What is meant by "LLLLLLLLLLLLL."?

GNU Radio
1

All,

I have an application where I need to send timed-bursts, so, I have
modifiedtags_demo.cca bit according to my needs. Now, when I run the
transmitter flow-graph in order to send timed-bursts, I occasionally see
a sequence of “LLLLLLLLL…” in the console. I assume it is some sort of
error message indicating that the time value carried by tx_time tag for
that burst has been past at the FPGA. Am I right?

Thanks,
Yu.

2

On 07/17/2012 03:50 PM, Zing Yu wrote:

All,

I have an application where I need to send timed-bursts, so, I have
modified tags_demo.cc a bit according to my needs. Now, when I run
the transmitter flow-graph in order to send timed-bursts, I
occasionally see a sequence of “LLLLLLLLL…” in the console. I
assume it is some sort of error message indicating that the time
value carried by tx_time tag for that burst has been past at the
FPGA. Am I right?

Correct, its a “late” packet. Time in packet > time on the fpga.

http://files.ettus.com/uhd_docs/doxygen/html/structuhd_1_1async__metadata__t.html#a2be1b5c0351746c78fa3bcb74a8ff5da

-josh

3

Hi Josh,

The “late” packet gets sent anyway. Right? At least I have verified this
at the receive side.Further, I hope that one L is printed for each
“late” burst (i.e., no. of L’s equals no. of late “bursts”). Correct?
Finally, how can one print the metadata to find which of the bursts were
late, assuming that there is a tag containing the burst number?

Thanks,
Yu.

From: Josh B. [email protected]
To: [email protected]
Sent: Tuesday, July 17, 2012 5:53 PM
Subject: Re: [Discuss-gnuradio] what is meant by “LLLLLLLLLLLLL…”?

On 07/17/2012 03:50 PM, Zing Yu wrote:

All,

I have an application where I need to send timed-bursts, so, I have
modified tags_demo.cc a bit according to my needs. Now, when I run
the transmitter flow-graph in order to send timed-bursts, I
occasionally see a sequence of “LLLLLLLLL…” in the console. I
assume it is some sort of error message indicating that the time
value carried by tx_time tag for that burst has been past at the
FPGA. Am I right?

Correct, its a “late” packet. Time in packet > time on the fpga.

http://files.ettus.com/uhd_docs/doxygen/html/structuhd_1_1async__metadata__t.html#a2be1b5c0351746c78fa3bcb74a8ff5da

-josh

4

On 07/17/2012 04:20 PM, Zing Yu wrote:

Hi Josh,

The “late” packet gets sent anyway. Right? At least I have verified

Late packets get dropped. Think of this as a catch-up mechanism.
Otherwise, all subsequent packets would also be late.

The late packet error follows the same policy as the underflow error:
http://files.ettus.com/uhd_docs/doxygen/html/structuhd_1_1stream__args__t.html#a4463f2eec2cc7ee70f84baacbb26e1ef

this at the receive side. Further, I hope that one L is printed for
each “late” burst (i.e., no. of L’s equals no. of late “bursts”).

Yup. That means everything transmitted 100%

Correct? Finally, how can one print the metadata to find which of the
bursts were late, assuming that there is a tag containing the burst
number?

There is time stamp on the async message.

You can retrieve this in the gnuradio environment via async message
block. Here is one implementation using the gr_message_queues:
http://gnuradio.org/cgit/gnuradio.git/tree/gr-uhd/include/gr_uhd_amsg_source.h

And there is a version in GrExtras that creates a custom PMT tag with
the async message:

-josh