animetosho

# apply with `patch -s -p1 < unrar-gcc.patch` in unrar 6.2.6 source directory
--- a/blake2s.cpp
+++ b/blake2s.cpp
@@ -18,7 +18,7 @@
   0x510E527FUL, 0x9B05688CUL, 0x1F83D9ABUL, 0x5BE0CD19UL
 };
 
-static const byte blake2s_sigma[10][16] =
+const byte blake2s_sigma[10][16] =
 {

animetosho

"use strict";

// arguments:
//  dummy-nntp-server.js [-S] [-d=delay|0] [port/socket|119/563]
const DEBUG = false;
var delay = 0; // in miliseconds

var y;
var junkArticle, junkArticleCrc, junkFileCrc;
try {

animetosho

# compute a*b
CRC32_POLYNOMIAL = 0xedb88320
def crc_multiply(a: int, b: int):
	prod = 0
	while b != 0:
		prod ^= -(b>>31) & a
		a = (a>>1) ^ (CRC32_POLYNOMIAL & -(a&1))
		b = (b<<1) & 0xffffffff
	return prod

animetosho

#include <stdio.h>
#include <x86intrin.h>
#include <stdint.h>
#include <stdlib.h>
#include <string.h>

int main(void) {
	__m128i polymask1 = _mm_set1_epi64x(0xff00ffffff000000ULL); // polynomial 0x11d: 0x1d = 0b11101, reverse bits -> 10111000, then replace 1->ff, 0->00
	
	for(int val=0; val<256; val++) {

animetosho

--- a/libavformat/hevc.c
+++ b/libavformat/hevc.c
@@ -1080,6 +1080,15 @@
         return 0;
     } else if (!(AV_RB24(data) == 1 || AV_RB32(data) == 1)) {
         /* Not a valid Annex B start code prefix */
+        if (*data == 0) {
+            /* The version field is is ignored by the decoder, so any value should be allowed. As this code relies on magic identifiers, we'll be a bit more strict and only accept '0' as a candidate to be fixed */
+            /* (these instances are outputs likely sourced from older/buggy tools) */
+            /* it might be ideal to do extra checks on the subsequent field (profile/tier) but I can't find open documentation on how they work */

animetosho

Unexpected Uses for the Galois Field Affine Transformation Instruction

Intel added the Galois Field instruction set (GFNI) extensions to their Sunny Cove and Tremont cores. What’s particularly interesting is that GFNI is the only new SIMD extension that came with SSE and VEX/AVX encodings (in addition to EVEX/AVX512), to allow it to be supported on all future Intel cores, including those which don’t support AVX512 (such as the Atom line, as well as Celeron/Pentium branded “big” cores).

I suspect GFNI was aimed at accelerating SM4 encryption, however, one of the instructions can be used for many other purposes. The extension includes three instructions, but of particular interest here is the Affine Transformation (GF2P8AFFINEQB), aka bit-matrix multiply, instruction.

There have been various articles which discuss out-of-band

animetosho

Count Leading/Trailing Zero Bits (Byte-wise)

Counting the trailing zero bit count (TZCNT) can be done by isolating the lowest bit, then depositing this into the appropriate locations for the count. The leading zero bit count (LZCNT) can be done by reversing bits, then computing the TZCNT.

__m128i _mm_tzcnt_epi8(__m128i a) {
	// isolate lowest bit
	a = _mm_andnot_si128(_mm_add_epi8(a, _mm_set1_epi8(0xff)), a);
	// convert lowest bit to index