#include #include namespace twiddle { /* This reproduces the twiddle index table shown in "3.6.2.1 Twiddled Format". x → 000 001 010 011 -------------------------------- | xyxyxy xyxyxy xyxyxy xyxyxy |=============================== y 000 | 000000 000010 001000 001010 ↓ 001 | 000001 000011 001001 001011 010 | 000100 000110 001100 001110 011 | 000101 000111 001101 001111 alternately, in verilog syntax: input [2:0] x; // x coordinate input [2:0] y; // y coordinate output [5:0] t; // twiddled index assign t = {x[2], y[2], x[1], y[1], x[0], y[0]}; */ constexpr inline int log2(uint32_t n) { switch (n) { default: case 8: return 3; case 16: return 4; case 32: return 5; case 64: return 6; case 128: return 7; case 256: return 8; case 512: return 9; case 1024: return 10; } } constexpr inline uint32_t from_xy(uint32_t x, uint32_t y, uint32_t width, uint32_t height) { // maximum texture size : 1024x1024 // maximum 1-dimensional index: 0xfffff // bits : 19-0 // y bits: 0, 2, 4, 6, 8, 10, 12, 14, 16, 18 // x bits: 1, 3, 5, 7, 9, 11, 13, 15, 17, 19 int width_max = log2(width); int height_max = log2(height); uint32_t twiddle_ix = 0; for (int i = 0; i < (20 / 2); i++) { if (i < width_max && i < height_max) { twiddle_ix |= ((y >> i) & 1) << (i * 2 + 0); twiddle_ix |= ((x >> i) & 1) << (i * 2 + 1); } else if (i < width_max) { twiddle_ix |= ((x >> i) & 1) << (i + height_max); } else if (i < height_max) { twiddle_ix |= ((y >> i) & 1) << (i + width_max); } else { break; } } return twiddle_ix; } static_assert(from_xy(0b000, 0b000, 8, 8) == 0); static_assert(from_xy(0b001, 0b000, 8, 8) == 2); static_assert(from_xy(0b010, 0b000, 8, 8) == 8); static_assert(from_xy(0b011, 0b000, 8, 8) == 10); static_assert(from_xy(0b100, 0b000, 8, 8) == 32); static_assert(from_xy(0b101, 0b000, 8, 8) == 34); static_assert(from_xy(0b110, 0b000, 8, 8) == 40); static_assert(from_xy(0b111, 0b000, 8, 8) == 42); static_assert(from_xy(0b000, 0b001, 8, 8) == 1); static_assert(from_xy(0b000, 0b010, 8, 8) == 4); static_assert(from_xy(0b000, 0b011, 8, 8) == 5); static_assert(from_xy(0b000, 0b100, 8, 8) == 16); static_assert(from_xy(0b000, 0b101, 8, 8) == 17); static_assert(from_xy(0b000, 0b110, 8, 8) == 20); static_assert(from_xy(0b000, 0b111, 8, 8) == 21); // 1 0 0 0 // x bits: 19, 17, 15, 13, 11, 9, 7, 5, 3, 1 // y bits: 18, 16, 14, 12, 10, 8, 6, 4, 2, 0 static_assert(from_xy(0b1000, 0b001, 16, 8) == 65); static_assert(from_xy(0b1010, 0b001, 16, 8) == 73); static_assert(from_xy(0b000, 0b1001, 8, 16) == 65); static_assert(from_xy(0b010, 0b1001, 8, 16) == 73); /* constexpr inline std::array from_ix(uint32_t curve_ix) { std::array x_y = {0, 0}; uint32_t curve_bit = 0; while (curve_ix != 0) { x_y[(curve_bit + 1) % 2] |= (curve_ix & 1) << (curve_bit / 2); curve_ix >>= 1; curve_bit += 1; } return x_y; } constexpr inline std::tuple from_ix(uint32_t curve_ix) { uint32_t y = (curve_ix >> 0) & 0x55555555; y = (y | (y >> 1)) & 0x33333333; y = (y | (y >> 2)) & 0x0f0f0f0f; y = (y | (y >> 4)) & 0x00ff00ff; y = (y | (y >> 8)) & 0x0000ffff; uint32_t x = (curve_ix >> 1) & 0x55555555; x = (x | (x >> 1)) & 0x33333333; x = (x | (x >> 2)) & 0x0f0f0f0f; x = (x | (x >> 4)) & 0x00ff00ff; x = (x | (x >> 8)) & 0x0000ffff; return {x, y}; } using xy_type = std::tuple; static_assert(from_ix(0) == xy_type{0b000, 0b000}); static_assert(from_ix(2) == xy_type{0b001, 0b000}); static_assert(from_ix(8) == xy_type{0b010, 0b000}); static_assert(from_ix(10) == xy_type{0b011, 0b000}); static_assert(from_ix(32) == xy_type{0b100, 0b000}); static_assert(from_ix(34) == xy_type{0b101, 0b000}); static_assert(from_ix(40) == xy_type{0b110, 0b000}); static_assert(from_ix(42) == xy_type{0b111, 0b000}); static_assert(from_ix(1) == xy_type{0b000, 0b001}); static_assert(from_ix(4) == xy_type{0b000, 0b010}); static_assert(from_ix(5) == xy_type{0b000, 0b011}); static_assert(from_ix(16) == xy_type{0b000, 0b100}); static_assert(from_ix(17) == xy_type{0b000, 0b101}); static_assert(from_ix(20) == xy_type{0b000, 0b110}); static_assert(from_ix(21) == xy_type{0b000, 0b111}); */ template void texture(volatile T * dst, const T * src, const uint32_t width, const uint32_t height) { for (uint32_t y = 0; y < height; y++) { for (uint32_t x = 0; x < width; x++) { uint32_t twiddle_ix = from_xy(x, y, width, height); T value = src[y * width + x]; dst[twiddle_ix] = value; } } } template void texture_4bpp(volatile T * dst, const T * src, const uint32_t width, const uint32_t height) { for (uint32_t y = 0; y < height; y++) { for (uint32_t x = 0; x < width; x++) { uint32_t twiddle_ix = from_xy(x, y, width, height); T value = src[y * width + x]; uint32_t shift = (4 * (twiddle_ix & 1)); dst[twiddle_ix / 2] &= ~(0b1111 << shift); dst[twiddle_ix / 2] |= value << shift; } } } /* the following code is not correct template void texture2(volatile T * dst, const U * src, const uint32_t src_stride, const uint32_t curve_end_ix) { constexpr uint32_t t_bits = (sizeof (T)) * 8; static_assert(t_bits >= dst_bits_per_pixel); static_assert((t_bits / dst_bits_per_pixel) * dst_bits_per_pixel == t_bits); constexpr uint32_t pixels_per_t = t_bits / dst_bits_per_pixel; static_assert(pixels_per_t == 1 || pixels_per_t == 2 || pixels_per_t == 4 || pixels_per_t == 8 || pixels_per_t == 16 || pixels_per_t == 32); T dst_val = 0; for (uint32_t curve_ix = 0; curve_ix <= curve_end_ix; curve_ix++) { auto [x, y] = from_ix(curve_ix); const U src_val = src[y * src_stride + x]; if constexpr (pixels_per_t == 1) { dst[curve_ix] = src_val; } else { const uint32_t curve_ix_mod = curve_ix & (pixels_per_t - 1); dst_val |= src_val << (dst_bits_per_pixel * curve_ix_mod); if (curve_ix_mod == (pixels_per_t - 1)) { dst[curve_ix / pixels_per_t] = dst_val; dst_val = 0; } } } } template void texture3(volatile T * dst, const U * src, const uint32_t src_stride, const uint32_t curve_end_ix) { constexpr uint32_t t_bits = (sizeof (T)) * 8; static_assert(t_bits >= dst_bits_per_pixel); static_assert((t_bits / dst_bits_per_pixel) * dst_bits_per_pixel == t_bits); constexpr uint32_t pixels_per_t = t_bits / dst_bits_per_pixel; static_assert(pixels_per_t == 1 || pixels_per_t == 2 || pixels_per_t == 4 || pixels_per_t == 8 || pixels_per_t == 16 || pixels_per_t == 32); constexpr uint32_t u_bits = (sizeof (U)) * 8; static_assert(u_bits >= src_bits_per_pixel); static_assert((u_bits / src_bits_per_pixel) * src_bits_per_pixel == u_bits); constexpr uint32_t pixels_per_u = u_bits / src_bits_per_pixel; static_assert(pixels_per_u == 1 || pixels_per_u == 2 || pixels_per_u == 4 || pixels_per_u == 8 || pixels_per_u == 16 || pixels_per_u == 32); constexpr uint32_t src_val_mask = ((1 << src_bits_per_pixel) - 1); constexpr uint32_t dst_val_mask = ((1 << dst_bits_per_pixel) - 1); constexpr uint32_t dst_src_shift = (dst_bits_per_pixel < src_bits_per_pixel) ? (src_bits_per_pixel - dst_bits_per_pixel) : 0; T dst_val = 0; for (uint32_t curve_ix = 0; curve_ix <= curve_end_ix; curve_ix++) { auto [x, y] = from_ix(curve_ix); const uint32_t src_ix = y * src_stride + (x / pixels_per_u); const uint32_t src_ix_mod = x & (pixels_per_u - 1); const U src_val = (src[src_ix] >> (src_bits_per_pixel * src_ix_mod)) & src_val_mask; if constexpr (pixels_per_t == 1) { dst[curve_ix] = src_val; } else { const uint32_t curve_ix_mod = curve_ix & (pixels_per_t - 1); dst_val |= ((src_val >> dst_src_shift) & dst_val_mask) << (dst_bits_per_pixel * curve_ix_mod); if (curve_ix_mod == (pixels_per_t - 1)) { dst[curve_ix / pixels_per_t] = dst_val; dst_val = 0; } } } } */ }