void speck128_encryptx(
const void *key,
void *in)
{
uint64_t i, t, k0, k1, k2, k3, x0, x1;
// copy 256-bit key to local space
k0 = ((uint64_t*)key)[0];
k1 = ((uint64_t*)key)[1];
k2 = ((uint64_t*)key)[2];
k3 = ((uint64_t*)key)[3];
// copy input to local space
x0 = ((uint64_t*)in)[0];
x1 = ((uint64_t*)in)[1];
for (i=0; i<34; i++)
{
// encrypt block
x1 = (ROTR64(x1, 8) + x0) ^ k0;
x0 = ROTL64(x0, 3) ^ x1;
// create next subkey
k1 = (ROTR64(k1, 8) + k0) ^ i;
k0 = ROTL64(k0, 3) ^ k1;
XCHG(k3, k2, t);
XCHG(k3, k1, t);
}
// save result
((uint64_t*)in)[0] = x0;
((uint64_t*)in)[1] = x1;
}
void speck128_encryptx(const void *key, void *in)
{
uint32_t i, t, k0, k1, k2, k3, x0, x1;
bc_blk *x=(bc_blk*)in;
bc_blk *k=(bc_blk*)key;
// copy 128-bit key to local registers
k0 = k->w[0]; k1 = k->w[1];
k2 = k->w[2]; k3 = k->w[3];
// copy M to local space
x0 = x->w[0]; x1 = x->w[1];
for (i=0; i<27; i++)
{
// encrypt block
x0 = (ROTR32(x0, 8) + x1) ^ k0;
x1 = ROTL32(x1, 3) ^ x0;
// create next subkey
k1 = (ROTR32(k1, 8) + k0) ^ i;
k0 = ROTL32(k0, 3) ^ k1;
XCHG(k3, k2, t);
XCHG(k3, k1, t);
}
// save result
x->w[0] = x0; x->w[1] = x1;
}
void speck32_encryptx(const void *key, void *data)
{
uint16_t i, t, k0, k1, k2, k3, x0, x1;
uint16_t *k=(uint16_t*)key;
uint16_t *x=(uint16_t*)data;
// copy 128-bit key to local space
k0 = k[0]; k1 = k[1];
k2 = k[2]; k3 = k[3];
// copy input to local space
x0 = x[0]; x1 = x[1];
for (i=0; i<22; i++)
{
// encrypt block
x0 = (ROTR16(x0, 7) + x1) ^ k0;
x1 = ROTL16(x1, 2) ^ x0;
// create next subkey
k1 = (ROTR16(k1, 7) + k0) ^ i;
k0 = ROTL16(k0, 2) ^ k1;
XCHG(k3, k2);
XCHG(k3, k1);
}
// save result
x[0] = x0; x[1] = x1;
}
void speck128_setkey(
const void *in,
void *out)
{
uint64_t i, t, k0, k1, k2, k3;
// copy 256-bit key to local space
k0 = ((uint64_t*)in)[0];
k1 = ((uint64_t*)in)[1];
k2 = ((uint64_t*)in)[2];
k3 = ((uint64_t*)in)[3];
// expand 256-bit key into round keys
for (i=0; i<34; i++)
{
((uint64_t*)out)[i] = k0;
k1 = (ROTR64(k1, 8) + k0) ^ i;
k0 = ROTL64(k0, 3) ^ k1;
// rotate left 64-bits
XCHG(k3, k2, t);
XCHG(k3, k1, t);
}
}
void xoodoo(void *state) {
uint32_t e[4], a0, a1, a2, a3, t;
int i, x, y, j;
uint32_t *p, *q;
uint32_t (*A)[4] = (uint32_t(*)[4])state;
p=(uint32_t*)A[0];
// 12 rounds by default
for (i=0; i<12; i++) {
// θ
for (x=0; x<4; x++) {
e[x] = ROTR32(p[x] ^ p[x+4] ^ p[x+8], 18);
e[x]^= ROTR32(e[x], 9);
}
for (x=0; x<12; x++) {
p[x] ^= e[(x - 1) & 3];
}
// ρ_west
XCHG(p[7], p[4]);
XCHG(p[7], p[5]);
XCHG(p[7], p[6]);
for (x=0; x<4; x++) {
p[x+8] = ROTR32(p[x+8], 21);
}
// ι
p[0] ^= (uint32_t)rc[i];
// χ
for (x=0; x<4; x++) {
a0 = p[x+0];
a1 = p[x+4];
a2 = p[x+8];
p[x+0] ^= ~a1 & a2;
p[x+4] ^= ~a2 & a0;
p[x+8] ^= ~a0 & a1;
}
// ******
// ρ_east
// ******
for (x=0;x<4; x++) {
p[x+4] = ROTR32(p[x+4], 31);
p[x+8] = ROTR32(p[x+8], 24);
}
XCHG(p[8], p[10]);
XCHG(p[9], p[11]);
}
}
static void transpose( uint8_t *buf, int w )
{
int i, j;
for( i = 0; i < w; i++ )
for( j = 0; j < i; j++ )
XCHG( uint8_t, buf[w*i+j], buf[w*j+i] );
}
static int read_frame( cli_pic_t *p_pic, hnd_t handle, int i_frame )
{
thread_hnd_t *h = handle;
int ret = 0;
if( h->next_frame >= 0 )
{
x264_threadpool_wait( h->pool, h->next_args );
ret |= h->next_args->status;
}
if( h->next_frame == i_frame )
XCHG( cli_pic_t, *p_pic, h->pic );
else
{
if( h->next_frame >= 0 )
thread_input.release_frame( &h->pic, handle );
ret |= h->input.read_frame( p_pic, h->p_handle, i_frame );
}
if( !h->frame_total || i_frame+1 < h->frame_total )
{
h->next_frame =
h->next_args->i_frame = i_frame+1;
h->next_args->pic = &h->pic;
x264_threadpool_run( h->pool, (void*)read_frame_thread_int, h->next_args );
}
else
h->next_frame = -1;
return ret;
}
static int read_frame( x264_picture_t *p_pic, hnd_t handle, int i_frame )
{
thread_hnd_t *h = handle;
int ret = 0;
if( h->next_frame >= 0 )
{
x264_pthread_join( h->tid, NULL );
ret |= h->next_args->status;
h->in_progress = 0;
}
if( h->next_frame == i_frame )
XCHG( x264_picture_t, *p_pic, h->pic );
else
ret |= h->input.read_frame( p_pic, h->p_handle, i_frame );
if( !h->frame_total || i_frame+1 < h->frame_total )
{
h->next_frame =
h->next_args->i_frame = i_frame+1;
h->next_args->pic = &h->pic;
if( x264_pthread_create( &h->tid, NULL, (void*)read_frame_thread_int, h->next_args ) )
return -1;
h->in_progress = 1;
}
else
h->next_frame = -1;
return ret;
}
void speck32_setkey(const void *in, void *out)
{
uint16_t i, t, k0, k1, k2, k3;
uint16_t *k=(uint16_t*)in;
uint16_t *ks=(uint16_t*)out;
// copy 64-bit key to local space
k0 = k[0]; k1 = k[1];
k2 = k[2]; k3 = k[3];
// expand 64-bit key into round keys
for (i=0; i<22; i++)
{
ks[i] = k0;
k1 = (ROTR16(k1, 7) + k0) ^ i;
k0 = ROTL16(k0, 2) ^ k1;
// rotate left 32-bits
XCHG(k3, k2);
XCHG(k3, k1);
}
}
void rc5_encryptx(void *key, void *data)
{
uint32_t A, B, t;
uint32_t L[4], S[RC5_KR];
int i, j,r;
uint32_t *k=(uint32_t*)key;
uint32_t *x=(uint32_t*)data;
// initialize L with 128-bit key
memcpy(&L, key, 16);
A = 0xC983AE2D; //RC5_P;
// initialize S with constants
for (i=RC5_KR-1; i>=0; i--) {
A -= RC5_Q;
S[i] = A;
}
A = B = 0;
r = (RC5_KR*3) - 1;
// create subkeys
for (i=0, j=0; r>0; i++, j++, r--) {
// i needs reset?
if (i==RC5_KR) i=0;
A = S[i%RC5_KR] = ROTR32(S[i%RC5_KR] + A+B, 32-3);
B = L[j%4] = ROTR32(L[j%4] + A+B, 32-U8V(A+B));
}
// assign subkeys to k ptr
k = S;
// load plaintext
A = x[0] + *k; k++;
B = x[1] + *k; k++;
// apply encryption
for (i=RC5_KR-2; i>0; i--) {
// original spec uses ROTL32
A = ROTR32(A ^ B, 32 - U8V(B)) + *k; k++;
// rotate
XCHG(A, B);
}
// save
x[0] = A; x[1] = B;
}
void gimlix(void *state)
{
int r, j;
uint32_t t, x, y, z;
uint32_t *s=(uint32_t*)state;
for (r=24; r>0; --r) {
// apply SP-box
for (j=0; j<4; j++) {
x = ROTR32(s[ j], 8);
y = ROTR32(s[4 + j], 23);
z = s[8 + j];
s[8 + j] = x ^ (z << 1) ^ ((y & z) << 2);
s[4 + j] = y ^ x ^ ((x | z) << 1);
s[j] = z ^ y ^ ((x & y) << 3);
}
// apply Linear layer
t = r & 3;
// if zero, do small swap
if (t == 0) {
XCHG(s[0], s[1]);
XCHG(s[2], s[3]);
// add constant
s[0] ^= (0x9e377900 | r);
}
// if 2, do big swap
if (t == 2) {
XCHG(s[0], s[2]);
XCHG(s[1], s[3]);
}
}
}
static void
p_sort_(uint* base, uint* l, uint* h)
{
uint* i;
for(i = l; i < h; i++)
{
uint* elm = base + *(i);
uint* j;
for(j = i+1; j < h; j++)
{
if(*elm > *(base + *(j)))
{
XCHG(*(i), *(j));
elm = base + *(i);
}
}
}
}
static int read_frame_internal( cli_pic_t *p_pic, lavf_hnd_t *h, int i_frame, video_info_t *info )
{
if( h->first_pic && !info )
{
/* see if the frame we are requesting is the frame we have already read and stored.
* if so, retrieve the pts and image data before freeing it. */
if( !i_frame )
{
XCHG( cli_image_t, p_pic->img, h->first_pic->img );
p_pic->pts = h->first_pic->pts;
XCHG( void*, p_pic->opaque, h->first_pic->opaque );
}
lavf_input.release_frame( h->first_pic, NULL );
lavf_input.picture_clean( h->first_pic );
free( h->first_pic );
h->first_pic = NULL;
if( !i_frame )
return 0;
}
static void __chfsnprb_tree_insert_color(CHFSNPRB_POOL *pool, const uint32_t node_pos, uint32_t *root_pos)
{
CHFSNPRB_NODE *node;
CHFSNPRB_NODE *root;
CHFSNPRB_NODE *parent;
uint32_t node_pos_t;
node_pos_t = node_pos;
node = CHFSNPRB_POOL_NODE(pool, node_pos_t);
while (NULL_PTR != (parent = CHFSNPRB_POOL_NODE(pool, CHFSNPRB_NODE_PARENT_POS(node))) /*parent is valid*/
&& CHFSNPRB_RED == CHFSNPRB_NODE_COLOR(parent))
{
uint32_t parent_pos;
uint32_t gparent_pos;
CHFSNPRB_NODE *gparent;
parent_pos = CHFSNPRB_NODE_PARENT_POS(node);
gparent_pos = CHFSNPRB_NODE_PARENT_POS(parent);
ASSERT(CHFSNPRB_ERR_POS != gparent_pos);
gparent = CHFSNPRB_POOL_NODE(pool, gparent_pos);
if (parent_pos == CHFSNPRB_NODE_LEFT_POS(gparent))
{
{
CHFSNPRB_NODE *uncle;
if (NULL_PTR != (uncle = CHFSNPRB_POOL_NODE(pool, CHFSNPRB_NODE_RIGHT_POS(gparent))) /*uncle is valid*/
&& CHFSNPRB_RED == CHFSNPRB_NODE_COLOR(uncle))
{
CHFSNPRB_NODE_COLOR(uncle) = CHFSNPRB_BLACK;
CHFSNPRB_NODE_COLOR(parent) = CHFSNPRB_BLACK;
CHFSNPRB_NODE_COLOR(gparent) = CHFSNPRB_RED;
node = gparent;
node_pos_t = gparent_pos;
continue;
}
}
if (CHFSNPRB_NODE_RIGHT_POS(parent) == node_pos_t)
{
__chfsnprb_tree_rotate_left(pool, parent_pos, root_pos);
XCHG(CHFSNPRB_NODE *, parent, node);
XCHG(uint32_t, parent_pos, node_pos_t);
}
CHFSNPRB_NODE_COLOR(parent) = CHFSNPRB_BLACK;
CHFSNPRB_NODE_COLOR(gparent) = CHFSNPRB_RED;
__chfsnprb_tree_rotate_right(pool, gparent_pos, root_pos);
}
else
{
{
CHFSNPRB_NODE *uncle;
if (NULL_PTR != (uncle = CHFSNPRB_POOL_NODE(pool, CHFSNPRB_NODE_LEFT_POS(gparent))) /*uncle is valid*/
&& CHFSNPRB_RED == CHFSNPRB_NODE_COLOR(uncle))
{
CHFSNPRB_NODE_COLOR(uncle) = CHFSNPRB_BLACK;
CHFSNPRB_NODE_COLOR(parent) = CHFSNPRB_BLACK;
CHFSNPRB_NODE_COLOR(gparent) = CHFSNPRB_RED;
node = gparent;
node_pos_t = gparent_pos;
continue;
}
}
if (CHFSNPRB_NODE_LEFT_POS(parent) == node_pos_t)
{
__chfsnprb_tree_rotate_right(pool, parent_pos, root_pos);
XCHG(CHFSNPRB_NODE *, parent, node);
XCHG(uint32_t, parent_pos, node_pos_t);
}
CHFSNPRB_NODE_COLOR(parent) = CHFSNPRB_BLACK;
CHFSNPRB_NODE_COLOR(gparent) = CHFSNPRB_RED;
__chfsnprb_tree_rotate_left(pool, gparent_pos, root_pos);
}
}
root = CHFSNPRB_POOL_NODE(pool, *root_pos);
CHFSNPRB_NODE_COLOR(root) = CHFSNPRB_BLACK;
return;
}
static int read_frame_internal( x264_picture_t *p_pic, lavf_hnd_t *h, int i_frame, video_info_t *info )
{
if( h->first_pic && !info )
{
/* see if the frame we are requesting is the frame we have already read and stored.
* if so, retrieve the pts and image data before freeing it. */
if( !i_frame )
{
XCHG( x264_image_t, p_pic->img, h->first_pic->img );
p_pic->i_pts = h->first_pic->i_pts;
}
lavf_input.picture_clean( h->first_pic );
free( h->first_pic );
h->first_pic = NULL;
if( !i_frame )
return 0;
}
AVCodecContext *c = h->lavf->streams[h->stream_id]->codec;
lavf_pic_t *pic_h = p_pic->opaque;
AVPacket *pkt = &pic_h->packet;
AVFrame *frame = &pic_h->frame;
while( i_frame >= h->next_frame )
{
int finished = 0;
while( !finished && av_read_frame( h->lavf, pkt ) >= 0 )
if( pkt->stream_index == h->stream_id )
{
c->reordered_opaque = pkt->pts;
if( avcodec_decode_video2( c, frame, &finished, pkt ) < 0 )
fprintf( stderr, "lavf [warning]: video decoding failed on frame %d\n", h->next_frame );
}
if( !finished )
{
if( avcodec_decode_video2( c, frame, &finished, pkt ) < 0 )
fprintf( stderr, "lavf [warning]: video decoding failed on frame %d\n", h->next_frame );
if( !finished )
return -1;
}
h->next_frame++;
}
if( check_swscale( h, c, i_frame ) )
return -1;
/* FIXME: avoid sws_scale where possible (no colorspace conversion). */
sws_scale( h->scaler, frame->data, frame->linesize, 0, c->height, p_pic->img.plane, p_pic->img.i_stride );
if( info )
info->interlaced = frame->interlaced_frame;
if( h->vfr_input )
{
p_pic->i_pts = 0;
if( frame->reordered_opaque != AV_NOPTS_VALUE )
p_pic->i_pts = frame->reordered_opaque;
else if( pkt->dts != AV_NOPTS_VALUE )
p_pic->i_pts = pkt->dts; // for AVI files
else if( info )
{
h->vfr_input = info->vfr = 0;
goto exit;
}
if( !h->pts_offset_flag )
{
h->pts_offset = p_pic->i_pts;
h->pts_offset_flag = 1;
}
p_pic->i_pts -= h->pts_offset;
}
exit:
if( pkt->destruct )
pkt->destruct( pkt );
avcodec_get_frame_defaults( frame );
return 0;
}