libvpx: Pull from upstream

source/libvpx/vp8/common/x86/postproc_mmx.asm

 ;
 ;  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
 ;
 ;  Use of this source code is governed by a BSD-style license
 ;  that can be found in the LICENSE file in the root of the source
 ;  tree. An additional intellectual property rights grant can be found
 ;  in the file PATENTS.  All contributing project authors may
 ;  be found in the AUTHORS file in the root of the source tree.
 ;
 
@@ skipping 43 matching lines @@
             ; this copies the last row down into the border 8 rows
             mov         rdi,        rsi
             mov         rdx,        arg(2)
             sub         rdx,        9
             imul        rdx,        rax
             lea         rdi,        [rdi+rdx]
             movq        mm1,        QWORD ptr[rdi]              ; first row
             mov         rcx,        8
 .init_borderd                                                    ; initialize borders
             lea         rdi,        [rdi + rax]
-            movq        [rdi],      xmm1
+            movq        [rdi],      mm1
 
             dec         rcx
             jne         .init_borderd
 
             neg         rax                                     ; rax = -pitch
 
             ; this copies the first row up into the border 8 rows
             mov         rdi,        rsi
             movq        mm1,        QWORD ptr[rdi]              ; first row
             mov         rcx,        8
@@ skipping 111 matching lines @@
             push        rax
             lea         rax,        [GLOBAL(sym(vp8_rv))]
             movq        mm4,        [rax + rcx*2] ;vp8_rv[rcx*2]
             pop         rax
 %elif ABI_IS_32BIT=0
             movq        mm4,        [r8 + rcx*2] ;vp8_rv[rcx*2]
 %else
             movq        mm4,        [sym(vp8_rv) + rcx*2]
 %endif
             paddw       mm1,        mm4
-            ;paddw     xmm1,       eight8s
             psraw       mm1,        4
 
             packuswb    mm1,        mm0
             pand        mm1,        mm3
 
             pandn       mm3,        mm2
             por         mm1,        mm3
 
             and         rcx,        15
             movd        DWORD PTR   [rsp+rcx*4], mm1 ;d[rcx*4]
@@ skipping 98 matching lines @@
 SECTION_RODATA
 align 16
 Blur:
     times 16 dw 16
     times  8 dw 64
     times 16 dw 16
     times  8 dw  0
 
 rd:
     times 4 dw 0x40