[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [linrad] VESA; and a question
- Subject: RE: [linrad] VESA; and a question
- From: Leif Asbrink <leif.asbrink@xxxxxxxxxxxxxxxxxxxx
- Date: Wed, 14 Apr 2004 03:27:04 +0200
> I tried linrad 18 last night (with svgalib - 1.4.3) and did get "speed
> warning" notices as I had with version 17 but not as frequently. I have
> Red Hat 9 on a 2.7 GHz Dell PC.
OK. Presumably your CPU usage is well below 90% so you should not
get any speed warnings.
Most probably you have one of my mistakes.
The speed warning (in Linrad-01.18) is issued if the main loop
has not reached the idle loop at least every 0.5 second with
all flags indicating there is something to do set to zero.
(It was 1.5 seconds in 00-17)
If the time exceeds 2 seconds you will see a SPEED ERROR in
01-18 (it was 10 seconds in 01-17) Very long times means that
Linrad most probably skipped something it was suppused to do.
(Probably some graph updata)
If you get a speed warning when clicking on a new signal, it
is not an indication of a problem if you use AFC and have a long
time delay. Linrad recomputes everyting back in time to present you
with the output from the new station immediately and it could easily
take 0.5 seconds if you have 10 seconds of data in the buffers.
As a first step I would like to check the timing on my system
with the exact parameters that you use. Please copy all files
par* and dsp* into a subdirectory and send them as a tar.bz2
archive. Also run Linrad, press "F" and "T". Then select a signal
and wait 10 seconds. Finally press "G" to generate a screen dump.
Send the gif file also. Maybe I can find out something.
> I also did get a frozen screen in one of the modes (unfortunately I
> don't remember which mode). It was frozen in the sense that I could not
> select another frequency. "ESC" took me out of linrad and I could
> restart it.
Please, and this is for everyone on this list running the latest
version. Please try to find out exactly how you can reproduce it
if you run into an obvious bug like this. It is not difficult for
me to find a bug that I can reproduce here and it is a reasonable
task to find a problem that you can reproduce (I would send various
modified packages that write debug files), but just knowin there
is an error without any idea where it is is a bit frustraring....
In case the thing you observed was that you could select a signal
but that there was no audio output. (The signal was visible in the
baseband window) there was an error in the sound drivers. These
problems occur sometimes buit it should not be often.
Try to change the parameter da stop/start (close and reopen D/A
when a new signal is selected.) Close and reopen should solve
problems in the D/A driver, but it can cause problems for the
A/D because it switches off the interrupt for too long sometimes.
Leif / SM5BSZ