transpose_16lane_inline()

void fl::transpose_16lane_inline ( const u8 *const lanes[16], u8 * output, size_t num_bytes )

inline

Low-level bit-interleaving primitive for 16 lanes (ISR-safe)

Transposes 16 input bytes into 16-way interleaved format. This function is ISR-safe: no allocations, no exceptions, minimal overhead. Inline functions are automatically placed where needed - no IRAM_ATTR required.

Parameters

lanes	Array of 16 lane byte pointers
output	Output buffer (must have space for num_bytes * 16 bytes)
num_bytes	Number of bytes to transpose per lane

Note

Inline function - inlined at call site (including ISR contexts)

Output size is num_bytes * 16

Examples

/home/runner/work/FastLED/FastLED/src/fl/math/transposition.h.

Definition at line 417 of file transposition.h.

              {
    for (size_t byte_idx = 0; byte_idx < num_bytes; byte_idx++) {
        // Pack lanes 0-7 into first 64-bit register
        u64 packed_lo =
            ((u64)lanes[0][byte_idx] << 0)  |
            ((u64)lanes[1][byte_idx] << 8)  |
            ((u64)lanes[2][byte_idx] << 16) |
            ((u64)lanes[3][byte_idx] << 24) |
            ((u64)lanes[4][byte_idx] << 32) |
            ((u64)lanes[5][byte_idx] << 40) |
            ((u64)lanes[6][byte_idx] << 48) |
            ((u64)lanes[7][byte_idx] << 56);
 
        // Pack lanes 8-15 into second 64-bit register
        u64 packed_hi =
            ((u64)lanes[8][byte_idx]  << 0)  |
            ((u64)lanes[9][byte_idx]  << 8)  |
            ((u64)lanes[10][byte_idx] << 16) |
            ((u64)lanes[11][byte_idx] << 24) |
            ((u64)lanes[12][byte_idx] << 32) |
            ((u64)lanes[13][byte_idx] << 40) |
            ((u64)lanes[14][byte_idx] << 48) |
            ((u64)lanes[15][byte_idx] << 56);
 
        u8* dest = &output[byte_idx * 16];
 
        // Extract bits in parallel from both packed registers
        for (int bit = 7; bit >= 0; bit--) {
            dest[7 - bit] =
                ((packed_lo >> (bit + 0))  & 0x01) << 0 |
                ((packed_lo >> (bit + 8))  & 0x01) << 1 |
                ((packed_lo >> (bit + 16)) & 0x01) << 2 |
                ((packed_lo >> (bit + 24)) & 0x01) << 3 |
                ((packed_lo >> (bit + 32)) & 0x01) << 4 |
                ((packed_lo >> (bit + 40)) & 0x01) << 5 |
                ((packed_lo >> (bit + 48)) & 0x01) << 6 |
                ((packed_lo >> (bit + 56)) & 0x01) << 7;
 
            dest[15 - bit] =
                ((packed_hi >> (bit + 0))  & 0x01) << 0 |
                ((packed_hi >> (bit + 8))  & 0x01) << 1 |
                ((packed_hi >> (bit + 16)) & 0x01) << 2 |
                ((packed_hi >> (bit + 24)) & 0x01) << 3 |
                ((packed_hi >> (bit + 32)) & 0x01) << 4 |
                ((packed_hi >> (bit + 40)) & 0x01) << 5 |
                ((packed_hi >> (bit + 48)) & 0x01) << 6 |
                ((packed_hi >> (bit + 56)) & 0x01) << 7;
        }
    }
}

References FL_NOEXCEPT.

Referenced by fl::SPITransposer::transpose16().

Here is the caller graph for this function: