libyuv/source/rotate_any.cc
Manojkumar Bhosale 6fa5e4eb78 Add MSA optimized TransposeWx8_MSA and TransposeUVWx8_MSA functions
R=fbarchard@google.com
BUG=libyuv:634

Performance Gain (vs C vectorized)
TransposeWx8_MSA          - ~2.7x
TransposeWx8_Any_MSA      - ~2.1x
TransposeUVWx8_MSA        - ~2.5x
TransposeUVWx8_Any_MSA    - ~2.7x

Performance Gain (vs C non-vectorized)
TransposeWx8_MSA          - ~4.6x
TransposeWx8_Any_MSA      - ~2.9x
TransposeUVWx8_MSA        - ~4.4x
TransposeUVWx8_Any_MSA    - ~3.7x

Review URL: https://codereview.chromium.org/2553403002 .
2016-12-15 10:06:01 +05:30

79 lines
3.0 KiB
C++

/*
* Copyright 2015 The LibYuv Project Authors. All rights reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include "libyuv/rotate.h"
#include "libyuv/rotate_row.h"
#include "libyuv/basic_types.h"
#ifdef __cplusplus
namespace libyuv {
extern "C" {
#endif
#define TANY(NAMEANY, TPOS_SIMD, MASK) \
void NAMEANY(const uint8* src, int src_stride, uint8* dst, int dst_stride, \
int width) { \
int r = width & MASK; \
int n = width - r; \
if (n > 0) { \
TPOS_SIMD(src, src_stride, dst, dst_stride, n); \
} \
TransposeWx8_C(src + n, src_stride, dst + n * dst_stride, dst_stride, r); \
}
#ifdef HAS_TRANSPOSEWX8_NEON
TANY(TransposeWx8_Any_NEON, TransposeWx8_NEON, 7)
#endif
#ifdef HAS_TRANSPOSEWX8_SSSE3
TANY(TransposeWx8_Any_SSSE3, TransposeWx8_SSSE3, 7)
#endif
#ifdef HAS_TRANSPOSEWX8_FAST_SSSE3
TANY(TransposeWx8_Fast_Any_SSSE3, TransposeWx8_Fast_SSSE3, 15)
#endif
#ifdef HAS_TRANSPOSEWX8_DSPR2
TANY(TransposeWx8_Any_DSPR2, TransposeWx8_DSPR2, 7)
#endif
#ifdef HAS_TRANSPOSEWX8_MSA
TANY(TransposeWx8_Any_MSA, TransposeWx8_MSA, 15)
#endif
#undef TANY
#define TUVANY(NAMEANY, TPOS_SIMD, MASK) \
void NAMEANY(const uint8* src, int src_stride, uint8* dst_a, \
int dst_stride_a, uint8* dst_b, int dst_stride_b, int width) { \
int r = width & MASK; \
int n = width - r; \
if (n > 0) { \
TPOS_SIMD(src, src_stride, dst_a, dst_stride_a, dst_b, dst_stride_b, n); \
} \
TransposeUVWx8_C(src + n * 2, src_stride, dst_a + n * dst_stride_a, \
dst_stride_a, dst_b + n * dst_stride_b, dst_stride_b, r); \
}
#ifdef HAS_TRANSPOSEUVWX8_NEON
TUVANY(TransposeUVWx8_Any_NEON, TransposeUVWx8_NEON, 7)
#endif
#ifdef HAS_TRANSPOSEUVWX8_SSE2
TUVANY(TransposeUVWx8_Any_SSE2, TransposeUVWx8_SSE2, 7)
#endif
#ifdef HAS_TRANSPOSEUVWX8_DSPR2
TUVANY(TransposeUVWx8_Any_DSPR2, TransposeUVWx8_DSPR2, 7)
#endif
#ifdef HAS_TRANSPOSEUVWX8_MSA
TUVANY(TransposeUVWx8_Any_MSA, TransposeUVWx8_MSA, 7)
#endif
#undef TUVANY
#ifdef __cplusplus
} // extern "C"
} // namespace libyuv
#endif