huh... so uh... i am observing weird behavior in code compiled using tinycc on windows...

it's like:

float d = 0.6f;
float x = pow(d/4.f, 4.f);
printf("x: %e pow(...): %e", x, pow(d/4.f, 4.f));

you'd expect both x and the just computed output of pow to be the same, ne?

but it isn't... for some reason x contains 256 instead of the expected 5.0625e-4

it appears the 64bit version is kinda borked...