Use the right type of infinity so we don't get a compiler warning.

[ThiagoIze]

Depending on the windows toolkit being used on Windows (and maybe whatever #defines are set?) we can get compiler warnings like:

cuBQL/queries/pointData/findClosest.h(161): error #221-D: floating-point value does not fit in required floating-point type
                      float squareOfMaxQueryDistance=((float)(1e+300)));
                                                      ^

This happens because INFINITY can be defined by Windows toolkit to be 1e300:
C:\Program Files (x86)\Windows Kits\10\Include\10.0.26100.0\ucrt\corecrt_math.h

#ifndef _HUGE_ENUF
    #define _HUGE_ENUF  1e+300  // _HUGE_ENUF*_HUGE_ENUF must overflow
#endif
#ifndef _UCRT_LEGACY_INFINITY
#define INFINITY   ((float)(_HUGE_ENUF))

cubql already has functions to do this properly, so I went ahead and used that.

[iwald-nvidia]

CI broke on some (tiny) missing typecast. Already fixed but haven't figure out to modify the existing PR, so going to fix this through a new PR #24 (which is a pull of your diffs plus that typecast)
Just waiting for CI to run through before I can merge

[iwald-nvidia]

BTW the entire PosInfTy() etc isn't actually used all that much in cuBQL - it's pretty much a leftover from embree.... but now that you point that it's there I agree it's probably the right way of doing these constants in other places too.
And that 1e300 warning was been driving me crazy for a while now (also in other projects), so if that fixes it then that's worth doing - thx for pointing my nose that way.

[iwald-nvidia]

handling this through PR #24, closing this.