2
* All Video Processing kernels
3
* Copyright © <2010>, Intel Corporation.
5
* Permission is hereby granted, free of charge, to any person obtaining a
6
* copy of this software and associated documentation files (the
7
* "Software"), to deal in the Software without restriction, including
8
* without limitation the rights to use, copy, modify, merge, publish,
9
* distribute, sub license, and/or sell copies of the Software, and to
10
* permit persons to whom the Software is furnished to do so, subject to
11
* the following conditions:
13
* The above copyright notice and this permission notice (including the
14
* next paragraph) shall be included in all copies or substantial portions
17
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
18
* OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
19
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
20
* IN NO EVENT SHALL PRECISION INSIGHT AND/OR ITS SUPPLIERS BE LIABLE FOR
21
* ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
22
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
23
* SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
25
* This file was originally licensed under the following license
27
* Licensed under the Apache License, Version 2.0 (the "License");
28
* you may not use this file except in compliance with the License.
29
* You may obtain a copy of the License at
31
* http://www.apache.org/licenses/LICENSE-2.0
33
* Unless required by applicable law or agreed to in writing, software
34
* distributed under the License is distributed on an "AS IS" BASIS,
35
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
36
* See the License for the specific language governing permissions and
37
* limitations under the License.
46
#define nY_NUM_OF_ROWS 8 // Number of Y rows per block (4 rows for each frame)
47
#undef nUV_NUM_OF_ROWS
48
#define nUV_NUM_OF_ROWS 8 // Number of U/V rows per block
51
#define nSMPL_RESP_LEN nSMPL_RESP_LEN_DNDI // set the number of GRF
52
#undef nDPW_BLOCK_SIZE_HIST
53
#define nDPW_BLOCK_SIZE_HIST nBLOCK_WIDTH_4+nBLOCK_HEIGHT_1 // HIST Block Size for Write is 4x2
54
#undef nDPW_BLOCK_SIZE_DN
55
#define nDPW_BLOCK_SIZE_DN nBLOCK_WIDTH_16+nBLOCK_HEIGHT_4 // DN Block Size for Write is 16x4
56
#undef nDPR_BLOCK_SIZE_UV
57
#define nDPR_BLOCK_SIZE_UV nBLOCK_WIDTH_16+nBLOCK_HEIGHT_2 // DN Block Size for UV Write/Read is 16x2
59
////////////////////////////////////// Run the DN Algorithm ///////////////////////////////////////
60
#include "DNDI_COMMAND.asm"
62
////////////////////////////////////// Rearrange for Internal Planar //////////////////////////////
63
// move the previous frame Y component to internal planar format
64
$for (0; <nY_NUM_OF_ROWS/2; 1) {
65
mov (16) uwDEST_Y(%1,0)<1> ubRESP(nDI_PREV_FRAME_LUMA_OFFSET,%1*16)
67
// move the previous frame U,V components to internal planar format
68
$for (0; <nUV_NUM_OF_ROWS/2; 1) {
69
mov (8) uwDEST_U(0,%1*8)<1> ubRESP(nDI_PREV_FRAME_CHROMA_OFFSET,%1*16+1)<16;8,2> //U pixels
70
mov (8) uwDEST_V(0,%1*8)<1> ubRESP(nDI_PREV_FRAME_CHROMA_OFFSET,%1*16)<16;8,2> //V pixels
72
// move the current frame Y component to internal planar format
73
$for (0; <nY_NUM_OF_ROWS/2; 1) {
74
mov (16) uwDEST_Y(%1+4,0)<1> ubRESP(nDI_CURR_FRAME_LUMA_OFFSET,%1*16)
76
// move the current frame U,V components to internal planar format
77
$for (0; <nUV_NUM_OF_ROWS/2; 1) {
78
mov (8) uwDEST_U(2,%1*8)<1> ubRESP(nDI_CURR_FRAME_CHROMA_OFFSET,%1*16+1)<16;8,2> //U pixels
79
mov (8) uwDEST_V(2,%1*8)<1> ubRESP(nDI_CURR_FRAME_CHROMA_OFFSET,%1*16)<16;8,2> //V pixels
82
////////////////////////////////////// Save the STMM Data for Next Run /////////////////////////
83
// Write STMM to memory
84
shr (1) rMSGSRC.0<1>:ud wORIX<0;1,0>:w 1:w // X origin / 2
85
mov (1) rMSGSRC.1<1>:ud wORIY<0;1,0>:w // Y origin
86
mov (1) rMSGSRC.2<1>:ud nDPW_BLOCK_SIZE_STMM:ud // block width and height (8x4)
87
mov (8) mudMSGHDR_STMM(0)<1> rMSGSRC.0<8;8,1>:ud // message header
88
mov (8) mudMSGHDR_STMM(1)<1> udRESP(nDI_STMM_OFFSET,0) // Move STMM to MRF
89
send (8) dNULLREG mMSGHDR_STMM udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPMW_MSG_LEN_STMM+nBI_STMM_HISTORY_OUTPUT:ud
91
////////////////////////////////////// Save the History Data for Next Run /////////////////////////
92
#include "DI_Hist_Save.asm"
94
////////////////////////////////////// Save the DN Curr Frame for Next Run ////////////////////////
96
$for (0; <nY_NUM_OF_ROWS/2; 1) {
97
mov (16) mubMSGHDR_DN(1, %1*16)<1> ubRESP(nDI_PREV_FRAME_LUMA_OFFSET,%1*16)
100
mov (2) rMSGSRC.0<1>:ud wORIX<2;2,1>:w // X origin and Y origin
101
mov (1) rMSGSRC.2<1>:ud nDPW_BLOCK_SIZE_DN:ud // block width and height (16x4)
102
mov (8) mudMSGHDR_DN(0)<1> rMSGSRC.0<8;8,1>:ud
103
send (8) dNULLREG mMSGHDR_DN udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPMW_MSG_LEN_PL_DN_DI+nBI_DESTINATION_Y:ud
105
//Write UV through DATAPORT
106
mov (2) rMSGSRC.0<1>:ud wORIX<2;2,1>:w // X origin and Y origin
107
asr (1) rMSGSRC.1<1>:d rMSGSRC.1<0;1,0>:d 1:w // U/V block origin should be half of Y's
108
mov (1) rMSGSRC.2<1>:ud nDPR_BLOCK_SIZE_UV:ud // block width and height (16x2)
109
mov (8) mudMSGHDR_DN(0)<1> rMSGSRC.0<8;8,1>:ud
111
mov (8) mubMSGHDR_DN(1, 0)<2> ubRESP(nDI_PREV_FRAME_CHROMA_OFFSET, 1)<16 ;8,2>
112
mov (8) mubMSGHDR_DN(1, 1)<2> ubRESP(nDI_PREV_FRAME_CHROMA_OFFSET, 16)<16 ;8,2>
113
mov (8) mubMSGHDR_DN(1, 16)<2> ubRESP(nDI_PREV_FRAME_CHROMA_OFFSET+1, 1)<16 ;8,2>
114
mov (8) mubMSGHDR_DN(1, 17)<2> ubRESP(nDI_PREV_FRAME_CHROMA_OFFSET+1, 16)<16 ;8,2>
115
send (8) dNULLREG mMSGHDR_DN udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nMSGLEN_1+nBI_DESTINATION_UV:ud
118
$for (0; <nY_NUM_OF_ROWS/2; 1) {
119
mov (16) mubMSGHDR_DN(1, %1*16)<1> ubRESP(nDI_CURR_FRAME_LUMA_OFFSET,%1*16)
122
mov (2) rMSGSRC.0<1>:ud wORIX<2;2,1>:w // X origin and Y origin
123
mov (1) rMSGSRC.2<1>:ud nDPW_BLOCK_SIZE_DN:ud // block width and height (16x4)
124
mov (8) mudMSGHDR_DN(0)<1> rMSGSRC.0<8;8,1>:ud
125
send (8) dNULLREG mMSGHDR_DN udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPMW_MSG_LEN_PL_DN_DI+nBI_DESTINATION_1_Y:ud
127
//Write UV through DATAPORT
128
mov (2) rMSGSRC.0<1>:ud wORIX<2;2,1>:w // X origin and Y origin
129
asr (1) rMSGSRC.1<1>:d rMSGSRC.1<0;1,0>:d 1:w // U/V block origin should be half of Y's
130
mov (1) rMSGSRC.2<1>:ud nDPR_BLOCK_SIZE_UV:ud // block width and height (16x2)
131
mov (8) mudMSGHDR_DN(0)<1> rMSGSRC.0<8;8,1>:ud
133
mov (8) mubMSGHDR_DN(1, 0)<2> ubRESP(nDI_CURR_FRAME_CHROMA_OFFSET, 1)<16 ;8,2>
134
mov (8) mubMSGHDR_DN(1, 1)<2> ubRESP(nDI_CURR_FRAME_CHROMA_OFFSET, 16)<16 ;8,2>
135
mov (8) mubMSGHDR_DN(1, 16)<2> ubRESP(nDI_CURR_FRAME_CHROMA_OFFSET+1, 1)<16 ;8,2>
136
mov (8) mubMSGHDR_DN(1, 17)<2> ubRESP(nDI_CURR_FRAME_CHROMA_OFFSET+1, 16)<16 ;8,2>
137
send (8) dNULLREG mMSGHDR_DN udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nMSGLEN_1+nBI_DESTINATION_1_UV:ud