Skip to content

Commit

Permalink
fix compile error.
Browse files Browse the repository at this point in the history
  • Loading branch information
@aikiriao
aikiriao committed Jun 30, 2024
1 parent aab5cf0 commit 88b271f
Show file tree
Hide file tree
Showing 2 changed files with 43 additions and 43 deletions.
44 changes: 22 additions & 22 deletions libs/srla_decoder/src/srla_lpc_synthesize.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -51,20 +51,20 @@ void SRLALPC_Synthesize(
__m128i vpred = _mm_set1_epi32(half);
for (ord = 0; ord < (int32_t)coef_order - 3 - 4; ord += 4) {
const int32_t *dat = &data[smpl - coef_order + ord];
vdata = _mm_load_si128(&dat[0]);
vdata = _mm_load_si128((const __m128i *)&dat[0]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 0], vdata));
vdata = _mm_loadu_si128(&dat[1]);
vdata = _mm_loadu_si128((const __m128i *)&dat[1]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 1], vdata));
vdata = _mm_loadu_si128(&dat[2]);
vdata = _mm_loadu_si128((const __m128i *)&dat[2]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 2], vdata));
vdata = _mm_loadu_si128(&dat[3]);
vdata = _mm_loadu_si128((const __m128i *)&dat[3]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 3], vdata));
}
for (; ord < coef_order - 3; ord++) {
vdata = _mm_loadu_si128(&data[smpl - coef_order + ord]);
vdata = _mm_loadu_si128((const __m128i *)&data[smpl - coef_order + ord]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord], vdata));
}
_mm_store_si128(predict, vpred);
_mm_store_si128((__m128i *)predict, vpred);

/* ord = coef_order - 3 */
/* data[smpl + 0] .. data[smpl + 2]に依存関係があるため処理
Expand Down Expand Up @@ -141,28 +141,28 @@ void SRLALPC_Synthesize(
__m256i vpred = _mm256_set1_epi32(half);
for (ord = 0; ord < (int32_t)coef_order - 7 - 8; ord += 8) {
const int32_t *dat = &data[smpl - coef_order + ord];
vdata = _mm256_loadu_si256(&dat[0]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[0]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 0], vdata));
vdata = _mm256_loadu_si256(&dat[1]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[1]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 1], vdata));
vdata = _mm256_loadu_si256(&dat[2]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[2]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 2], vdata));
vdata = _mm256_loadu_si256(&dat[3]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[3]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 3], vdata));
vdata = _mm256_loadu_si256(&dat[4]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[4]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 4], vdata));
vdata = _mm256_loadu_si256(&dat[5]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[5]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 5], vdata));
vdata = _mm256_loadu_si256(&dat[6]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[6]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 6], vdata));
vdata = _mm256_loadu_si256(&dat[7]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[7]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 7], vdata));
}
for (; ord < coef_order - 7; ord++) {
vdata = _mm256_loadu_si256(&data[smpl - coef_order + ord]);
vdata = _mm256_loadu_si256((const __m256i *)&data[smpl - coef_order + ord]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord], vdata));
}
_mm256_store_si256(predict, vpred);
_mm256_store_si256((__m256i *)predict, vpred);

/* ord = coef_order - 7 */
for (i = 0; i < 8; i++) {
Expand Down Expand Up @@ -190,20 +190,20 @@ void SRLALPC_Synthesize(
__m128i vpred = _mm_set1_epi32(half);
for (ord = 0; ord < (int32_t)coef_order - 3 - 4; ord += 4) {
const int32_t *dat = &data[smpl - coef_order + ord];
vdata = _mm_loadu_si128(&dat[0]);
vdata = _mm_loadu_si128((const __m128i *)&dat[0]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 0], vdata));
vdata = _mm_loadu_si128(&dat[1]);
vdata = _mm_loadu_si128((const __m128i *)&dat[1]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 1], vdata));
vdata = _mm_loadu_si128(&dat[2]);
vdata = _mm_loadu_si128((const __m128i *)&dat[2]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 2], vdata));
vdata = _mm_loadu_si128(&dat[3]);
vdata = _mm_loadu_si128((const __m128i *)&dat[3]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 3], vdata));
}
for (; ord < coef_order - 3; ord++) {
vdata = _mm_loadu_si128(&data[smpl - coef_order + ord]);
vdata = _mm_loadu_si128((const __m128i *)&data[smpl - coef_order + ord]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord], vdata));
}
_mm_store_si128(predict, vpred);
_mm_store_si128((__m128i *)predict, vpred);

/* ord = coef_order - 3 */
for (i = 0; i < 4; i++) {
Expand Down
42 changes: 21 additions & 21 deletions libs/srla_encoder/src/srla_lpc_predict.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -54,20 +54,20 @@ void SRLALPC_Predict(
__m128i vpred = _mm_set1_epi32(half);
for (ord = 0; ord < (int32_t)coef_order - 3 - 4; ord += 4) {
const int32_t *dat = &data[smpl - coef_order + ord];
vdata = _mm_loadu_si128(&dat[0]);
vdata = _mm_loadu_si128((const __m128i *)&dat[0]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 0], vdata));
vdata = _mm_loadu_si128(&dat[1]);
vdata = _mm_loadu_si128((const __m128i *)&dat[1]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 1], vdata));
vdata = _mm_loadu_si128(&dat[2]);
vdata = _mm_loadu_si128((const __m128i *)&dat[2]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 2], vdata));
vdata = _mm_loadu_si128(&dat[3]);
vdata = _mm_loadu_si128((const __m128i *)&dat[3]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 3], vdata));
}
for (; ord < coef_order - 3; ord++) {
vdata = _mm_loadu_si128(&data[smpl - coef_order + ord]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord], vdata));
}
_mm_store_si128(predict, vpred);
_mm_store_si128((__m128i *)predict, vpred);

/* ord = coef_order - 3 */
/* data[smpl + 0] .. data[smpl + 2]に依存関係があるため処理
Expand Down Expand Up @@ -149,28 +149,28 @@ void SRLALPC_Predict(
__m256i vpred = _mm256_set1_epi32(half);
for (ord = 0; ord < (int32_t)coef_order - 7 - 8; ord += 8) {
const int32_t *dat = &data[smpl - coef_order + ord];
vdata = _mm256_loadu_si256(&dat[0]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[0]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 0], vdata));
vdata = _mm256_loadu_si256(&dat[1]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[1]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 1], vdata));
vdata = _mm256_loadu_si256(&dat[2]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[2]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 2], vdata));
vdata = _mm256_loadu_si256(&dat[3]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[3]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 3], vdata));
vdata = _mm256_loadu_si256(&dat[4]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[4]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 4], vdata));
vdata = _mm256_loadu_si256(&dat[5]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[5]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 5], vdata));
vdata = _mm256_loadu_si256(&dat[6]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[6]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 6], vdata));
vdata = _mm256_loadu_si256(&dat[7]);
vdata = _mm256_loadu_si256((const __m256i *)&dat[7]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord + 7], vdata));
}
for (; ord < coef_order - 7; ord++) {
vdata = _mm256_loadu_si256(&data[smpl - coef_order + ord]);
vdata = _mm256_loadu_si256((const __m256i *)&data[smpl - coef_order + ord]);
vpred = _mm256_add_epi32(vpred, _mm256_mullo_epi32(vcoef[ord], vdata));
}
_mm256_store_si256(predict, vpred);
_mm256_store_si256((__m256i *)predict, vpred);

/* ord = coef_order - 7 */
for (i = 0; i < 8; i++) {
Expand Down Expand Up @@ -198,20 +198,20 @@ void SRLALPC_Predict(
__m128i vpred = _mm_set1_epi32(half);
for (ord = 0; ord < (int32_t)coef_order - 3 - 4; ord += 4) {
const int32_t *dat = &data[smpl - coef_order + ord];
vdata = _mm_loadu_si128(&dat[0]);
vdata = _mm_loadu_si128((const __m128i *)&dat[0]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 0], vdata));
vdata = _mm_loadu_si128(&dat[1]);
vdata = _mm_loadu_si128((const __m128i *)&dat[1]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 1], vdata));
vdata = _mm_loadu_si128(&dat[2]);
vdata = _mm_loadu_si128((const __m128i *)&dat[2]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 2], vdata));
vdata = _mm_loadu_si128(&dat[3]);
vdata = _mm_loadu_si128((const __m128i *)&dat[3]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord + 3], vdata));
}
for (; ord < coef_order - 3; ord++) {
vdata = _mm_loadu_si128(&data[smpl - coef_order + ord]);
vdata = _mm_loadu_si128((const __m128i *)&data[smpl - coef_order + ord]);
vpred = _mm_add_epi32(vpred, _mm_mullo_epi32(vcoef[ord], vdata));
}
_mm_store_si128(predict, vpred);
_mm_store_si128((__m128i *)predict, vpred);

/* ord = coef_order - 3 */
for (i = 0; i < 4; i++) {
Expand Down

0 comments on commit 88b271f

Please sign in to comment.