[03/23] vp9lpf/x86: add x86 SSSE3/AVX SIMD for vp9_loop_filter_[vh]_16_16.

Message ID 1475436610-25770-3-git-send-email-anton@khirnov.net
State Committed
Commit a692724c587859c8157d1aefcae9fbdb23328c61
Headers show

Commit Message

Anton Khirnov Oct. 2, 2016, 7:29 p.m.
From: Clément Bœsch <u@pkh.me>

Signed-off-by: Anton Khirnov <anton@khirnov.net>
---
 libavcodec/x86/Makefile      |   3 +-
 libavcodec/x86/vp9dsp_init.c |  15 ++
 libavcodec/x86/vp9lpf.asm    | 601 +++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 618 insertions(+), 1 deletion(-)
 create mode 100644 libavcodec/x86/vp9lpf.asm

Patch

diff --git a/libavcodec/x86/Makefile b/libavcodec/x86/Makefile
index 4aae594..a38535b 100644
--- a/libavcodec/x86/Makefile
+++ b/libavcodec/x86/Makefile
@@ -125,4 +125,5 @@  YASM-OBJS-$(CONFIG_V210_ENCODER)       += x86/v210enc.o
 YASM-OBJS-$(CONFIG_VORBIS_DECODER)     += x86/vorbisdsp.o
 YASM-OBJS-$(CONFIG_VP3_DECODER)        += x86/hpeldsp_vp3.o
 YASM-OBJS-$(CONFIG_VP6_DECODER)        += x86/vp6dsp.o
-YASM-OBJS-$(CONFIG_VP9_DECODER)        += x86/vp9mc.o
+YASM-OBJS-$(CONFIG_VP9_DECODER)        += x86/vp9mc.o                   \
+                                          x86/vp9lpf.o
diff --git a/libavcodec/x86/vp9dsp_init.c b/libavcodec/x86/vp9dsp_init.c
index 41fa35a..93453b8 100644
--- a/libavcodec/x86/vp9dsp_init.c
+++ b/libavcodec/x86/vp9dsp_init.c
@@ -217,6 +217,11 @@  filters_8tap_1d_fn2(avg, 32, avx2, ssse3)
 #undef filters_8tap_1d_fn3
 #undef filter_8tap_1d_fn
 
+void ff_vp9_loop_filter_v_16_16_ssse3(uint8_t *dst, ptrdiff_t stride, int E, int I, int H);
+void ff_vp9_loop_filter_v_16_16_avx  (uint8_t *dst, ptrdiff_t stride, int E, int I, int H);
+void ff_vp9_loop_filter_h_16_16_ssse3(uint8_t *dst, ptrdiff_t stride, int E, int I, int H);
+void ff_vp9_loop_filter_h_16_16_avx  (uint8_t *dst, ptrdiff_t stride, int E, int I, int H);
+
 #endif /* HAVE_YASM */
 
 av_cold void ff_vp9dsp_init_x86(VP9DSPContext *dsp)
@@ -283,11 +288,21 @@  av_cold void ff_vp9dsp_init_x86(VP9DSPContext *dsp)
     if (EXTERNAL_SSSE3(cpu_flags)) {
         init_subpel3(0, put, ssse3);
         init_subpel3(1, avg, ssse3);
+
+        if (ARCH_X86_64) {
+            dsp->loop_filter_16[0] = ff_vp9_loop_filter_h_16_16_ssse3;
+            dsp->loop_filter_16[1] = ff_vp9_loop_filter_v_16_16_ssse3;
+        }
     }
 
     if (EXTERNAL_AVX(cpu_flags)) {
         init_fpel(1, 0, 32, put, avx);
         init_fpel(0, 0, 64, put, avx);
+
+        if (ARCH_X86_64) {
+            dsp->loop_filter_16[0] = ff_vp9_loop_filter_h_16_16_avx;
+            dsp->loop_filter_16[1] = ff_vp9_loop_filter_v_16_16_avx;
+        }
     }
 
     if (EXTERNAL_AVX2(cpu_flags)) {
diff --git a/libavcodec/x86/vp9lpf.asm b/libavcodec/x86/vp9lpf.asm
new file mode 100644
index 0000000..8568f1d
--- /dev/null
+++ b/libavcodec/x86/vp9lpf.asm
@@ -0,0 +1,601 @@ 
+;******************************************************************************
+;* VP9 loop filter SIMD optimizations
+;*
+;* Copyright (C) 2013-2014 Clément Bœsch <u pkh me>
+;*
+;* This file is part of Libav.
+;*
+;* Libav is free software; you can redistribute it and/or
+;* modify it under the terms of the GNU Lesser General Public
+;* License as published by the Free Software Foundation; either
+;* version 2.1 of the License, or (at your option) any later version.
+;*
+;* Libav is distributed in the hope that it will be useful,
+;* but WITHOUT ANY WARRANTY; without even the implied warranty of
+;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+;* Lesser General Public License for more details.
+;*
+;* You should have received a copy of the GNU Lesser General Public
+;* License along with Libav; if not, write to the Free Software
+;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+;******************************************************************************
+
+
+%include "libavutil/x86/x86util.asm"
+
+%if ARCH_X86_64
+
+SECTION_RODATA
+
+cextern pb_3
+cextern pb_80
+
+pb_4:   times 16 db 0x04
+pb_10:  times 16 db 0x10
+pb_40:  times 16 db 0x40
+pb_81:  times 16 db 0x81
+pb_f8:  times 16 db 0xf8
+pb_fe:  times 16 db 0xfe
+
+pw_4:   times  8 dw 4
+pw_8:   times  8 dw 8
+
+SECTION .text
+
+; %1 = abs(%2-%3)
+%macro ABSSUB 4 ; dst, src1 (RO), src2 (RO), tmp
+    psubusb             %1, %3, %2
+    psubusb             %4, %2, %3
+    por                 %1, %4
+%endmacro
+
+; %1 = %1<=%2
+%macro CMP_LTE 3-4 ; src/dst, cmp, tmp, pb_80
+%if %0 == 4
+    pxor                %1, %4
+%endif
+    pcmpgtb             %3, %2, %1          ; cmp > src?
+    pcmpeqb             %1, %2              ; cmp == src? XXX: avoid this with a -1/+1 well placed?
+    por                 %1, %3              ; cmp >= src?
+%endmacro
+
+; %1 = abs(%2-%3) <= %4
+%macro ABSSUB_CMP 6-7 [pb_80]; dst, src1, src2, cmp, tmp1, tmp2, [pb_80]
+    ABSSUB              %1, %2, %3, %6      ; dst = abs(src1-src2)
+    CMP_LTE             %1, %4, %6, %7      ; dst <= cmp
+%endmacro
+
+%macro MASK_APPLY 4 ; %1=new_data/dst %2=old_data %3=mask %4=tmp
+    pand                %1, %3              ; new &= mask
+    pandn               %4, %3, %2          ; tmp = ~mask & old
+    por                 %1, %4              ; new&mask | old&~mask
+%endmacro
+
+%macro FILTER_SUBx2_ADDx2 8 ; %1=dst %2=h/l %3=cache %4=sub1 %5=sub2 %6=add1 %7=add2 %8=rshift
+    punpck%2bw          %3, %4, m0
+    psubw               %1, %3
+    punpck%2bw          %3, %5, m0
+    psubw               %1, %3
+    punpck%2bw          %3, %6, m0
+    paddw               %1, %3
+    punpck%2bw          %3, %7, m0
+    paddw               %3, %1
+    psraw               %1, %3, %8
+%endmacro
+
+%macro FILTER_INIT 8 ; tmp1, tmp2, cacheL, cacheH, dstp, filterid, mask, source
+    FILTER%6_INIT       %1, l, %3
+    FILTER%6_INIT       %2, h, %4
+    packuswb            %1, %2
+    MASK_APPLY          %1, %8, %7, %2
+    mova                %5, %1
+%endmacro
+
+%macro FILTER_UPDATE 11-14 ; tmp1, tmp2, cacheL, cacheH, dstp, -, -, +, +, rshift, mask, [source], [preload reg + value]
+%if %0 == 13 ; no source + preload
+    mova                %12, %13
+%elif %0 == 14 ; source + preload
+    mova                %13, %14
+%endif
+    FILTER_SUBx2_ADDx2  %1, l, %3, %6, %7, %8, %9, %10
+    FILTER_SUBx2_ADDx2  %2, h, %4, %6, %7, %8, %9, %10
+    packuswb            %1, %2
+%if %0 == 12 || %0 == 14
+    MASK_APPLY          %1, %12, %11, %2
+%else
+    MASK_APPLY          %1, %5, %11, %2
+%endif
+    mova                %5, %1
+%endmacro
+
+%macro SRSHIFT3B_2X 4 ; reg1, reg2, [pb_10], tmp
+    mova                %4, [pb_f8]
+    pand                %1, %4
+    pand                %2, %4
+    psrlq               %1, 3
+    psrlq               %2, 3
+    pxor                %1, %3
+    pxor                %2, %3
+    psubb               %1, %3
+    psubb               %2, %3
+%endmacro
+
+%macro EXTRACT_POS_NEG 3 ; i8, neg, pos
+    pxor                %3, %3
+    pxor                %2, %2
+    pcmpgtb             %3, %1                          ; i8 < 0 mask
+    psubb               %2, %1                          ; neg values (only the originally - will be kept)
+    pand                %2, %3                          ; negative values of i8 (but stored as +)
+    pandn               %3, %1                          ; positive values of i8
+%endmacro
+
+; clip_u8(u8 + i8)
+%macro SIGN_ADD 5 ; dst, u8, i8, tmp1, tmp2
+    EXTRACT_POS_NEG     %3, %4, %5
+    psubusb             %1, %2, %4                      ; sub the negatives
+    paddusb             %1, %5                          ; add the positives
+%endmacro
+
+; clip_u8(u8 - i8)
+%macro SIGN_SUB 5 ; dst, u8, i8, tmp1, tmp2
+    EXTRACT_POS_NEG     %3, %4, %5
+    psubusb             %1, %2, %5                      ; sub the positives
+    paddusb             %1, %4                          ; add the negatives
+%endmacro
+
+%macro FILTER6_INIT 3 ; %1=dst %2=h/l %3=cache
+    punpck%2bw          %1, m14, m0                     ; p3: B->W
+    paddw               %3, %1, %1                      ; p3*2
+    paddw               %3, %1                          ; p3*3
+    punpck%2bw          %1, m15, m0                     ; p2: B->W
+    paddw               %3, %1                          ; p3*3 + p2
+    paddw               %3, %1                          ; p3*3 + p2*2
+    punpck%2bw          %1, m10, m0                     ; p1: B->W
+    paddw               %3, %1                          ; p3*3 + p2*2 + p1
+    punpck%2bw          %1, m11, m0                     ; p0: B->W
+    paddw               %3, %1                          ; p3*3 + p2*2 + p1 + p0
+    punpck%2bw          %1, m12, m0                     ; q0: B->W
+    paddw               %3, %1                          ; p3*3 + p2*2 + p1 + p0 + q0
+    paddw               %3, [pw_4]                      ; p3*3 + p2*2 + p1 + p0 + q0 + 4
+    psraw               %1, %3, 3                       ; (p3*3 + p2*2 + p1 + p0 + q0 + 4) >> 3
+%endmacro
+
+%macro FILTER14_INIT 3 ; %1=dst %2=h/l %3=cache
+    punpck%2bw          %1, m2, m0                      ; p7: B->W
+    psllw               %3, %1, 3                       ; p7*8
+    psubw               %3, %1                          ; p7*7
+    punpck%2bw          %1, m3, m0                      ; p6: B->W
+    paddw               %3, %1                          ; p7*7 + p6
+    paddw               %3, %1                          ; p7*7 + p6*2
+    punpck%2bw          %1, m8, m0                      ; p5: B->W
+    paddw               %3, %1                          ; p7*7 + p6*2 + p5
+    punpck%2bw          %1, m9, m0                      ; p4: B->W
+    paddw               %3, %1                          ; p7*7 + p6*2 + p5 + p4
+    punpck%2bw          %1, m14, m0                     ; p3: B->W
+    paddw               %3, %1                          ; p7*7 + p6*2 + p5 + p4 + p3
+    punpck%2bw          %1, m15, m0                     ; p2: B->W
+    paddw               %3, %1                          ; p7*7 + p6*2 + p5 + .. + p2
+    punpck%2bw          %1, m10, m0                     ; p1: B->W
+    paddw               %3, %1                          ; p7*7 + p6*2 + p5 + .. + p1
+    punpck%2bw          %1, m11, m0                     ; p0: B->W
+    paddw               %3, %1                          ; p7*7 + p6*2 + p5 + .. + p0
+    punpck%2bw          %1, m12, m0                     ; q0: B->W
+    paddw               %3, %1                          ; p7*7 + p6*2 + p5 + .. + p0 + q0
+    paddw               %3, [pw_8]                      ; p7*7 + p6*2 + p5 + .. + p0 + q0 + 8
+    psraw               %1, %3, 4                       ; (p7*7 + p6*2 + p5 + .. + p0 + q0 + 8) >> 4
+%endmacro
+
+%macro TRANSPOSE16x16B 17
+    mova %17, m%16
+    SBUTTERFLY bw,  %1,  %2,  %16
+    SBUTTERFLY bw,  %3,  %4,  %16
+    SBUTTERFLY bw,  %5,  %6,  %16
+    SBUTTERFLY bw,  %7,  %8,  %16
+    SBUTTERFLY bw,  %9,  %10, %16
+    SBUTTERFLY bw,  %11, %12, %16
+    SBUTTERFLY bw,  %13, %14, %16
+    mova m%16,  %17
+    mova  %17, m%14
+    SBUTTERFLY bw,  %15, %16, %14
+    SBUTTERFLY wd,  %1,  %3,  %14
+    SBUTTERFLY wd,  %2,  %4,  %14
+    SBUTTERFLY wd,  %5,  %7,  %14
+    SBUTTERFLY wd,  %6,  %8,  %14
+    SBUTTERFLY wd,  %9,  %11, %14
+    SBUTTERFLY wd,  %10, %12, %14
+    SBUTTERFLY wd,  %13, %15, %14
+    mova m%14,  %17
+    mova  %17, m%12
+    SBUTTERFLY wd,  %14, %16, %12
+    SBUTTERFLY dq,  %1,  %5,  %12
+    SBUTTERFLY dq,  %2,  %6,  %12
+    SBUTTERFLY dq,  %3,  %7,  %12
+    SBUTTERFLY dq,  %4,  %8,  %12
+    SBUTTERFLY dq,  %9,  %13, %12
+    SBUTTERFLY dq,  %10, %14, %12
+    SBUTTERFLY dq,  %11, %15, %12
+    mova m%12, %17
+    mova  %17, m%8
+    SBUTTERFLY dq,  %12, %16, %8
+    SBUTTERFLY qdq, %1,  %9,  %8
+    SBUTTERFLY qdq, %2,  %10, %8
+    SBUTTERFLY qdq, %3,  %11, %8
+    SBUTTERFLY qdq, %4,  %12, %8
+    SBUTTERFLY qdq, %5,  %13, %8
+    SBUTTERFLY qdq, %6,  %14, %8
+    SBUTTERFLY qdq, %7,  %15, %8
+    mova m%8, %17
+    mova %17, m%1
+    SBUTTERFLY qdq, %8,  %16, %1
+    mova m%1, %17
+    SWAP %2,  %9
+    SWAP %3,  %5
+    SWAP %4,  %13
+    SWAP %6,  %11
+    SWAP %8,  %15
+    SWAP %12, %14
+%endmacro
+
+%macro DEFINE_REAL_P7_TO_Q7 0
+%define P7 dst1q + 2*mstrideq
+%define P6 dst1q +   mstrideq
+%define P5 dst1q
+%define P4 dst1q +    strideq
+%define P3 dstq  + 4*mstrideq
+%define P2 dstq  +   mstride3q
+%define P1 dstq  + 2*mstrideq
+%define P0 dstq  +   mstrideq
+%define Q0 dstq
+%define Q1 dstq  +   strideq
+%define Q2 dstq  + 2*strideq
+%define Q3 dstq  +   stride3q
+%define Q4 dstq  + 4*strideq
+%define Q5 dst2q + mstrideq
+%define Q6 dst2q
+%define Q7 dst2q +  strideq
+%endmacro
+
+%macro LPF_16_16 1
+    lea mstrideq, [strideq]
+    neg mstrideq
+
+    lea stride3q, [strideq+2*strideq]
+    mov mstride3q, stride3q
+    neg mstride3q
+
+%ifidn %1, h
+    lea dstq, [dstq + 8*strideq - 8] ; go from top center (h pos) to center left (v pos)
+%endif
+
+    lea dst1q, [dstq + 2*mstride3q]                         ; dst1q = &dst[stride * -6]
+    lea dst2q, [dstq + 2* stride3q]                         ; dst2q = &dst[stride * +6]
+
+    DEFINE_REAL_P7_TO_Q7
+
+%ifidn %1, h
+    movu                    m0, [P7]
+    movu                    m1, [P6]
+    movu                    m2, [P5]
+    movu                    m3, [P4]
+    movu                    m4, [P3]
+    movu                    m5, [P2]
+    movu                    m6, [P1]
+    movu                    m7, [P0]
+    movu                    m8, [Q0]
+    movu                    m9, [Q1]
+    movu                   m10, [Q2]
+    movu                   m11, [Q3]
+    movu                   m12, [Q4]
+    movu                   m13, [Q5]
+    movu                   m14, [Q6]
+    movu                   m15, [Q7]
+    TRANSPOSE16x16B 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, [rsp]
+
+%define P7 rsp +   0
+%define P6 rsp +  16
+%define P5 rsp +  32
+%define P4 rsp +  48
+%define P3 rsp +  64
+%define P2 rsp +  80
+%define P1 rsp +  96
+%define P0 rsp + 112
+%define Q0 rsp + 128
+%define Q1 rsp + 144
+%define Q2 rsp + 160
+%define Q3 rsp + 176
+%define Q4 rsp + 192
+%define Q5 rsp + 208
+%define Q6 rsp + 224
+%define Q7 rsp + 240
+
+    mova           [P7],  m0
+    mova           [P6],  m1
+    mova           [P5],  m2
+    mova           [P4],  m3
+    mova           [P3],  m4
+    mova           [P2],  m5
+    mova           [P1],  m6
+    mova           [P0],  m7
+    mova           [Q0],  m8
+    mova           [Q1],  m9
+    mova           [Q2], m10
+    mova           [Q3], m11
+    mova           [Q4], m12
+    mova           [Q5], m13
+    mova           [Q6], m14
+    mova           [Q7], m15
+%endif
+
+    ; calc fm mask
+    pxor                m0, m0
+    movd                m2, Id
+    movd                m3, Ed
+    pshufb              m2, m0                          ; I I I I ...
+    pshufb              m3, m0                          ; E E E E ...
+    mova                m0, [pb_80]
+    pxor                m2, m0
+    pxor                m3, m0
+%ifidn %1, v
+    mova                m8, [P3]
+    mova                m9, [P2]
+    mova               m10, [P1]
+    mova               m11, [P0]
+    mova               m12, [Q0]
+    mova               m13, [Q1]
+    mova               m14, [Q2]
+    mova               m15, [Q3]
+%else
+    ; In case of horizontal, P3..Q3 are already present in some registers due
+    ; to the previous transpose, so we just swap registers.
+    SWAP                 8,  4, 12
+    SWAP                 9,  5, 13
+    SWAP                10,  6, 14
+    SWAP                11,  7, 15
+%endif
+    ABSSUB_CMP          m5,  m8,  m9, m2, m6, m7, m0    ; m5 = abs(p3-p2) <= I
+    ABSSUB_CMP          m1,  m9, m10, m2, m6, m7, m0    ; m1 = abs(p2-p1) <= I
+    pand                m5, m1
+    ABSSUB_CMP          m1, m10, m11, m2, m6, m7, m0    ; m1 = abs(p1-p0) <= I
+    pand                m5, m1
+    ABSSUB_CMP          m1, m12, m13, m2, m6, m7, m0    ; m1 = abs(q1-q0) <= I
+    pand                m5, m1
+    ABSSUB_CMP          m1, m13, m14, m2, m6, m7, m0    ; m1 = abs(q2-q1) <= I
+    pand                m5, m1
+    ABSSUB_CMP          m1, m14, m15, m2, m6, m7, m0    ; m1 = abs(q3-q2) <= I
+    pand                m5, m1
+    ABSSUB              m1, m11, m12, m7                ; abs(p0-q0)
+    paddusb             m1, m1                          ; abs(p0-q0) * 2
+    ABSSUB              m2, m10, m13, m7                ; abs(p1-q1)
+    pand                m2, [pb_fe]                     ; drop lsb so shift can work
+    psrlq               m2, 1                           ; abs(p1-q1)/2
+    paddusb             m1, m2                          ; abs(p0-q0)*2 + abs(p1-q1)/2
+    pxor                m1, m0
+    pcmpgtb             m4, m3, m1                      ; E > X?
+    pcmpeqb             m3, m1                          ; E == X?
+    por                 m3, m4                          ; E >= X?
+    pand                m3, m5                          ; fm final value
+
+    ; (m3: fm, m8..15: p3 p2 p1 p0 q0 q1 q2 q3)
+    ; calc flat8in and hev masks
+    mova                m6, [pb_81]                     ; [1 1 1 1 ...] ^ 0x80
+    ABSSUB_CMP          m2, m8, m11, m6, m4, m5         ; abs(p3 - p0) <= 1
+    mova                m8, [pb_80]
+    ABSSUB_CMP          m1, m9, m11, m6, m4, m5, m8     ; abs(p2 - p0) <= 1
+    pand                m2, m1
+    ABSSUB              m4, m10, m11, m5                ; abs(p1 - p0)
+    pxor                m0, m0
+    movd                m7, Hd
+    pshufb              m7, m0                          ; H H H H ...
+    pxor                m7, m8
+    pxor                m4, m8
+    pcmpgtb             m0, m4, m7                      ; abs(p1 - p0) > H (1/2 hev condition)
+    CMP_LTE             m4, m6, m5                      ; abs(p1 - p0) <= 1
+    pand                m2, m4                          ; (flat8in)
+    ABSSUB              m4, m13, m12, m1                ; abs(q1 - q0)
+    pxor                m4, m8
+    pcmpgtb             m5, m4, m7                      ; abs(q1 - q0) > H (2/2 hev condition)
+    por                 m0, m5                          ; hev final value
+    CMP_LTE             m4, m6, m5                      ; abs(q1 - q0) <= 1
+    pand                m2, m4                          ; (flat8in)
+    ABSSUB_CMP          m1, m14, m12, m6, m4, m5, m8    ; abs(q2 - q0) <= 1
+    pand                m2, m1
+    ABSSUB_CMP          m1, m15, m12, m6, m4, m5, m8    ; abs(q3 - q0) <= 1
+    pand                m2, m1                          ; flat8in final value
+
+    ; (m0: hev, m2: flat8in, m3: fm, m6: pb_81, m9..15: p2 p1 p0 q0 q1 q2 q3)
+    ; calc flat8out mask
+    mova                m8, [P7]
+    mova                m9, [P6]
+    ABSSUB_CMP          m1, m8, m11, m6, m4, m5         ; abs(p7 - p0) <= 1
+    ABSSUB_CMP          m7, m9, m11, m6, m4, m5         ; abs(p6 - p0) <= 1
+    pand                m1, m7
+    mova                m8, [P5]
+    mova                m9, [P4]
+    ABSSUB_CMP          m7, m8, m11, m6, m4, m5         ; abs(p5 - p0) <= 1
+    pand                m1, m7
+    ABSSUB_CMP          m7, m9, m11, m6, m4, m5         ; abs(p4 - p0) <= 1
+    pand                m1, m7
+    mova                m14, [Q4]
+    mova                m15, [Q5]
+    ABSSUB_CMP          m7, m14, m12, m6, m4, m5        ; abs(q4 - q0) <= 1
+    pand                m1, m7
+    ABSSUB_CMP          m7, m15, m12, m6, m4, m5        ; abs(q5 - q0) <= 1
+    pand                m1, m7
+    mova                m14, [Q6]
+    mova                m15, [Q7]
+    ABSSUB_CMP          m7, m14, m12, m6, m4, m5        ; abs(q4 - q0) <= 1
+    pand                m1, m7
+    ABSSUB_CMP          m7, m15, m12, m6, m4, m5        ; abs(q5 - q0) <= 1
+    pand                m1, m7                          ; flat8out final value
+
+    ; if (fm) {
+    ;     if (out && in) filter_14()
+    ;     else if (in)   filter_6()
+    ;     else if (hev)  filter_2()
+    ;     else           filter_4()
+    ; }
+    ;
+    ; f14:                                                                            fm &  out &  in
+    ; f6:  fm & ~f14 & in        => fm & ~(out & in) & in                          => fm & ~out &  in
+    ; f2:  fm & ~f14 & ~f6 & hev => fm & ~(out & in) & ~(~out & in) & hev          => fm &  ~in &  hev
+    ; f4:  fm & ~f14 & ~f6 & ~f2 => fm & ~(out & in) & ~(~out & in) & ~(~in & hev) => fm &  ~in & ~hev
+
+    ; (m0: hev, m1: flat8out, m2: flat8in, m3: fm, m8..15: p5 p4 p1 p0 q0 q1 q6 q7)
+    ; filter2()
+    mova                m6, [pb_80]
+    pxor                m15, m12, m6                    ; q0 ^ 0x80
+    pxor                m14, m11, m6                    ; p0 ^ 0x80
+    psubsb              m15, m14                        ; (signed) q0 - p0
+    pxor                m4, m10, m6                     ; p1 ^ 0x80
+    pxor                m5, m13, m6                     ; q1 ^ 0x80
+    psubsb              m4, m5                          ; (signed) p1 - q1
+    paddsb              m4, m15                         ;   (q0 - p0) + (p1 - q1)
+    paddsb              m4, m15                         ; 2*(q0 - p0) + (p1 - q1)
+    paddsb              m4, m15                         ; 3*(q0 - p0) + (p1 - q1)
+    paddsb              m6, m4, [pb_4]                  ; m6: f1 = clip(f + 4, 127)
+    paddsb              m4, [pb_3]                      ; m4: f2 = clip(f + 3, 127)
+    mova                m14, [pb_10]                    ; will be reused in filter4()
+    SRSHIFT3B_2X        m6, m4, m14, m7                 ; f1 and f2 sign byte shift by 3
+    SIGN_SUB            m7, m12, m6, m5, m9             ; m7 = q0 - f1
+    SIGN_ADD            m8, m11, m4, m5, m9             ; m8 = p0 + f2
+    pandn               m6, m2, m3                      ;  ~mask(in) & mask(fm)
+    pand                m6, m0                          ; (~mask(in) & mask(fm)) & mask(hev)
+    MASK_APPLY          m7, m12, m6, m5                 ; m7 = filter2(q0) & mask / we write it in filter4()
+    MASK_APPLY          m8, m11, m6, m5                 ; m8 = filter2(p0) & mask / we write it in filter4()
+
+    ; (m0: hev, m1: flat8out, m2: flat8in, m3: fm, m7..m8: q0' p0', m10..13: p1 p0 q0 q1, m14: pb_10, m15: q0-p0)
+    ; filter4()
+    mova                m4, m15
+    paddsb              m15, m4                         ; 2 * (q0 - p0)
+    paddsb              m15, m4                         ; 3 * (q0 - p0)
+    paddsb              m6, m15, [pb_4]                 ; m6:  f1 = clip(f + 4, 127)
+    paddsb              m15, [pb_3]                     ; m15: f2 = clip(f + 3, 127)
+    SRSHIFT3B_2X        m6, m15, m14, m9                ; f1 and f2 sign byte shift by 3
+    pandn               m5, m2, m3                      ;               ~mask(in) & mask(fm)
+    pandn               m0, m5                          ; ~mask(hev) & (~mask(in) & mask(fm))
+    SIGN_SUB            m9, m12, m6, m4, m14            ; q0 - f1
+    MASK_APPLY          m9, m7, m0, m5                  ; m9 = filter4(q0) & mask
+    mova                [Q0], m9
+    SIGN_ADD            m7, m11, m15, m4, m14           ; p0 + f2
+    MASK_APPLY          m7, m8, m0, m5                  ; m7 = filter4(p0) & mask
+    mova                [P0], m7
+    paddb               m6, [pb_80]                     ;
+    pxor                m8, m8                          ;   f=(f1+1)>>1
+    pavgb               m6, m8                          ;
+    psubb               m6, [pb_40]                     ;
+    SIGN_ADD            m7, m10, m6, m8, m9             ; p1 + f
+    SIGN_SUB            m4, m13, m6, m8, m9             ; q1 - f
+    MASK_APPLY          m7, m10, m0, m14                ; m7 = filter4(p1)
+    MASK_APPLY          m4, m13, m0, m14                ; m4 = filter4(q1)
+    mova                [P1], m7
+    mova                [Q1], m4
+
+    ; (m1: flat8out, m2: flat8in, m3: fm, m10..13: p1 p0 q0 q1)
+    ; filter6()
+    pxor                m0, m0
+    pand                m2, m3                          ;               mask(fm) & mask(in)
+    pandn               m3, m1, m2                      ; ~mask(out) & (mask(fm) & mask(in))
+    mova               m14, [P3]
+    mova               m15, [P2]
+    mova                m8, [Q2]
+    mova                m9, [Q3]
+    FILTER_INIT         m4, m5, m6, m7, [P2], 6,                     m3, m15    ; [p2]
+    FILTER_UPDATE       m6, m7, m4, m5, [P1], m14, m15, m10, m13, 3, m3         ; [p1] -p3 -p2 +p1 +q1
+    FILTER_UPDATE       m4, m5, m6, m7, [P0], m14, m10, m11,  m8, 3, m3         ; [p0] -p3 -p1 +p0 +q2
+    FILTER_UPDATE       m6, m7, m4, m5, [Q0], m14, m11, m12,  m9, 3, m3         ; [q0] -p3 -p0 +q0 +q3
+    FILTER_UPDATE       m4, m5, m6, m7, [Q1], m15, m12, m13,  m9, 3, m3         ; [q1] -p2 -q0 +q1 +q3
+    FILTER_UPDATE       m6, m7, m4, m5, [Q2], m10, m13,  m8,  m9, 3, m3,  m8    ; [q2] -p1 -q1 +q2 +q3
+
+    ; (m0: 0, m1: flat8out, m2: fm & flat8in, m8..15: q2 q3 p1 p0 q0 q1 p3 p2)
+    ; filter14()
+    ;
+    ;                            m2  m3  m8  m9 m14 m15 m10 m11 m12 m13
+    ;
+    ;                                    q2  q3  p3  p2  p1  p0  q0  q1
+    ; p6  -7                     p7  p6  p5  p4   .   .   .   .   .
+    ; p5  -6  -p7 -p6 +p5 +q1     .   .   .                           .
+    ; p4  -5  -p7 -p5 +p4 +q2     .       .   .                      q2
+    ; p3  -4  -p7 -p4 +p3 +q3     .           .   .                  q3
+    ; p2  -3  -p7 -p3 +p2 +q4     .               .   .              q4
+    ; p1  -2  -p7 -p2 +p1 +q5     .                   .   .          q5
+    ; p0  -1  -p7 -p1 +p0 +q6     .                       .   .      q6
+    ; q0  +0  -p7 -p0 +q0 +q7     .                           .   .  q7
+    ; q1  +1  -p6 -q0 +q1 +q7    q1   .                           .   .
+    ; q2  +2  -p5 -q1 +q2 +q7     .  q2   .                           .
+    ; q3  +3  -p4 -q2 +q3 +q7         .  q3   .                       .
+    ; q4  +4  -p3 -q3 +q4 +q7             .  q4   .                   .
+    ; q5  +5  -p2 -q4 +q5 +q7                 .  q5   .               .
+    ; q6  +6  -p1 -q5 +q6 +q7                     .  q6   .           .
+
+    pand            m1, m2                                                              ; mask(out) & (mask(fm) & mask(in))
+    mova            m2, [P7]
+    mova            m3, [P6]
+    mova            m8, [P5]
+    mova            m9, [P4]
+    FILTER_INIT     m4, m5, m6, m7, [P6],  14,                   m1,  m3
+    FILTER_UPDATE   m6, m7, m4, m5, [P5],  m2,  m3,  m8, m13, 4, m1,  m8                ; [p5] -p7 -p6 +p5 +q1
+    FILTER_UPDATE   m4, m5, m6, m7, [P4],  m2,  m8,  m9, m13, 4, m1,  m9, m13, [Q2]     ; [p4] -p7 -p5 +p4 +q2
+    FILTER_UPDATE   m6, m7, m4, m5, [P3],  m2,  m9, m14, m13, 4, m1, m14, m13, [Q3]     ; [p3] -p7 -p4 +p3 +q3
+    FILTER_UPDATE   m4, m5, m6, m7, [P2],  m2, m14, m15, m13, 4, m1,      m13, [Q4]     ; [p2] -p7 -p3 +p2 +q4
+    FILTER_UPDATE   m6, m7, m4, m5, [P1],  m2, m15, m10, m13, 4, m1,      m13, [Q5]     ; [p1] -p7 -p2 +p1 +q5
+    FILTER_UPDATE   m4, m5, m6, m7, [P0],  m2, m10, m11, m13, 4, m1,      m13, [Q6]     ; [p0] -p7 -p1 +p0 +q6
+    FILTER_UPDATE   m6, m7, m4, m5, [Q0],  m2, m11, m12, m13, 4, m1,      m13, [Q7]     ; [q0] -p7 -p0 +q0 +q7
+    FILTER_UPDATE   m4, m5, m6, m7, [Q1],  m3, m12,  m2, m13, 4, m1,       m2, [Q1]     ; [q1] -p6 -q0 +q1 +q7
+    FILTER_UPDATE   m6, m7, m4, m5, [Q2],  m8,  m2,  m3, m13, 4, m1,       m3, [Q2]     ; [q2] -p5 -q1 +q2 +q7
+    FILTER_UPDATE   m4, m5, m6, m7, [Q3],  m9,  m3,  m8, m13, 4, m1,  m8,  m8, [Q3]     ; [q3] -p4 -q2 +q3 +q7
+    FILTER_UPDATE   m6, m7, m4, m5, [Q4], m14,  m8,  m9, m13, 4, m1,  m9,  m9, [Q4]     ; [q4] -p3 -q3 +q4 +q7
+    FILTER_UPDATE   m4, m5, m6, m7, [Q5], m15,  m9, m14, m13, 4, m1, m14, m14, [Q5]     ; [q5] -p2 -q4 +q5 +q7
+    FILTER_UPDATE   m6, m7, m4, m5, [Q6], m10, m14, m15, m13, 4, m1, m15, m15, [Q6]     ; [q6] -p1 -q5 +q6 +q7
+
+%ifidn %1, h
+    mova                    m0, [P7]
+    mova                    m1, [P6]
+    mova                    m2, [P5]
+    mova                    m3, [P4]
+    mova                    m4, [P3]
+    mova                    m5, [P2]
+    mova                    m6, [P1]
+    mova                    m7, [P0]
+    mova                    m8, [Q0]
+    mova                    m9, [Q1]
+    mova                   m10, [Q2]
+    mova                   m11, [Q3]
+    mova                   m12, [Q4]
+    mova                   m13, [Q5]
+    mova                   m14, [Q6]
+    mova                   m15, [Q7]
+    TRANSPOSE16x16B 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, [rsp]
+    DEFINE_REAL_P7_TO_Q7
+    movu  [P7],  m0
+    movu  [P6],  m1
+    movu  [P5],  m2
+    movu  [P4],  m3
+    movu  [P3],  m4
+    movu  [P2],  m5
+    movu  [P1],  m6
+    movu  [P0],  m7
+    movu  [Q0],  m8
+    movu  [Q1],  m9
+    movu  [Q2], m10
+    movu  [Q3], m11
+    movu  [Q4], m12
+    movu  [Q5], m13
+    movu  [Q6], m14
+    movu  [Q7], m15
+%endif
+%endmacro
+
+%macro LPF_16_16_VH 1
+INIT_XMM %1
+cglobal vp9_loop_filter_v_16_16, 5,10,16,      dst, stride, E, I, H, mstride, dst1, dst2, stride3, mstride3
+    LPF_16_16 v
+    RET
+cglobal vp9_loop_filter_h_16_16, 5,10,16, 256, dst, stride, E, I, H, mstride, dst1, dst2, stride3, mstride3
+    LPF_16_16 h
+    RET
+%endmacro
+
+LPF_16_16_VH ssse3
+LPF_16_16_VH avx
+
+%endif ; x86-64