libvpx: Pull from upstream

source/libvpx/vp9/common/x86/vp9_high_intrapred_sse2.asm

 ;
 ;  Copyright (c) 2014 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.
 ;
 
 %include "third_party/x86inc/x86inc.asm"
 
 SECTION_RODATA
 pw_4:  times 8 dw 4
 pw_8:  times 8 dw 8
 pw_16: times 4 dd 16
 pw_32: times 4 dd 32
 
 SECTION .text
 INIT_MMX sse
-cglobal high_dc_predictor_4x4, 4, 5, 4, dst, stride, above, left, goffset
+cglobal highbd_dc_predictor_4x4, 4, 5, 4, dst, stride, above, left, goffset
   GET_GOT     goffsetq
 
   movq                  m0, [aboveq]
   movq                  m2, [leftq]
   DEFINE_ARGS dst, stride, one
   mov                 oned, 0x0001
   pxor                  m1, m1
   movd                  m3, oned
   pshufw                m3, m3, 0x0
   paddw                 m0, m2
   pmaddwd               m0, m3
   packssdw              m0, m1
   pmaddwd               m0, m3
   paddw                 m0, [GLOBAL(pw_4)]
   psraw                 m0, 3
   pshufw                m0, m0, 0x0
   movq    [dstq          ], m0
   movq    [dstq+strideq*2], m0
   lea                 dstq, [dstq+strideq*4]
   movq    [dstq          ], m0
   movq    [dstq+strideq*2], m0
 
   RESTORE_GOT
   RET
 
 INIT_XMM sse2
-cglobal high_dc_predictor_8x8, 4, 5, 4, dst, stride, above, left, goffset
+cglobal highbd_dc_predictor_8x8, 4, 5, 4, dst, stride, above, left, goffset
   GET_GOT     goffsetq
 
   pxor                  m1, m1
   mova                  m0, [aboveq]
   mova                  m2, [leftq]
   DEFINE_ARGS dst, stride, stride3, one
   mov                 oned, 0x00010001
   lea             stride3q, [strideq*3]
   movd                  m3, oned
   pshufd                m3, m3, 0x0
@@ skipping 14 matching lines @@
   lea                 dstq, [dstq+strideq*8]
   mova   [dstq           ], m0
   mova   [dstq+strideq*2 ], m0
   mova   [dstq+strideq*4 ], m0
   mova   [dstq+stride3q*2], m0
 
   RESTORE_GOT
   RET
 
 INIT_XMM sse2
-cglobal high_dc_predictor_16x16, 4, 5, 5, dst, stride, above, left, goffset
+cglobal highbd_dc_predictor_16x16, 4, 5, 5, dst, stride, above, left, goffset
   GET_GOT     goffsetq
 
   pxor                  m1, m1
   mova                  m0, [aboveq]
   mova                  m3, [aboveq+16]
   mova                  m2, [leftq]
   mova                  m4, [leftq+16]
   DEFINE_ARGS dst, stride, stride3, lines4
   lea             stride3q, [strideq*3]
   mov              lines4d, 4
@@ skipping 23 matching lines @@
   mova   [dstq+stride3q*2+16], m0
   lea                 dstq, [dstq+strideq*8]
   dec              lines4d
   jnz .loop
 
   RESTORE_GOT
   REP_RET
 
 %if ARCH_X86_64
 INIT_XMM sse2
-cglobal high_dc_predictor_32x32, 4, 5, 9, dst, stride, above, left, goffset
+cglobal highbd_dc_predictor_32x32, 4, 5, 9, dst, stride, above, left, goffset
   GET_GOT     goffsetq
 
   pxor                  m1, m1
   mova                  m0, [aboveq]
   mova                  m2, [aboveq+16]
   mova                  m3, [aboveq+32]
   mova                  m4, [aboveq+48]
   mova                  m5, [leftq]
   mova                  m6, [leftq+16]
   mova                  m7, [leftq+32]
@@ skipping 39 matching lines @@
   mova [dstq+stride3q*2 +48], m0
   lea                 dstq, [dstq+strideq*8]
   dec              lines4d
   jnz .loop
 
   RESTORE_GOT
   REP_RET
 %endif
 
 INIT_MMX sse
-cglobal high_v_predictor_4x4, 3, 3, 1, dst, stride, above
+cglobal highbd_v_predictor_4x4, 3, 3, 1, dst, stride, above
   movq                  m0, [aboveq]
   movq    [dstq          ], m0
   movq    [dstq+strideq*2], m0
   lea                 dstq, [dstq+strideq*4]
   movq    [dstq          ], m0
   movq    [dstq+strideq*2], m0
   RET
 
 INIT_XMM sse2
-cglobal high_v_predictor_8x8, 3, 3, 1, dst, stride, above
+cglobal highbd_v_predictor_8x8, 3, 3, 1, dst, stride, above
   mova                  m0, [aboveq]
   DEFINE_ARGS dst, stride, stride3
   lea             stride3q, [strideq*3]
   mova   [dstq           ], m0
   mova   [dstq+strideq*2 ], m0
   mova   [dstq+strideq*4 ], m0
   mova   [dstq+stride3q*2], m0
   lea                 dstq, [dstq+strideq*8]
   mova   [dstq           ], m0
   mova   [dstq+strideq*2 ], m0
   mova   [dstq+strideq*4 ], m0
   mova   [dstq+stride3q*2], m0
   RET
 
 INIT_XMM sse2
-cglobal high_v_predictor_16x16, 3, 4, 2, dst, stride, above
+cglobal highbd_v_predictor_16x16, 3, 4, 2, dst, stride, above
   mova                  m0, [aboveq]
   mova                  m1, [aboveq+16]
   DEFINE_ARGS dst, stride, stride3, nlines4
   lea             stride3q, [strideq*3]
   mov              nlines4d, 4
 .loop:
   mova    [dstq              ], m0
   mova    [dstq           +16], m1
   mova    [dstq+strideq*2    ], m0
   mova    [dstq+strideq*2 +16], m1
   mova    [dstq+strideq*4    ], m0
   mova    [dstq+strideq*4 +16], m1
   mova    [dstq+stride3q*2   ], m0
   mova    [dstq+stride3q*2+16], m1
   lea                 dstq, [dstq+strideq*8]
   dec             nlines4d
   jnz .loop
   REP_RET
 
 INIT_XMM sse2
-cglobal high_v_predictor_32x32, 3, 4, 4, dst, stride, above
+cglobal highbd_v_predictor_32x32, 3, 4, 4, dst, stride, above
   mova                  m0, [aboveq]
   mova                  m1, [aboveq+16]
   mova                  m2, [aboveq+32]
   mova                  m3, [aboveq+48]
   DEFINE_ARGS dst, stride, stride3, nlines4
   lea             stride3q, [strideq*3]
   mov              nlines4d, 8
 .loop:
   mova [dstq               ], m0
   mova [dstq            +16], m1
@@ skipping 10 matching lines @@
   mova [dstq+stride3q*2    ], m0
   mova [dstq+stride3q*2 +16], m1
   mova [dstq+stride3q*2 +32], m2
   mova [dstq+stride3q*2 +48], m3
   lea                 dstq, [dstq+strideq*8]
   dec             nlines4d
   jnz .loop
   REP_RET
 
 INIT_MMX sse
-cglobal high_tm_predictor_4x4, 5, 6, 5, dst, stride, above, left, bps, one
+cglobal highbd_tm_predictor_4x4, 5, 6, 5, dst, stride, above, left, bps, one
   movd                  m1, [aboveq-2]
   movq                  m0, [aboveq]
   pshufw                m1, m1, 0x0
   ; Get the values to compute the maximum value at this bit depth
   mov                 oned, 1
   movd                  m3, oned
   movd                  m4, bpsd
   pshufw                m3, m3, 0x0
   DEFINE_ARGS dst, stride, line, left
   mov                lineq, -2
@@ skipping 17 matching lines @@
   pmaxsw                m2, m4
   ;Store the values
   movq    [dstq          ], m1
   movq    [dstq+strideq*2], m2
   lea                 dstq, [dstq+strideq*4]
   inc                lineq
   jnz .loop
   REP_RET
 
 INIT_XMM sse2
-cglobal high_tm_predictor_8x8, 5, 6, 5, dst, stride, above, left, bps, one
+cglobal highbd_tm_predictor_8x8, 5, 6, 5, dst, stride, above, left, bps, one
   movd                  m1, [aboveq-2]
   mova                  m0, [aboveq]
   pshuflw               m1, m1, 0x0
   ; Get the values to compute the maximum value at this bit depth
   mov                 oned, 1
   pxor                  m3, m3
   pxor                  m4, m4
   pinsrw                m3, oned, 0
   pinsrw                m4, bpsd, 0
   pshuflw               m3, m3, 0x0
@@ skipping 24 matching lines @@
   ;Store the values
   mova      [dstq          ], m1
   mova      [dstq+strideq*2], m2
   lea                 dstq, [dstq+strideq*4]
   inc                lineq
   jnz .loop
   REP_RET
 
 %if ARCH_X86_64
 INIT_XMM sse2
-cglobal high_tm_predictor_16x16, 5, 6, 8, dst, stride, above, left, bps, one
+cglobal highbd_tm_predictor_16x16, 5, 6, 8, dst, stride, above, left, bps, one
   movd                  m2, [aboveq-2]
   mova                  m0, [aboveq]
   mova                  m1, [aboveq+16]
   pshuflw               m2, m2, 0x0
   ; Get the values to compute the maximum value at this bit depth
   mov                 oned, 1
   pxor                  m7, m7
   pxor                  m8, m8
   pinsrw                m7, oned, 0
   pinsrw                m8, bpsd, 0
@@ skipping 33 matching lines @@
   mova   [dstq             ], m4
   mova   [dstq+strideq*2   ], m5
   mova   [dstq          +16], m2
   mova   [dstq+strideq*2+16], m3
   lea                 dstq, [dstq+strideq*4]
   inc                lineq
   jnz .loop
   REP_RET
 
 INIT_XMM sse2
-cglobal high_tm_predictor_32x32, 5, 6, 12, dst, stride, above, left, bps, one
+cglobal highbd_tm_predictor_32x32, 5, 6, 12, dst, stride, above, left, bps, one
   movd                  m0, [aboveq-2]
   mova                  m1, [aboveq]
   mova                  m2, [aboveq+16]
   mova                  m3, [aboveq+32]
   mova                  m4, [aboveq+48]
   pshuflw               m0, m0, 0x0
   ; Get the values to compute the maximum value at this bit depth
   mov                 oned, 1
   pxor                 m10, m10
   pxor                 m11, m11
@@ skipping 54 matching lines @@
   ;Store these values
   mova   [dstq+strideq*2   ], m7
   mova   [dstq+strideq*2+16], m8
   mova   [dstq+strideq*2+32], m9
   mova   [dstq+strideq*2+48], m6
   lea                 dstq, [dstq+strideq*4]
   inc                lineq
   jnz .loop
   REP_RET
 %endif