Merging upstream version 127.0.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

8 files changed, 45 insertions, 136 deletions

diff --git a/media/ffvpx/libavcodec/x86/h264_intrapred.asm b/media/ffvpx/libavcodec/x86/h264_intrapred.asm
index 8a38ba2bb5..a8a630dbe6 100644
--- a/media/ffvpx/libavcodec/x86/h264_intrapred.asm
+++ b/media/ffvpx/libavcodec/x86/h264_intrapred.asm
@@ -86,8 +86,6 @@ cglobal pred16x16_horizontal_8, 2,3
     punpcklbw m1, m1
     SPLATW    m0, m0, 3
     SPLATW    m1, m1, 3
-    mova [r0+r1*0+8], m0
-    mova [r0+r1*1+8], m1
 %endif
 
     mova [r0+r1*0], m0
@@ -98,7 +96,7 @@ cglobal pred16x16_horizontal_8, 2,3
     RET
 %endmacro
 
-INIT_MMX mmxext
+INIT_XMM sse2
 PRED16x16_H
 INIT_XMM ssse3
 PRED16x16_H
@@ -568,17 +566,17 @@ H264_PRED8x8_PLANE
 ; void ff_pred8x8_vertical_8(uint8_t *src, ptrdiff_t stride)
 ;-----------------------------------------------------------------------------
 
-INIT_MMX mmx
+INIT_XMM sse2
 cglobal pred8x8_vertical_8, 2,2
     sub    r0, r1
-    movq  mm0, [r0]
+    movq   m0, [r0]
 %rep 3
-    movq [r0+r1*1], mm0
-    movq [r0+r1*2], mm0
+    movq [r0+r1*1], m0
+    movq [r0+r1*2], m0
     lea    r0, [r0+r1*2]
 %endrep
-    movq [r0+r1*1], mm0
-    movq [r0+r1*2], mm0
+    movq [r0+r1*1], m0
+    movq [r0+r1*2], m0
     RET
 
 ;-----------------------------------------------------------------------------
@@ -1313,10 +1311,7 @@ PRED8x8L_DOWN_RIGHT
 ;-----------------------------------------------------------------------------
 
 %macro PRED8x8L_VERTICAL_RIGHT 0
-cglobal pred8x8l_vertical_right_8, 4,5,7
-    ; manually spill XMM registers for Win64 because
-    ; the code here is initialized with INIT_MMX
-    WIN64_SPILL_XMM 7
+cglobal pred8x8l_vertical_right_8, 4,5,6
     sub          r0, r3
     lea          r4, [r0+r3*2]
     movq        mm0, [r0+r3*1-8]
@@ -1386,7 +1381,6 @@ cglobal pred8x8l_vertical_right_8, 4,5,7
     movq2dq     xmm4, mm6
     pslldq      xmm4, 8
     por         xmm0, xmm4
-    movdqa      xmm6, [pw_ff00]
     movdqa      xmm1, xmm0
     lea           r2, [r1+r3*2]
     movdqa      xmm2, xmm0
@@ -1396,15 +1390,16 @@ cglobal pred8x8l_vertical_right_8, 4,5,7
     pavgb       xmm2, xmm0
 INIT_XMM cpuname
     PRED4x4_LOWPASS xmm4, xmm3, xmm1, xmm0, xmm5
-    pandn       xmm6, xmm4
+    movdqa      xmm0, [pw_ff00]
+    pandn       xmm0, xmm4
     movdqa      xmm5, xmm4
     psrlw       xmm4, 8
-    packuswb    xmm6, xmm4
-    movhlps     xmm4, xmm6
+    packuswb    xmm0, xmm4
+    movhlps     xmm4, xmm0
     movhps [r0+r3*2], xmm5
     movhps [r0+r3*1], xmm2
     psrldq      xmm5, 4
-    movss       xmm5, xmm6
+    movss       xmm5, xmm0
     psrldq      xmm2, 4
     movss       xmm2, xmm4
     lea           r0, [r2+r3*2]
diff --git a/media/ffvpx/libavcodec/x86/h264_intrapred_init.c b/media/ffvpx/libavcodec/x86/h264_intrapred_init.c
index ee46927a24..aa9bc721f0 100644
--- a/media/ffvpx/libavcodec/x86/h264_intrapred_init.c
+++ b/media/ffvpx/libavcodec/x86/h264_intrapred_init.c
@@ -100,7 +100,7 @@ PRED16x16(horizontal, 10, sse2)
 
 /* 8-bit versions */
 PRED16x16(vertical, 8, sse)
-PRED16x16(horizontal, 8, mmxext)
+PRED16x16(horizontal, 8, sse2)
 PRED16x16(horizontal, 8, ssse3)
 PRED16x16(dc, 8, sse2)
 PRED16x16(dc, 8, ssse3)
@@ -116,7 +116,7 @@ PRED16x16(tm_vp8, 8, avx2)
 PRED8x8(top_dc, 8, mmxext)
 PRED8x8(dc_rv40, 8, mmxext)
 PRED8x8(dc, 8, mmxext)
-PRED8x8(vertical, 8, mmx)
+PRED8x8(vertical, 8, sse2)
 PRED8x8(horizontal, 8, mmxext)
 PRED8x8(horizontal, 8, ssse3)
 PRED8x8(plane, 8, sse2)
@@ -163,14 +163,7 @@ av_cold void ff_h264_pred_init_x86(H264PredContext *h, int codec_id,
     int cpu_flags = av_get_cpu_flags();
 
     if (bit_depth == 8) {
-        if (EXTERNAL_MMX(cpu_flags)) {
-            if (chroma_format_idc <= 1) {
-                h->pred8x8  [VERT_PRED8x8     ] = ff_pred8x8_vertical_8_mmx;
-            }
-        }
-
         if (EXTERNAL_MMXEXT(cpu_flags)) {
-            h->pred16x16[HOR_PRED8x8            ] = ff_pred16x16_horizontal_8_mmxext;
             if (chroma_format_idc <= 1)
                 h->pred8x8[HOR_PRED8x8          ] = ff_pred8x8_horizontal_8_mmxext;
             h->pred8x8l [TOP_DC_PRED            ] = ff_pred8x8l_top_dc_8_mmxext;
@@ -210,12 +203,15 @@ av_cold void ff_h264_pred_init_x86(H264PredContext *h, int codec_id,
         }
 
         if (EXTERNAL_SSE2(cpu_flags)) {
+            h->pred16x16[HOR_PRED8x8          ] = ff_pred16x16_horizontal_8_sse2;
             h->pred16x16[DC_PRED8x8           ] = ff_pred16x16_dc_8_sse2;
             h->pred8x8l [DIAG_DOWN_LEFT_PRED  ] = ff_pred8x8l_down_left_8_sse2;
             h->pred8x8l [DIAG_DOWN_RIGHT_PRED ] = ff_pred8x8l_down_right_8_sse2;
             h->pred8x8l [VERT_RIGHT_PRED      ] = ff_pred8x8l_vertical_right_8_sse2;
             h->pred8x8l [VERT_LEFT_PRED       ] = ff_pred8x8l_vertical_left_8_sse2;
             h->pred8x8l [HOR_DOWN_PRED        ] = ff_pred8x8l_horizontal_down_8_sse2;
+            if (chroma_format_idc <= 1)
+                h->pred8x8  [VERT_PRED8x8     ] = ff_pred8x8_vertical_8_sse2;
             if (codec_id == AV_CODEC_ID_VP7 || codec_id == AV_CODEC_ID_VP8) {
                 h->pred16x16[PLANE_PRED8x8    ] = ff_pred16x16_tm_vp8_8_sse2;
                 h->pred8x8  [PLANE_PRED8x8    ] = ff_pred8x8_tm_vp8_8_sse2;
diff --git a/media/ffvpx/libavcodec/x86/moz.build b/media/ffvpx/libavcodec/x86/moz.build
index bd721affad..c358b5fdb5 100644
--- a/media/ffvpx/libavcodec/x86/moz.build
+++ b/media/ffvpx/libavcodec/x86/moz.build
@@ -37,6 +37,8 @@ SOURCES += [
     'vp9mc_16bpp.asm',
 ]
 
+LOCAL_INCLUDES += [ "../" ]
+
 if CONFIG['TARGET_CPU'] == 'x86':
     SOURCES += [ 'simple_idct.asm' ]
 
diff --git a/media/ffvpx/libavcodec/x86/simple_idct.asm b/media/ffvpx/libavcodec/x86/simple_idct.asm
index 982b2f0bbb..c79519372a 100644
--- a/media/ffvpx/libavcodec/x86/simple_idct.asm
+++ b/media/ffvpx/libavcodec/x86/simple_idct.asm
@@ -783,68 +783,33 @@ SECTION .text
 %macro PUT_PIXELS_CLAMPED_HALF 1
     mova     m0, [blockq+mmsize*0+%1]
     mova     m1, [blockq+mmsize*2+%1]
-%if mmsize == 8
-    mova     m2, [blockq+mmsize*4+%1]
-    mova     m3, [blockq+mmsize*6+%1]
-%endif
     packuswb m0, [blockq+mmsize*1+%1]
     packuswb m1, [blockq+mmsize*3+%1]
-%if mmsize == 8
-    packuswb m2, [blockq+mmsize*5+%1]
-    packuswb m3, [blockq+mmsize*7+%1]
-    movq           [pixelsq], m0
-    movq    [lsizeq+pixelsq], m1
-    movq  [2*lsizeq+pixelsq], m2
-    movq   [lsize3q+pixelsq], m3
-%else
     movq           [pixelsq], m0
     movhps  [lsizeq+pixelsq], m0
     movq  [2*lsizeq+pixelsq], m1
     movhps [lsize3q+pixelsq], m1
-%endif
 %endmacro
 
 %macro ADD_PIXELS_CLAMPED 1
     mova       m0, [blockq+mmsize*0+%1]
     mova       m1, [blockq+mmsize*1+%1]
-%if mmsize == 8
-    mova       m5, [blockq+mmsize*2+%1]
-    mova       m6, [blockq+mmsize*3+%1]
-%endif
     movq       m2, [pixelsq]
     movq       m3, [pixelsq+lsizeq]
-%if mmsize == 8
-    mova       m7, m2
-    punpcklbw  m2, m4
-    punpckhbw  m7, m4
-    paddsw     m0, m2
-    paddsw     m1, m7
-    mova       m7, m3
-    punpcklbw  m3, m4
-    punpckhbw  m7, m4
-    paddsw     m5, m3
-    paddsw     m6, m7
-%else
     punpcklbw  m2, m4
     punpcklbw  m3, m4
     paddsw     m0, m2
     paddsw     m1, m3
-%endif
     packuswb   m0, m1
-%if mmsize == 8
-    packuswb   m5, m6
-    movq       [pixelsq], m0
-    movq       [pixelsq+lsizeq], m5
-%else
     movq       [pixelsq], m0
     movhps     [pixelsq+lsizeq], m0
-%endif
 %endmacro
 
 INIT_MMX mmx
 
 cglobal simple_idct, 1, 2, 8, 128, block, t0
     IDCT
+    emms
 RET
 
 INIT_XMM sse2
diff --git a/media/ffvpx/libavcodec/x86/vp56_arith.h b/media/ffvpx/libavcodec/x86/vp56_arith.h
deleted file mode 100644
index 9f7639980c..0000000000
--- a/media/ffvpx/libavcodec/x86/vp56_arith.h
+++ /dev/null
@@ -1,53 +0,0 @@
-/**
- * VP5 and VP6 compatible video decoder (arith decoder)
- *
- * Copyright (C) 2006  Aurelien Jacobs <aurel@gnuage.org>
- * Copyright (C) 2010  Eli Friedman
- *
- * This file is part of FFmpeg.
- *
- * FFmpeg 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.
- *
- * FFmpeg 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 FFmpeg; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- */
-
-#ifndef AVCODEC_X86_VP56_ARITH_H
-#define AVCODEC_X86_VP56_ARITH_H
-
-#if HAVE_INLINE_ASM && HAVE_FAST_CMOV && HAVE_6REGS
-#include "libavutil/attributes.h"
-
-#define vp56_rac_get_prob vp56_rac_get_prob
-static av_always_inline int vp56_rac_get_prob(VP56RangeCoder *c, uint8_t prob)
-{
-    unsigned int code_word = vp56_rac_renorm(c);
-    unsigned int low = 1 + (((c->high - 1) * prob) >> 8);
-    unsigned int low_shift = low << 16;
-    int bit = 0;
-    c->code_word = code_word;
-
-    __asm__(
-        "subl  %4, %1      \n\t"
-        "subl  %3, %2      \n\t"
-        "setae %b0         \n\t"
-        "cmovb %4, %1      \n\t"
-        "cmovb %5, %2      \n\t"
-        : "+q"(bit), "+&r"(c->high), "+&r"(c->code_word)
-        : "r"(low_shift), "r"(low), "r"(code_word)
-    );
-
-    return bit;
-}
-#endif
-
-#endif /* AVCODEC_X86_VP56_ARITH_H */
diff --git a/media/ffvpx/libavcodec/x86/vp8dsp.asm b/media/ffvpx/libavcodec/x86/vp8dsp.asm
index 6ac5a7721b..231c21ea0d 100644
--- a/media/ffvpx/libavcodec/x86/vp8dsp.asm
+++ b/media/ffvpx/libavcodec/x86/vp8dsp.asm
@@ -114,7 +114,7 @@ bilinear_filter_vb_m: times 8 db 7, 1
                       times 8 db 2, 6
                       times 8 db 1, 7
 
-%ifdef PIC
+%if PIC
 %define fourtap_filter_hw  picregq
 %define sixtap_filter_hw   picregq
 %define fourtap_filter_hb  picregq
@@ -166,7 +166,7 @@ cglobal put_vp8_epel%1_h6, 6, 6 + npicregs, 8, dst, dststride, src, srcstride, h
     lea      mxd, [mxq*3]
     mova      m3, [filter_h6_shuf2]
     mova      m4, [filter_h6_shuf3]
-%ifdef PIC
+%if PIC
     lea  picregq, [sixtap_filter_hb_m]
 %endif
     mova      m5, [sixtap_filter_hb+mxq*8-48] ; set up 6tap filter in bytes
@@ -207,7 +207,7 @@ cglobal put_vp8_epel%1_h4, 6, 6 + npicregs, 7, dst, dststride, src, srcstride, h
     mova      m2, [pw_256]
     mova      m3, [filter_h2_shuf]
     mova      m4, [filter_h4_shuf]
-%ifdef PIC
+%if PIC
     lea  picregq, [fourtap_filter_hb_m]
 %endif
     mova      m5, [fourtap_filter_hb+mxq-16] ; set up 4tap filter in bytes
@@ -234,7 +234,7 @@ cglobal put_vp8_epel%1_h4, 6, 6 + npicregs, 7, dst, dststride, src, srcstride, h
 
 cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my
     shl      myd, 4
-%ifdef PIC
+%if PIC
     lea  picregq, [fourtap_filter_hb_m]
 %endif
     mova      m5, [fourtap_filter_hb+myq-16]
@@ -272,7 +272,7 @@ cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picr
 
 cglobal put_vp8_epel%1_v6, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my
     lea      myd, [myq*3]
-%ifdef PIC
+%if PIC
     lea  picregq, [sixtap_filter_hb_m]
 %endif
     lea      myq, [sixtap_filter_hb+myq*8]
@@ -326,7 +326,7 @@ FILTER_SSSE3 8
 INIT_MMX mmxext
 cglobal put_vp8_epel4_h4, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, height, mx, picreg
     shl       mxd, 4
-%ifdef PIC
+%if PIC
     lea   picregq, [fourtap_filter_hw_m]
 %endif
     movq      mm4, [fourtap_filter_hw+mxq-16] ; set up 4tap filter in words
@@ -374,7 +374,7 @@ cglobal put_vp8_epel4_h4, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, he
 INIT_MMX mmxext
 cglobal put_vp8_epel4_h6, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, height, mx, picreg
     lea       mxd, [mxq*3]
-%ifdef PIC
+%if PIC
     lea   picregq, [sixtap_filter_hw_m]
 %endif
     movq      mm4, [sixtap_filter_hw+mxq*8-48] ; set up 4tap filter in words
@@ -431,7 +431,7 @@ cglobal put_vp8_epel4_h6, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, he
 INIT_XMM sse2
 cglobal put_vp8_epel8_h4, 6, 6 + npicregs, 10, dst, dststride, src, srcstride, height, mx, picreg
     shl      mxd, 5
-%ifdef PIC
+%if PIC
     lea  picregq, [fourtap_filter_v_m]
 %endif
     lea      mxq, [fourtap_filter_v+mxq-32]
@@ -480,7 +480,7 @@ INIT_XMM sse2
 cglobal put_vp8_epel8_h6, 6, 6 + npicregs, 14, dst, dststride, src, srcstride, height, mx, picreg
     lea      mxd, [mxq*3]
     shl      mxd, 4
-%ifdef PIC
+%if PIC
     lea  picregq, [sixtap_filter_v_m]
 %endif
     lea      mxq, [sixtap_filter_v+mxq-96]
@@ -543,7 +543,7 @@ cglobal put_vp8_epel8_h6, 6, 6 + npicregs, 14, dst, dststride, src, srcstride, h
 ; 4x4 block, V-only 4-tap filter
 cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my
     shl      myd, 5
-%ifdef PIC
+%if PIC
     lea  picregq, [fourtap_filter_v_m]
 %endif
     lea      myq, [fourtap_filter_v+myq-32]
@@ -597,7 +597,7 @@ cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picr
 cglobal put_vp8_epel%1_v6, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my
     shl      myd, 4
     lea      myq, [myq*3]
-%ifdef PIC
+%if PIC
     lea  picregq, [sixtap_filter_v_m]
 %endif
     lea      myq, [sixtap_filter_v+myq-96]
@@ -667,7 +667,7 @@ FILTER_V 8
 %if cpuflag(ssse3)
 cglobal put_vp8_bilinear%1_v, 7, 7, 5, dst, dststride, src, srcstride, height, picreg, my
     shl      myd, 4
-%ifdef PIC
+%if PIC
     lea  picregq, [bilinear_filter_vb_m]
 %endif
     pxor      m4, m4
@@ -697,7 +697,7 @@ cglobal put_vp8_bilinear%1_v, 7, 7, 5, dst, dststride, src, srcstride, height, p
 %else ; cpuflag(ssse3)
 cglobal put_vp8_bilinear%1_v, 7, 7, 7, dst, dststride, src, srcstride, height, picreg, my
     shl      myd, 4
-%ifdef PIC
+%if PIC
     lea  picregq, [bilinear_filter_vw_m]
 %endif
     pxor      m6, m6
@@ -743,7 +743,7 @@ cglobal put_vp8_bilinear%1_v, 7, 7, 7, dst, dststride, src, srcstride, height, p
 %if cpuflag(ssse3)
 cglobal put_vp8_bilinear%1_h, 6, 6 + npicregs, 5, dst, dststride, src, srcstride, height, mx, picreg
     shl      mxd, 4
-%ifdef PIC
+%if PIC
     lea  picregq, [bilinear_filter_vb_m]
 %endif
     pxor      m4, m4
@@ -773,7 +773,7 @@ cglobal put_vp8_bilinear%1_h, 6, 6 + npicregs, 5, dst, dststride, src, srcstride
 %else ; cpuflag(ssse3)
 cglobal put_vp8_bilinear%1_h, 6, 6 + npicregs, 7, dst, dststride, src, srcstride, height, mx, picreg
     shl      mxd, 4
-%ifdef PIC
+%if PIC
     lea  picregq, [bilinear_filter_vw_m]
 %endif
     pxor      m6, m6
diff --git a/media/ffvpx/libavcodec/x86/vp9itxfm.asm b/media/ffvpx/libavcodec/x86/vp9itxfm.asm
index 2c63fe514a..2f290f2f88 100644
--- a/media/ffvpx/libavcodec/x86/vp9itxfm.asm
+++ b/media/ffvpx/libavcodec/x86/vp9itxfm.asm
@@ -330,7 +330,9 @@ IDCT_4x4_FN ssse3
 INIT_MMX %5
 cglobal vp9_%1_%3_4x4_add, 3, 3, 0, dst, stride, block, eob
 %if WIN64 && notcpuflag(ssse3)
+INIT_XMM cpuname
     WIN64_SPILL_XMM 8
+INIT_MMX cpuname
 %endif
     movdqa            xmm5, [pd_8192]
     mova                m0, [blockq+ 0]
diff --git a/media/ffvpx/libavcodec/x86/vp9itxfm_16bpp.asm b/media/ffvpx/libavcodec/x86/vp9itxfm_16bpp.asm
index 902685edf6..ebe6222285 100644
--- a/media/ffvpx/libavcodec/x86/vp9itxfm_16bpp.asm
+++ b/media/ffvpx/libavcodec/x86/vp9itxfm_16bpp.asm
@@ -303,7 +303,9 @@ IDCT4_10_FN
 %macro IADST4_FN 4
 cglobal vp9_%1_%3_4x4_add_10, 3, 3, 0, dst, stride, block, eob
 %if WIN64 && notcpuflag(ssse3)
+INIT_XMM cpuname
     WIN64_SPILL_XMM 8
+INIT_MMX cpuname
 %endif
     movdqa            xmm5, [pd_8192]
     mova                m0, [blockq+0*16+0]
@@ -672,7 +674,7 @@ cglobal vp9_idct_idct_8x8_add_10, 4, 6 + ARCH_X86_64, 14, \
     mov            dstbakq, dstq
     movsxd            cntq, cntd
 %endif
-%ifdef PIC
+%if PIC
     lea               ptrq, [default_8x8]
     movzx             cntd, byte [ptrq+cntq-1]
 %else
@@ -921,7 +923,7 @@ cglobal vp9_%1_%3_8x8_add_10, 4, 6 + ARCH_X86_64, 16, \
     mov            dstbakq, dstq
     movsxd            cntq, cntd
 %endif
-%ifdef PIC
+%if PIC
     lea               ptrq, [%5_8x8]
     movzx             cntd, byte [ptrq+cntq-1]
 %else
@@ -1128,7 +1130,7 @@ cglobal vp9_idct_idct_16x16_add_10, 4, 6 + ARCH_X86_64, 16, \
     mov            dstbakq, dstq
     movsxd            cntq, cntd
 %endif
-%ifdef PIC
+%if PIC
     lea               ptrq, [default_16x16]
     movzx             cntd, byte [ptrq+cntq-1]
 %else
@@ -1445,7 +1447,7 @@ cglobal vp9_%1_%4_16x16_add_10, 4, 6 + ARCH_X86_64, 16, \
     mov            dstbakq, dstq
     movsxd            cntq, cntd
 %endif
-%ifdef PIC
+%if PIC
     lea               ptrq, [%7_16x16]
     movzx             cntd, byte [ptrq+cntq-1]
 %else
@@ -1958,7 +1960,7 @@ cglobal vp9_idct_idct_32x32_add_10, 4, 6 + ARCH_X86_64, 16, \
     mov            dstbakq, dstq
     movsxd            cntq, cntd
 %endif
-%ifdef PIC
+%if PIC
     lea               ptrq, [default_32x32]
     movzx             cntd, byte [ptrq+cntq-1]
 %else