wave8Transpose_16x2_pipe2()

FL_OPTIMIZE_FUNCTION FL_IRAM void fl::wave8Transpose_16x2_pipe2	(	const u8(&)	lanes_a[16],
		const u8(&)	lanes_b[16],
		const Wave8ByteExpansionLut &	lut,
		u8(&)	output_a[16 *sizeof(Wave8Byte)],
		u8(&)	output_b[16 *sizeof(Wave8Byte)] )

Pipe2: transpose 16-lane × 2-byte-positions (#2548).

Bit-identical to two sequential wave8Transpose_16 calls. Internally interleaves the two independent OR-trees inside the symbol loop so the in-order RV32 P4 can fill load-use stalls from position A with ALU ops from position B. Measured 9655 → 7625 µs/frame (+26%) on P4 v1.3 (16-lane × 256-LED, byte-LUT path) — first variant to beat the 7680 µs WS2812B 16-lane TX target.

Definition at line 138 of file wave8.cpp.hpp.

                                                                                                      {
    // Expand both positions independently — compiler can interleave the two
    // loops freely because they share no data.
    Wave8Byte laneWaveformsA[16];
    Wave8Byte laneWaveformsB[16];
    for (int lane = 0; lane < 16; lane++) {
        detail::wave8_expand_byte(lanes_a[lane], lut, &laneWaveformsA[lane]);
    }
    for (int lane = 0; lane < 16; lane++) {
        detail::wave8_expand_byte(lanes_b[lane], lut, &laneWaveformsB[lane]);
    }
    // Symbol-major loop with both transposes inlined back-to-back: the two
    // OR-trees share no dependencies, so the compiler can interleave them and
    // fill the in-order pipeline bubbles. See #2548.
    detail::wave8_transpose_16x2_pipe2(laneWaveformsA, laneWaveformsB,
                                       output_a, output_b);
}

References FL_NO_EXCEPT, FL_RESTRICT_PARAM, fl::detail::wave8_expand_byte(), and fl::detail::wave8_transpose_16x2_pipe2().

Here is the call graph for this function: