fly/flightgear
1
0
Fork 0
Code Activity

Fix a class of problem that could lead to needless extra time "jitter" in the flight dynamics

Browse source 
calculations.  We run the FDM at 120hz and compute how many loops can fit into each FG loop.
Floating point rounding could lead to a situation where we could end up running
1, 3, 1, 3, 1, 3... loops of the FDM when in fact we want to run 2, 2, 2, 2, 2...

If we artificially inflate ml above by a tiny amount to get the
closest integer, then subtract the integer from the original
slightly smaller value, we can get a negative remainder.
Logically this should never happen, and we definitely don't want
to carry a negative remainder over to the next iteration, so
never let the remainder go below zero.
This commit is contained in:
curt 2006-07-26 14:18:06 +00:00
parent 5d10e338d5
commit f31088070c
1 changed files with 12 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
src/FDM/flight.cxx
View file

@ -87,6 +87,18 @@ FGInterface::_calc_multiloop (double dt)
// ... ok, two times the roundoff to have enough room.
int multiloop = int(floor(ml * (1.0 + 2.0*DBL_EPSILON)));
remainder = (ml - multiloop) / hz;
// If we artificially inflate ml above by a tiny amount to get the
// closest integer, then subtract the integer from the original
// slightly smaller value, we can get a negative remainder.
// Logically this should never happen, and we definitely don't want
// to carry a negative remainder over to the next iteration, so
// never let the remainder go below zero.
//
// Note: this fixes a problem where we run 1, 3, 1, 3, 1, 3... loops
// of the FDM when in fact we want to run 2, 2, 2, 2, 2...
if ( remainder < 0 ) { remainder = 0; }
return (multiloop * speedup);
}