1 /*
2 * Copyright 2021 Advanced Micro Devices, Inc.
3 *
4 * Permission is hereby granted, free of charge, to any person obtaining a
5 * copy of this software and associated documentation files (the "Software"),
6 * to deal in the Software without restriction, including without limitation
7 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
8 * and/or sell copies of the Software, and to permit persons to whom the
9 * Software is furnished to do so, subject to the following conditions:
10 *
11 * The above copyright notice and this permission notice shall be included in
12 * all copies or substantial portions of the Software.
13 *
14 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
15 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
16 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
17 * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
18 * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
19 * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
20 * OTHER DEALINGS IN THE SOFTWARE.
21 *
22 * Authors: AMD
23 *
24 */
25
26 #include "dm_services.h"
27 #include "core_types.h"
28 #include "reg_helper.h"
29 #include "dcn32/dcn32_dpp.h"
30 #include "basics/conversion.h"
31 #include "dcn30/dcn30_cm_common.h"
32
33 /* Compute the maximum number of lines that we can fit in the line buffer */
dscl32_calc_lb_num_partitions(const struct scaler_data * scl_data,enum lb_memory_config lb_config,int * num_part_y,int * num_part_c)34 static void dscl32_calc_lb_num_partitions(
35 const struct scaler_data *scl_data,
36 enum lb_memory_config lb_config,
37 int *num_part_y,
38 int *num_part_c)
39 {
40 int memory_line_size_y, memory_line_size_c, memory_line_size_a,
41 lb_memory_size, lb_memory_size_c, lb_memory_size_a, num_partitions_a;
42
43 int line_size = scl_data->viewport.width < scl_data->recout.width ?
44 scl_data->viewport.width : scl_data->recout.width;
45 int line_size_c = scl_data->viewport_c.width < scl_data->recout.width ?
46 scl_data->viewport_c.width : scl_data->recout.width;
47
48 if (line_size == 0)
49 line_size = 1;
50
51 if (line_size_c == 0)
52 line_size_c = 1;
53
54 memory_line_size_y = (line_size + 5) / 6; /* +5 to ceil */
55 memory_line_size_c = (line_size_c + 5) / 6; /* +5 to ceil */
56 memory_line_size_a = (line_size + 5) / 6; /* +5 to ceil */
57
58 if (lb_config == LB_MEMORY_CONFIG_1) {
59 lb_memory_size = 970;
60 lb_memory_size_c = 970;
61 lb_memory_size_a = 970;
62 } else if (lb_config == LB_MEMORY_CONFIG_2) {
63 lb_memory_size = 1290;
64 lb_memory_size_c = 1290;
65 lb_memory_size_a = 1290;
66 } else if (lb_config == LB_MEMORY_CONFIG_3) {
67 if (scl_data->viewport.width == scl_data->h_active &&
68 scl_data->viewport.height == scl_data->v_active) {
69 /* 420 mode: luma using all 3 mem from Y, plus 3rd mem from Cr and Cb */
70 /* use increased LB size for calculation only if Scaler not enabled */
71 lb_memory_size = 970 + 1290 + 1170 + 1170 + 1170;
72 lb_memory_size_c = 970 + 1290;
73 lb_memory_size_a = 970 + 1290 + 1170;
74 } else {
75 /* 420 mode: luma using all 3 mem from Y, plus 3rd mem from Cr and Cb */
76 lb_memory_size = 970 + 1290 + 484 + 484 + 484;
77 lb_memory_size_c = 970 + 1290;
78 lb_memory_size_a = 970 + 1290 + 484;
79 }
80 } else {
81 if (scl_data->viewport.width == scl_data->h_active &&
82 scl_data->viewport.height == scl_data->v_active) {
83 /* use increased LB size for calculation only if Scaler not enabled */
84 lb_memory_size = 970 + 1290 + 1170;
85 lb_memory_size_c = 970 + 1290 + 1170;
86 lb_memory_size_a = 970 + 1290 + 1170;
87 } else {
88 lb_memory_size = 970 + 1290 + 484;
89 lb_memory_size_c = 970 + 1290 + 484;
90 lb_memory_size_a = 970 + 1290 + 484;
91 }
92 }
93 *num_part_y = lb_memory_size / memory_line_size_y;
94 *num_part_c = lb_memory_size_c / memory_line_size_c;
95 num_partitions_a = lb_memory_size_a / memory_line_size_a;
96
97 if (scl_data->lb_params.alpha_en
98 && (num_partitions_a < *num_part_y))
99 *num_part_y = num_partitions_a;
100
101 if (*num_part_y > 32)
102 *num_part_y = 32;
103 if (*num_part_c > 32)
104 *num_part_c = 32;
105 }
106
107 static struct dpp_funcs dcn32_dpp_funcs = {
108 .dpp_program_gamcor_lut = dpp3_program_gamcor_lut,
109 .dpp_read_state = dpp30_read_state,
110 .dpp_reset = dpp_reset,
111 .dpp_set_scaler = dpp1_dscl_set_scaler_manual_scale,
112 .dpp_get_optimal_number_of_taps = dpp3_get_optimal_number_of_taps,
113 .dpp_set_gamut_remap = dpp3_cm_set_gamut_remap,
114 .dpp_set_csc_adjustment = NULL,
115 .dpp_set_csc_default = NULL,
116 .dpp_program_regamma_pwl = NULL,
117 .dpp_set_pre_degam = dpp3_set_pre_degam,
118 .dpp_program_input_lut = NULL,
119 .dpp_full_bypass = dpp1_full_bypass,
120 .dpp_setup = dpp3_cnv_setup,
121 .dpp_program_degamma_pwl = NULL,
122 .dpp_program_cm_dealpha = dpp3_program_cm_dealpha,
123 .dpp_program_cm_bias = dpp3_program_cm_bias,
124
125 .dpp_program_blnd_lut = NULL, // BLNDGAM is removed completely in DCN3.2 DPP
126 .dpp_program_shaper_lut = NULL, // CM SHAPER block is removed in DCN3.2 DPP, (it is in MPCC, programmable before or after BLND)
127 .dpp_program_3dlut = NULL, // CM 3DLUT block is removed in DCN3.2 DPP, (it is in MPCC, programmable before or after BLND)
128
129 .dpp_program_bias_and_scale = NULL,
130 .dpp_cnv_set_alpha_keyer = dpp2_cnv_set_alpha_keyer,
131 .set_cursor_attributes = dpp3_set_cursor_attributes,
132 .set_cursor_position = dpp1_set_cursor_position,
133 .set_optional_cursor_attributes = dpp1_cnv_set_optional_cursor_attributes,
134 .dpp_dppclk_control = dpp1_dppclk_control,
135 .dpp_set_hdr_multiplier = dpp3_set_hdr_multiplier,
136 .dpp_get_gamut_remap = dpp3_cm_get_gamut_remap,
137 .dpp_read_reg_state = dpp30_read_reg_state,
138 };
139
140
141 static struct dpp_caps dcn32_dpp_cap = {
142 .dscl_data_proc_format = DSCL_DATA_PRCESSING_FLOAT_FORMAT,
143 .max_lb_partitions = 31,
144 .dscl_calc_lb_num_partitions = dscl32_calc_lb_num_partitions,
145 };
146
dpp32_construct(struct dcn3_dpp * dpp,struct dc_context * ctx,uint32_t inst,const struct dcn3_dpp_registers * tf_regs,const struct dcn3_dpp_shift * tf_shift,const struct dcn3_dpp_mask * tf_mask)147 bool dpp32_construct(
148 struct dcn3_dpp *dpp,
149 struct dc_context *ctx,
150 uint32_t inst,
151 const struct dcn3_dpp_registers *tf_regs,
152 const struct dcn3_dpp_shift *tf_shift,
153 const struct dcn3_dpp_mask *tf_mask)
154 {
155 dpp->base.ctx = ctx;
156
157 dpp->base.inst = inst;
158 dpp->base.funcs = &dcn32_dpp_funcs;
159 dpp->base.caps = &dcn32_dpp_cap;
160
161 dpp->tf_regs = tf_regs;
162 dpp->tf_shift = tf_shift;
163 dpp->tf_mask = tf_mask;
164
165 return true;
166 }
dscl32_spl_calc_lb_num_partitions(bool alpha_en,const struct spl_scaler_data * scl_data,enum lb_memory_config lb_config,int * num_part_y,int * num_part_c)167 void dscl32_spl_calc_lb_num_partitions(
168 bool alpha_en,
169 const struct spl_scaler_data *scl_data,
170 enum lb_memory_config lb_config,
171 int *num_part_y,
172 int *num_part_c)
173 {
174 int memory_line_size_y, memory_line_size_c, memory_line_size_a,
175 lb_memory_size, lb_memory_size_c, lb_memory_size_a, num_partitions_a;
176
177 int line_size = scl_data->viewport.width < scl_data->recout.width ?
178 scl_data->viewport.width : scl_data->recout.width;
179 int line_size_c = scl_data->viewport_c.width < scl_data->recout.width ?
180 scl_data->viewport_c.width : scl_data->recout.width;
181
182 if (line_size == 0)
183 line_size = 1;
184
185 if (line_size_c == 0)
186 line_size_c = 1;
187
188 memory_line_size_y = (line_size + 5) / 6; /* +5 to ceil */
189 memory_line_size_c = (line_size_c + 5) / 6; /* +5 to ceil */
190 memory_line_size_a = (line_size + 5) / 6; /* +5 to ceil */
191
192 if (lb_config == LB_MEMORY_CONFIG_1) {
193 lb_memory_size = 970;
194 lb_memory_size_c = 970;
195 lb_memory_size_a = 970;
196 } else if (lb_config == LB_MEMORY_CONFIG_2) {
197 lb_memory_size = 1290;
198 lb_memory_size_c = 1290;
199 lb_memory_size_a = 1290;
200 } else if (lb_config == LB_MEMORY_CONFIG_3) {
201 if (scl_data->viewport.width == scl_data->h_active &&
202 scl_data->viewport.height == scl_data->v_active) {
203 /* 420 mode: luma using all 3 mem from Y, plus 3rd mem from Cr and Cb */
204 /* use increased LB size for calculation only if Scaler not enabled */
205 lb_memory_size = 970 + 1290 + 1170 + 1170 + 1170;
206 lb_memory_size_c = 970 + 1290;
207 lb_memory_size_a = 970 + 1290 + 1170;
208 } else {
209 /* 420 mode: luma using all 3 mem from Y, plus 3rd mem from Cr and Cb */
210 lb_memory_size = 970 + 1290 + 484 + 484 + 484;
211 lb_memory_size_c = 970 + 1290;
212 lb_memory_size_a = 970 + 1290 + 484;
213 }
214 } else {
215 if (scl_data->viewport.width == scl_data->h_active &&
216 scl_data->viewport.height == scl_data->v_active) {
217 /* use increased LB size for calculation only if Scaler not enabled */
218 lb_memory_size = 970 + 1290 + 1170;
219 lb_memory_size_c = 970 + 1290 + 1170;
220 lb_memory_size_a = 970 + 1290 + 1170;
221 } else {
222 lb_memory_size = 970 + 1290 + 484;
223 lb_memory_size_c = 970 + 1290 + 484;
224 lb_memory_size_a = 970 + 1290 + 484;
225 }
226 }
227 *num_part_y = lb_memory_size / memory_line_size_y;
228 *num_part_c = lb_memory_size_c / memory_line_size_c;
229 num_partitions_a = lb_memory_size_a / memory_line_size_a;
230
231 if (alpha_en
232 && (num_partitions_a < *num_part_y))
233 *num_part_y = num_partitions_a;
234
235 if (*num_part_y > 32)
236 *num_part_y = 32;
237 if (*num_part_c > 32)
238 *num_part_c = 32;
239 }
240