[FFmpeg-devel] [PATCH 5/6] diracdec: avx2 dd97

[James Darnley]

---
 libavcodec/x86/dirac_dwt_10bit.asm    |  3 ++-
 libavcodec/x86/dirac_dwt_init_10bit.c | 13 +++++++++++++
 2 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/libavcodec/x86/dirac_dwt_10bit.asm b/libavcodec/x86/dirac_dwt_10bit.asm
index ae110d2945..2e039e11ea 100644
--- a/libavcodec/x86/dirac_dwt_10bit.asm
+++ b/libavcodec/x86/dirac_dwt_10bit.asm
@@ -25,7 +25,7 @@ SECTION_RODATA
 
 cextern pd_1
 pd_2: times 8 dd 2
-pd_8: times 4 dd 8
+pd_8: times 8 dd 8
 
 SECTION .text
 
@@ -202,6 +202,7 @@ HAAR_HORIZONTAL
 HAAR_VERTICAL
 
 INIT_YMM avx2
+DD97_VERTICAL_HI
 HAAR_HORIZONTAL
 HAAR_VERTICAL
 LEGALL53_VERTICAL_HI
diff --git a/libavcodec/x86/dirac_dwt_init_10bit.c b/libavcodec/x86/dirac_dwt_init_10bit.c
index 51d6eeae93..f103a56176 100644
--- a/libavcodec/x86/dirac_dwt_init_10bit.c
+++ b/libavcodec/x86/dirac_dwt_init_10bit.c
@@ -24,6 +24,7 @@
 #include "libavcodec/dirac_dwt.h"
 
 void ff_dd97_vertical_hi_sse2(int32_t *b0, int32_t *b1, int32_t *b2, int32_t *b3, int32_t *b4, int width);
+void ff_dd97_vertical_hi_avx2(int32_t *b0, int32_t *b1, int32_t *b2, int32_t *b3, int32_t *b4, int width);
 
 void ff_legall53_vertical_hi_sse2(int32_t *b0, int32_t *b1, int32_t *b2, int width);
 void ff_legall53_vertical_lo_sse2(int32_t *b0, int32_t *b1, int32_t *b2, int width);
@@ -137,7 +138,15 @@ static void dd97_vertical_hi_sse2(int32_t *b0, int32_t *b1, int32_t *b2,
     ff_dd97_vertical_hi_sse2(b0, b1, b2, b3, b4, i);
     for(; i<width; i++)
         b2[i] = COMPOSE_DD97iH0(b0[i], b1[i], b2[i], b3[i], b4[i]);
+}
 
+static void dd97_vertical_hi_avx2(int32_t *b0, int32_t *b1, int32_t *b2,
+                                  int32_t *b3, int32_t *b4, int width)
+{
+    int i = width & ~7;
+    ff_dd97_vertical_hi_avx2(b0, b1, b2, b3, b4, i);
+    for(; i<width; i++)
+        b2[i] = COMPOSE_DD97iH0(b0[i], b1[i], b2[i], b3[i], b4[i]);
 }
 
 av_cold void ff_spatial_idwt_init_10bit_x86(DWTContext *d, enum dwt_type type)
@@ -179,6 +188,10 @@ av_cold void ff_spatial_idwt_init_10bit_x86(DWTContext *d, enum dwt_type type)
 
     if (EXTERNAL_AVX2(cpu_flags)) {
         switch (type) {
+            case DWT_DIRAC_DD9_7:
+                d->vertical_compose_h0 = (void*)dd97_vertical_hi_avx2;
+                d->vertical_compose_l0 = (void*)legall53_vertical_lo_avx2;
+                break;
             case DWT_DIRAC_LEGALL5_3:
                 d->vertical_compose_h0 = (void*)legall53_vertical_hi_avx2;
                 d->vertical_compose_l0 = (void*)legall53_vertical_lo_avx2;
-- 
2.17.1