[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Linrad] Re: Linrad-03.06



Hello Andrea,
 
> Leif Asbrink wrote:
> 
> > BIG PROBLEM: I have not been able to find out how to get the CPU
> > load per thread under modern Linux kernels. When searching the Internet
> > I get too much information on other things. The good old program top
> > has the information and I suppose one could look at the source 
> > code of top. But I can not find that either:-(
> 
> on modern unices top reads /proc/stat and /proc/meminfo.
> More details and sources on
> 
> http://procps.sourceforge.net/

Thanks for this info:-)

Linrad-03.07 will have timing by threads when the user presses 'T'
Each thread has a name in Linrad like this:

char *thread_names[THREAD_MAX]={"RxAD",  //0 = THREAD_RX_ADINPUT
                                "Rxrn",  //1 = THREAD_RX_RAW_NETINPUT
                                "Rxfn",  //2 = THREAD_RX_FFT1_NETINPUT
                                "Rxfi",  //3 = THREAD_RX_FILE_INPUT
                                "SR14",  //4 = THREAD_SDR14_INPUT
                                "RxDA",  //5 = THREAD_RX_OUTPUT
                                "Scre",  //6 = THREAD_SCREEN
.
.
.

The full list in thrvar.c

The CPU time for a thread can become near 100% and on a multi-processor
computer several threads may have near 100% CPU load. The sum of the
thread loads may become 100% times the number of processor cores.
The cpu load shown in the lower left corner of the screen will be the
average load on the computer. It will become 100% only if Linrad
loads all the processors by 100% all the time which would be very 
unlikely on a dual core and impossible on a quad core machine. 
Depending on parameters, the thread doing FFT1, the baseband thread or
the screen thread is likely to saturate at 100% and lead to some
kind of malfunction that prevents the load of other threads to reach 100%
simultaneously.

Linrad-03.06 and earlier show a load in the lower left corner that
may go up to 100% times the number of processor cores.

The thread timings would be useful to inform users about overload
conditions and what to do about them. This is not something I will
put priority on in the near future though.

Thank you again Andrea for the information. Now I know that everything
I needed was obtainable by "man proc" I could extract that from the 
link you posted to the list. Such pieces of info can be really hard 
to find....

   73


  Leif / SM5BSZ


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Linrad" group.
To post to this group, send email to linrad@xxxxxxxxxxxxxxxx
To unsubscribe from this group, send email to linrad+unsubscribe@xxxxxxxxxxxxxxxx
For more options, visit this group at http://groups.google.com/group/linrad?hl=en
-~----------~----~----~----~------~----~------~--~---

LINRADDARNIL