[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Linrad] Re: Linrad-02.54
- Subject: [Linrad] Re: Linrad-02.54
- From: Leif Asbrink <sm5bsz.com; leif@xxxxxxxxxxxxxxxx>
- Date: Sat, 13 Dec 2008 23:33:33 -0400
Hi Mart and All,
> There seems to be a problem with the "total amplitude calibration"
> function in ver 2.53 and 2.54 . When you get to the last page in the
> pulse calibration sequence and all data has been calculated, at the
> Y=>Accept prompt the program hangs (win version).
And the same under Linux.....
This bug is a typical example on why it is important to me
to get bug reports as fast as possible. It is much easier
to find the reason why something that used to work no
longer does so if the time since I made a mistake is short.
I surely understand that those of you who have an old version
that works well are reluctant to take the trouble to install
a new version and possibly find that the new version creates
problems. Your trouble - if it happens - is my help!
The calibration error is caused by a change I made in order
to make Linrad more tolerant to unfriendly programs that
use the CPU heavily now and then. (For X11 and Windows users.)
The Linrad user may ask for a low DMA rate to allocate
large buffers for the input and output. A user who
also sets a large output delay margin can have a lot
of data waiting in the various buffers within Linrad
and with 2.53 the different threads (hopefully) wait for
free space in the subsequent processing step to avoid
overwriting data and thus the output of linrad can
stay correct even under difficult CPU load conditions.
(But with a somewhat longer processing delay)
As a side effect the calibration thread hangs when one
presses Y because the fft1 thread finds that the buffer
of the calibration routine is full. I forgot to do
the appropriate modification to mark the buffer as
empty after the (long) time it takes the user to
The calibration thread waits on a semaphore for a new
fft1 transform but the fft1 thread waits (sleeping) for
the calibration thread to mark old data as used.....
This is a typical example when a small change at one place
causes an unexpected error at some other place.
(Of course it should not have been unexpected.....)
Again: Rapid feedback on errors is most helpful:-)
The bug is corrected in Linrad-02.55
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