Closed FrenkelS closed 3 months ago
somehow in 16-bit Watcom
fixed_t xr = CENTERX * FRACUNIT - FRACUNIT + FixedMul(tx + (((int32_t)patch->width) << FRACBITS), xscale);
is different from
fixed_t xr = CENTERX * FRACUNIT + FixedMul(tx + (((int32_t)patch->width) << FRACBITS), xscale) - FRACUNIT;
Using 286 instructions is fine, but 8086 instructions results in an infinite loop in
R_DrawMaskedColumn()
when compiled with Watcom half way through the demo.