
> I'm trying not to get drug into this thread sine I am on vacation next
> week ...
>
> xu3 -> https://github.com/akuster/meta-odroid or
> https://github.com/ARM-software/meta-mali ?
>
> The ARM-software one looks interesting since it may
Yes, that’s a workaround. My point is that this issue is not obvious to
somebody who isn’t familiar with how the scheduler works.
When using null sources, it makes sense to just use two of them. But what if
you’re supplying a data stream and need to inject a synchronization sequence in
the midd
Hi Zswx,
On 07/09/2015 04:49 PM, zs wrote:
>1.How did the system know how many items I can consume in a work call?
Because it's the one who supplies these -- hence, it knows how many exist!
>
>2.Do you means that every times call the block,the block will throw
> the ninput_items[0] item
Hi Marcus:
Thank you so much.
1.How did the system know how many items I can consume in a work call?
2.Do you means that every times call the block,the block will throw the
ninput_items[0] items?
For example,the input steam 0 item is 1,2,3,4,5... and the
ninput_items[0]=2.
Hi Marcus:
Thank you so much.
1.How did the system know how many items I can consume in a work call?
2.Do you means that every times call the block,the block will throw the
ninput_items[0] items?
For example,the input steam 0 item is 1,2,3,4,5... and the
ninput_items[0]=2.
UHD version is 3.008.004.
Looking at the connection properties in the Windows Network/Sharing
Center, I see "Realtek PCIe GBE Family Controller" listed as the
ethernet controller. Looking at the network settings through the VM, the
adapter type "Intel PRO/1000 MT Desktop (82540EM) is listed.
Ah! This is something I did not know about the VMWare NAT
implementation. TCP can code, UDP just falls apart in situations where
the stack re-orders. Other applications using UDP would have similar
issues, assuming they have out-of-order detection at all, which isn't
guaranteed.
On 2015-07-09
What version of UHD are you using?
Any idea what ethernet controller is used on that dual-port card?
You'd have to go into windows configuration to determine this, since I
think that 'lspci' on the VM side will see some kind of emulated network
card.
I've taken the liberty of copying this
To clarify:
Antonny Caesar wrote in post #1176215:
> I'm trying to do it for a long time and I didn't understand the problem
> I'm having, so I came here to ask you for help.
It's not my fault, I tried a lot (using tutorials, etc) and then I came
here. But it's okay. Thank you for your efforts
I installed UHD in windows but didn't find FastDatagramThreshold.reg in
/share/uhd/FastSendDatagramThreshold.reg as suggested in
the Windows specific notes. I pulled the UHD source off github and ran
the file in question, setting the threshold to 2048. Unfortunately, this
did not have any appar
Hi Zswx,
ninput_items[0] is really the number of items that you can consume in a
work call (on input stream 0). It's usually > 1.
Best regards,
Marcus
On 07/09/2015 03:29 PM, zs wrote:
> Hi,
>
> Thank you in advance.I have a question about the "ninput_items".I
> find the meaning of it is "ve
> + Adapter enabled and attached to NAT
there we have our culprit.
NAT doesn't work very well with UHD, because for reasons obscure to my
mortal mind, VirtualBox (VMWare, also) likes to reorder packets. Now, as
a realtime device, the USRP can't handle that, and thus, UHD can't take
car
Hi,
Thank you in advance.I have a question about the "ninput_items".I find the
meaning of it is "vector of items available on all input buffers".
When I read the source code of "crc32_bb_impl.cc",I see the code below,
"
int
65 crc32_bb_impl::work (int noutput_items,
66
See the windows-specific notes here:
http://files.ettus.com/manual/page_transport.html#transport_udp_params
Does setting the FastDatagramTheshold help?
Do you two ethernet ports have distinct addresses, and do your two USRPs
have distinct addresses to match?
On 2015-07-09 09:21, George H
I am indeed running inside a VM presently. Details below:
Hardware:
USRP N210 (2x) connected to dual-port gigabit ethernet card:
http://www.newegg.com/Product/Product.aspx?Item=N82E16833166096
Software:
Windows 7 host OS
VirtualBox VM running Mint Linux 17.1 (Mate Desktop edition)
This is usually due to your network subsystem re-ordering packets, which
is never supposed to happen but certain
1GiGe interfaces, particularly USB ones, do this quite a bit.
Are you running inside a VM?
What kind of network interface do you have?
On 2015-07-09 09:09, George Hadley wrote:
Hello,
I'm working on developing some gnuradio exercises, some of
which involve using USRP N210 devices as transmitters. When I attempt to
run the related flowgraphs, I get a stream of sequence errors (S) in the
GRC log window. Attached are a pair of GRC flowgraphs in which I am
getti
I'm trying not to get drug into this thread sine I am on vacation next
week ...
xu3 -> https://github.com/akuster/meta-odroid or
https://github.com/ARM-software/meta-mali ?
The ARM-software one looks interesting since it may deal with the Mali
stuff. I'll try and get some time to build test it on
Hi Dennis,
yes, that's probably 83%. From gnuradio-runtime/lib/block_detail.cc:
297 float pfull = 1.0f -
static_cast(d_output[i]->space_available()) /
298 static_cast(d_output[i]->bufsize());
299 d_ins_output_buffers_full[i] = pfull;
and
393 std::vector
394 block_d
I think it's the fact that you are feeding two different ports of the mux by
one null source that is the issue. The first ports buffer fills before the
third ports buffer does. When the first is full, it prevents null source from
creating new outputs, which will cause the graph to stall on the t
21 matches
Mail list logo
