Re: [Discuss-gnuradio] Interesting observation regarding varying time duration it takes to execute fg.start() [a flowgraph start ala thread spawning??]

[Bishal Thapa]

Yes Matt, you were correct. Now the range between min-time recorded and max-time recorded is much less. Min = 309us (still), Max = 2349us. So range ~= 2ms.

I believe it is due to Eric's reasoning, multiple services running on my host machine, although I have to mention I am running these programs as "sudo". Anyways, I will try to run it on a different host with nothing much going on, and see if I can reduce the range variation of 2ms.

Thank you Eric and Matt for your time,
Sincerely,

On Mon, Mar 8, 2010 at 10:07 AM, Matt Ettus <address@hidden> wrote:

On 03/07/2010 01:12 PM, Eric Blossom wrote:

On Sun, Mar 07, 2010 at 04:07:58PM -0500, Bishal Thapa wrote:

Thank you for your reply Eric.
  I did print delta and delta.microseconds and guess what I did get this:

Receiver (0) =  BABA0002
Requested RX Bitrate: 100k
Probe Level 39
0:00:00.000327  [*This is from printing delta*]
327  [*This is from printing delta.microseconds*]

Why do you think I should not get 309us? I am sorry if I don't get something
fundamental here. Thanks again for helping me out :)

No problem.  309us may be reasonable.

In general the wide variance that you're seeeing could be caused by
pretty much any other activity on the machine.  You are measuring
"wall time".

Does this flowgraph have a USRP1?  If so, the very long time could be when you have first powered up the USRP1 and so it needs to have its FPGA image downloaded.

Matt