Matt Hess
7730b6993f
- Create assembly/vm.ts with full RandomX VM implementation
- 256 instructions, 2048 iterations per hash
- Native u64/f64 operations in WebAssembly
- Full mode dataset lookups (2GB pre-computed)
- Update miner to use WASM VM for full mode
- mining-worker-asm.js uses pre-compiled WASM
- ~32 H/s per thread (4x faster than light mode)
- 8 threads achieves ~260 H/s
- Clean up redundant code
- Remove mining-worker-full.js (old JIT approach)
- Consolidate 'asm' mode into 'full' mode
76 lines
3.2 KiB
JavaScript
76 lines
3.2 KiB
JavaScript
/**
|
|
* Debug: Verify AES key byte ordering matches C++
|
|
*
|
|
* C++ rx_set_int_vec_i128(a, b, c, d) uses _mm_set_epi32(d, c, b, a)
|
|
* which puts 'a' at bytes 0-3, 'b' at bytes 4-7, etc.
|
|
*
|
|
* Example: AES_GEN_1R_KEY0 = 0xb4f44917, 0xdbb5552b, 0x62716609, 0x6daca553
|
|
*
|
|
* Expected byte order (little-endian per 32-bit value):
|
|
* bytes 0-3: 0xb4f44917 -> 17 49 f4 b4
|
|
* bytes 4-7: 0xdbb5552b -> 2b 55 b5 db
|
|
* bytes 8-11: 0x62716609 -> 09 66 71 62
|
|
* bytes 12-15: 0x6daca553 -> 53 a5 ac 6d
|
|
*
|
|
* Full key: 1749f4b42b55b5db09667162 53a5ac6d
|
|
*/
|
|
|
|
function bytesToHex(bytes) {
|
|
return Array.from(bytes).map(b => b.toString(16).padStart(2, '0')).join('');
|
|
}
|
|
|
|
// Current implementation (with v3, v2, v1, v0 parameter order)
|
|
function i128ToBytesOld(v3, v2, v1, v0) {
|
|
const arr = new Uint8Array(16);
|
|
arr[0] = v0 & 0xff; arr[1] = (v0 >> 8) & 0xff; arr[2] = (v0 >> 16) & 0xff; arr[3] = (v0 >> 24) & 0xff;
|
|
arr[4] = v1 & 0xff; arr[5] = (v1 >> 8) & 0xff; arr[6] = (v1 >> 16) & 0xff; arr[7] = (v1 >> 24) & 0xff;
|
|
arr[8] = v2 & 0xff; arr[9] = (v2 >> 8) & 0xff; arr[10] = (v2 >> 16) & 0xff; arr[11] = (v2 >> 24) & 0xff;
|
|
arr[12] = v3 & 0xff; arr[13] = (v3 >> 8) & 0xff; arr[14] = (v3 >> 16) & 0xff; arr[15] = (v3 >> 24) & 0xff;
|
|
return arr;
|
|
}
|
|
|
|
// Alternative: v0 first parameter order
|
|
function i128ToBytesNew(v0, v1, v2, v3) {
|
|
const arr = new Uint8Array(16);
|
|
arr[0] = v0 & 0xff; arr[1] = (v0 >> 8) & 0xff; arr[2] = (v0 >> 16) & 0xff; arr[3] = (v0 >> 24) & 0xff;
|
|
arr[4] = v1 & 0xff; arr[5] = (v1 >> 8) & 0xff; arr[6] = (v1 >> 16) & 0xff; arr[7] = (v1 >> 24) & 0xff;
|
|
arr[8] = v2 & 0xff; arr[9] = (v2 >> 8) & 0xff; arr[10] = (v2 >> 16) & 0xff; arr[11] = (v2 >> 24) & 0xff;
|
|
arr[12] = v3 & 0xff; arr[13] = (v3 >> 8) & 0xff; arr[14] = (v3 >> 16) & 0xff; arr[15] = (v3 >> 24) & 0xff;
|
|
return arr;
|
|
}
|
|
|
|
console.log('=== AES Key Byte Order Debug ===\n');
|
|
|
|
// Test with AES_GEN_1R_KEY0 values: 0xb4f44917, 0xdbb5552b, 0x62716609, 0x6daca553
|
|
const a = 0xb4f44917;
|
|
const b = 0xdbb5552b;
|
|
const c = 0x62716609;
|
|
const d = 0x6daca553;
|
|
|
|
console.log('Input values (in C++ macro order):');
|
|
console.log(` a = 0x${a.toString(16)}`);
|
|
console.log(` b = 0x${b.toString(16)}`);
|
|
console.log(` c = 0x${c.toString(16)}`);
|
|
console.log(` d = 0x${d.toString(16)}`);
|
|
console.log();
|
|
|
|
// C++ rx_set_int_vec_i128(a, b, c, d) uses _mm_set_epi32(d, c, b, a)
|
|
// which results in: a at bytes 0-3, b at bytes 4-7, c at bytes 8-11, d at bytes 12-15
|
|
console.log('Expected (C++ rx_set_int_vec_i128 behavior):');
|
|
console.log(' a at bytes 0-3, b at bytes 4-7, c at bytes 8-11, d at bytes 12-15');
|
|
console.log(` = 1749f4b42b55b5db0966716253a5ac6d`);
|
|
console.log();
|
|
|
|
console.log('Current i128ToBytes(a,b,c,d) with (v3,v2,v1,v0) signature:');
|
|
const old = i128ToBytesOld(a, b, c, d);
|
|
console.log(` = ${bytesToHex(old)}`);
|
|
console.log(` v0 (${d.toString(16)}) at bytes 0-3: ${bytesToHex(old.slice(0, 4))}`);
|
|
console.log(` v3 (${a.toString(16)}) at bytes 12-15: ${bytesToHex(old.slice(12, 16))}`);
|
|
console.log();
|
|
|
|
console.log('Alternative i128ToBytes(a,b,c,d) with (v0,v1,v2,v3) signature:');
|
|
const newKey = i128ToBytesNew(a, b, c, d);
|
|
console.log(` = ${bytesToHex(newKey)}`);
|
|
console.log(` v0 (${a.toString(16)}) at bytes 0-3: ${bytesToHex(newKey.slice(0, 4))}`);
|
|
console.log(` v3 (${d.toString(16)}) at bytes 12-15: ${bytesToHex(newKey.slice(12, 16))}`);
|