---
 libavresample/x86/audio_convert.asm    |   60 ++++++++++++++++++++++++++++++++
 libavresample/x86/audio_convert_init.c |    9 +++++
 2 files changed, 69 insertions(+), 0 deletions(-)

diff --git a/libavresample/x86/audio_convert.asm 
b/libavresample/x86/audio_convert.asm
index 70519e1..4690419 100644
--- a/libavresample/x86/audio_convert.asm
+++ b/libavresample/x86/audio_convert.asm
@@ -772,3 +772,63 @@ CONV_S16_TO_S16P_2CH
 INIT_XMM avx
 CONV_S16_TO_S16P_2CH
 %endif
+
+;------------------------------------------------------------------------------
+; void ff_conv_s16_to_s16p_6ch(int16_t *const *dst, int16_t *src, int len,
+;                              int channels);
+;------------------------------------------------------------------------------
+
+%macro CONV_S16_TO_S16P_6CH 0
+cglobal conv_s16_to_s16p_6ch, 2,8,5, dst, src, dst1, dst2, dst3, dst4, dst5, 
len
+%if ARCH_X86_64
+    mov      lend, r2d
+%else
+    %define lend dword r2m
+%endif
+    mov     dst1q, [dstq+  gprsize]
+    mov     dst2q, [dstq+2*gprsize]
+    mov     dst3q, [dstq+3*gprsize]
+    mov     dst4q, [dstq+4*gprsize]
+    mov     dst5q, [dstq+5*gprsize]
+    mov      dstq, [dstq          ]
+    sub     dst1q, dstq
+    sub     dst2q, dstq
+    sub     dst3q, dstq
+    sub     dst4q, dstq
+    sub     dst5q, dstq
+    ALIGN 16
+.loop:
+    mova       m0, [srcq         ]      ; m0 =  0,  1,  2,  3,  4,  5,  6,  7
+    mova       m2, [srcq+2*mmsize]      ; m2 = 16, 17, 18, 19, 20, 21, 22, 23
+    movq       m3, [srcq+  mmsize+mmsize/2]
+    movlhps    m3, m2                   ; m3 = 12, 13, 14, 15, 16, 17, 18, 19
+    movhpd     m1, [srcq+  mmsize]
+    movhlps    m1, m0                   ; m1 =  4,  5,  6,  7,  8,  9, 10, 11
+    psrldq     m1, 4                    ; m1 =  6,  7,  8,  9, 10, 11,  x,  x
+    psrldq     m2, 4                    ; m2 = 18, 19, 20, 21, 22, 23,  x,  x
+    punpcklwd  m4, m0, m1               ; m4 =  0,  6,  1,  7,  2,  8,  3,  9
+    punpckhwd  m0, m1                   ; m0 =  4, 10,  5, 11,  x,  x,  x,  x
+    punpcklwd  m1, m3, m2               ; m1 = 12, 18, 13, 19, 14, 20, 15, 21
+    punpckhwd  m3, m2                   ; m3 = 16, 22, 17, 23,  x,  x,  x,  x
+    punpckldq  m2, m4, m1               ; m2 =  0,  6, 12, 18,  1,  7, 13, 19
+    punpckhdq  m4, m1                   ; m4 =  2,  8, 14, 20,  3,  9, 15, 21
+    punpckldq  m0, m3                   ; m0 =  4, 10, 16, 22,  5, 11, 17, 23
+    movq    [dstq      ], m2
+    movhps  [dstq+dst1q], m2
+    movq    [dstq+dst2q], m4
+    movhps  [dstq+dst3q], m4
+    movq    [dstq+dst4q], m0
+    movhps  [dstq+dst5q], m0
+    add      srcq, mmsize*3
+    add      dstq, mmsize/2
+    sub      lend, mmsize/4
+    jg .loop
+    REP_RET
+%endmacro
+
+INIT_XMM sse2
+CONV_S16_TO_S16P_6CH
+%if HAVE_AVX
+INIT_XMM avx
+CONV_S16_TO_S16P_6CH
+%endif
diff --git a/libavresample/x86/audio_convert_init.c 
b/libavresample/x86/audio_convert_init.c
index 65c6e10..e907aa4 100644
--- a/libavresample/x86/audio_convert_init.c
+++ b/libavresample/x86/audio_convert_init.c
@@ -93,6 +93,11 @@ extern void ff_conv_s16_to_s16p_2ch_sse2(int16_t *const 
*dst, int16_t *src,
 extern void ff_conv_s16_to_s16p_2ch_avx (int16_t *const *dst, int16_t *src,
                                          int len, int channels);
 
+extern void ff_conv_s16_to_s16p_6ch_sse2(int16_t *const *dst, int16_t *src,
+                                         int len, int channels);
+extern void ff_conv_s16_to_s16p_6ch_avx (int16_t *const *dst, int16_t *src,
+                                         int len, int channels);
+
 av_cold void ff_audio_convert_init_x86(AudioConvert *ac)
 {
 #if HAVE_YASM
@@ -139,6 +144,8 @@ av_cold void ff_audio_convert_init_x86(AudioConvert *ac)
                                   6, 16, 4, "SSE2", 
ff_conv_fltp_to_s16_6ch_sse2);
         ff_audio_convert_set_func(ac, AV_SAMPLE_FMT_S16P, AV_SAMPLE_FMT_S16,
                                   2, 16, 8, "SSE2", 
ff_conv_s16_to_s16p_2ch_sse2);
+        ff_audio_convert_set_func(ac, AV_SAMPLE_FMT_S16P, AV_SAMPLE_FMT_S16,
+                                  6, 16, 4, "SSE2", 
ff_conv_s16_to_s16p_6ch_sse2);
     }
     if (mm_flags & AV_CPU_FLAG_SSE4 && HAVE_SSE) {
         ff_audio_convert_set_func(ac, AV_SAMPLE_FMT_FLT, AV_SAMPLE_FMT_S16,
@@ -169,6 +176,8 @@ av_cold void ff_audio_convert_init_x86(AudioConvert *ac)
                                   6, 16, 4, "AVX", 
ff_conv_fltp_to_flt_6ch_avx);
         ff_audio_convert_set_func(ac, AV_SAMPLE_FMT_S16P, AV_SAMPLE_FMT_S16,
                                   2, 16, 8, "AVX", 
ff_conv_s16_to_s16p_2ch_avx);
+        ff_audio_convert_set_func(ac, AV_SAMPLE_FMT_S16P, AV_SAMPLE_FMT_S16,
+                                  6, 16, 4, "AVX", 
ff_conv_s16_to_s16p_6ch_avx);
     }
 #endif
 }
-- 
1.7.1

_______________________________________________
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to