Date: Tue, 4 Apr 1995 11:10:25 -0600
From: craig AT sunspot DOT noao DOT edu (Craig Gullixson)
To: djgpp AT sun DOT soe DOT clarkson DOT edu
Subject: timers and smartdrive.exe

FYI, all

I am in the final stages of completing an image acquisition system
using the djgpp 1.12m4 development system.  The way the system works is
that a large number of 1024x1024 images are accumulated into a DSP
board, the final image is transferred to a long memory buffer in the pc
then written to disk (the final image sizes are between 1 and 4 Mb per
image).  I use the hrtimer.c code, posted by Christopher Christensen
sometime back, to time the exposures ( this code works by reading the
counts from the timer chip).  I ran into a snag with the sequence of
taking 100 images; writing to disk, taking another 100 images
immediately after the write.  During the acquisition of the second 100
images, smartdrive started to flush the disk cache resulting in longer
than requested exposure times.

I suspect that the disk I/O interrupts during the timer delay loop
results in the timer not catching when the delay count has just reached
the desired count resulting in the longer exposure times --- the
magnitude of the exposure increase reached about 30 ms for a 100 ms
exposure.

I am now loading smartdrive with the /C opton to force the flushing of
the cache during the write operation.

---Craig

________________________________________________________________________
Craig A. Gullixson
Senior Research Associate		INTERNET: craig AT sunspot DOT noao DOT edu
National Solar Observatory/Sac. Peak    PHONE: (505) 434-7065
Sunspot, NM 88349 USA