why would sin(0). return nan?

greenmuzz · 10-23-2003, 05:07 AM

just entered the function, as shown in the output from gdb below.
'ang' comes from the parameter of the funciton, and can be seen to be 0.
single stepping through the next few lines should calculate the sin & cos
of ang, but the result is as you can see below.

What can cause this?

GetRay (isurf=0x804ff10, startx=80, starty=119, ang=0, oray=0xbfffea50) at actray.cpp:96
96 float cosAng = cos(ang);
(gdb)
97 float sinAng = sin(ang);
(gdb)
98 int lambda = 0;
(gdb) print sinAng
$3 = -nan(0x400000)

jim mcnamara · 10-23-2003, 05:50 AM

I think it's a datatype overflow.
cos() and sin() for most implementations of C return a double, not a float.

greenmuzz · 10-23-2003, 06:03 AM

just tried, no change.

Did a test program too just to make sure there was nothing funny going on with the c++ libraries (well, you never know...). But they worked fine using float to get the answer back from cos & sin.

Could this be something to do with g++ version 2.95.3 ?

sn32/src> gcc -v
Reading specs from /usr/lib/gcc-lib/i486-suse-linux/2.95.3/specs
gcc version 2.95.3 20010315 (SuSE)

this is really annoying me now!