diff options
| author | Michael Niedermayer <michaelni@gmx.at> | 2013-01-23 17:31:55 +0400 |
|---|---|---|
| committer | Michael Niedermayer <michaelni@gmx.at> | 2013-01-23 17:31:55 +0400 |
| commit | 8102f27b5b3dff54f8099019c2df4701ac5e5d4f (patch) | |
| tree | f623e5063c24e6c4afc50338dfeecf7054c673a2 /libavutil/x86 | |
| parent | 24604ebaf85b5436d10da811f455dd710b353aca (diff) | |
| parent | 73b704ac609d83e0be124589f24efd9b94947cf9 (diff) | |
Merge commit '73b704ac609d83e0be124589f24efd9b94947cf9'
* commit '73b704ac609d83e0be124589f24efd9b94947cf9':
arm: Add some missing header #includes
floatdsp: move scalarproduct_float from dsputil to avfloatdsp.
Conflicts:
libavcodec/acelp_pitch_delay.c
libavcodec/amrnbdec.c
libavcodec/amrwbdec.c
libavcodec/ra288.c
libavcodec/x86/dsputil_mmx.c
libavutil/x86/float_dsp.asm
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Diffstat (limited to 'libavutil/x86')
| -rw-r--r-- | libavutil/x86/float_dsp.asm | 28 | ||||
| -rw-r--r-- | libavutil/x86/float_dsp_init.c | 3 |
2 files changed, 30 insertions, 1 deletions
diff --git a/libavutil/x86/float_dsp.asm b/libavutil/x86/float_dsp.asm index 3e5e91ad07..004e6cf1fe 100644 --- a/libavutil/x86/float_dsp.asm +++ b/libavutil/x86/float_dsp.asm @@ -236,4 +236,30 @@ VECTOR_FMUL_REVERSE %if HAVE_AVX_EXTERNAL INIT_YMM avx VECTOR_FMUL_REVERSE -%endif
\ No newline at end of file +%endif + +; float scalarproduct_float_sse(const float *v1, const float *v2, int len) +INIT_XMM sse +cglobal scalarproduct_float, 3,3,2, v1, v2, offset + neg offsetq + shl offsetq, 2 + sub v1q, offsetq + sub v2q, offsetq + xorps xmm0, xmm0 +.loop: + movaps xmm1, [v1q+offsetq] + mulps xmm1, [v2q+offsetq] + addps xmm0, xmm1 + add offsetq, 16 + js .loop + movhlps xmm1, xmm0 + addps xmm0, xmm1 + movss xmm1, xmm0 + shufps xmm0, xmm0, 1 + addss xmm0, xmm1 +%if ARCH_X86_64 == 0 + movss r0m, xmm0 + fld dword r0m +%endif + RET + diff --git a/libavutil/x86/float_dsp_init.c b/libavutil/x86/float_dsp_init.c index 9c58e2bc30..5c6383bc74 100644 --- a/libavutil/x86/float_dsp_init.c +++ b/libavutil/x86/float_dsp_init.c @@ -51,6 +51,8 @@ void ff_vector_fmul_reverse_sse(float *dst, const float *src0, void ff_vector_fmul_reverse_avx(float *dst, const float *src0, const float *src1, int len); +float ff_scalarproduct_float_sse(const float *v1, const float *v2, int order); + #if HAVE_6REGS && HAVE_INLINE_ASM static void vector_fmul_window_3dnowext(float *dst, const float *src0, const float *src1, const float *win, @@ -135,6 +137,7 @@ void ff_float_dsp_init_x86(AVFloatDSPContext *fdsp) fdsp->vector_fmul_scalar = ff_vector_fmul_scalar_sse; fdsp->vector_fmul_add = ff_vector_fmul_add_sse; fdsp->vector_fmul_reverse = ff_vector_fmul_reverse_sse; + fdsp->scalarproduct_float = ff_scalarproduct_float_sse; } if (EXTERNAL_SSE2(mm_flags)) { fdsp->vector_dmul_scalar = ff_vector_dmul_scalar_sse2; |