applyFluxScalar()

void fl::applyFluxScalar	(	FluxScalar	scalar,
		span< i32 >	drives )

Scale every drive by one scalar, in place.

Definition at line 40 of file flux_scalar.cpp.hpp.

                                                                       {
    const i64 raw = static_cast<i64>(scalar.rawQ16());
    if (raw == kUnityQ16) {
        // Pure short-circuit: the general path below is already bit-exact at
        // unity, since drive * 65536 + 32768 >> 16 == drive. This only avoids
        // walking the buffer on the default full-brightness path. Exactness is
        // pinned by a test, not by this branch -- verified by removing it.
        return;
    }
    for (size i = 0; i < drives.size(); ++i) {
        const i64 scaled = static_cast<i64>(drives[i]) * raw;
        // Round half away from zero so negative signed-wide intermediates
        // are treated symmetrically.
        drives[i] = static_cast<i32>(scaled >= 0 ? (scaled + 32768) >> 16
                                                 : -((-scaled + 32768) >> 16));
    }
}

References FL_NO_EXCEPT.

Referenced by encodeLinearEmitterQ16(), processPixelQ16(), and processPixelWideQ16().

Here is the caller graph for this function: