freedreno/ir3: Lower output precision
[mesa.git] / src / freedreno / ir3 / ir3_compiler.h
1 /*
2 * Copyright (C) 2013 Rob Clark <robclark@freedesktop.org>
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 (including the next
12 * paragraph) shall be included in all copies or substantial portions of the
13 * Software.
14 *
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
18 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 * SOFTWARE.
22 *
23 * Authors:
24 * Rob Clark <robclark@freedesktop.org>
25 */
26
27 #ifndef IR3_COMPILER_H_
28 #define IR3_COMPILER_H_
29
30 #include "ir3_shader.h"
31
32 struct ir3_ra_reg_set;
33
34 struct ir3_compiler {
35 struct fd_device *dev;
36 uint32_t gpu_id;
37 struct ir3_ra_reg_set *set;
38 uint32_t shader_count;
39
40 /*
41 * Configuration options for things that are handled differently on
42 * different generations:
43 */
44
45 /* a4xx (and later) drops SP_FS_FLAT_SHAD_MODE_REG_* for flat-interpolate
46 * so we need to use ldlv.u32 to load the varying directly:
47 */
48 bool flat_bypass;
49
50 /* on a3xx, we need to add one to # of array levels:
51 */
52 bool levels_add_one;
53
54 /* on a3xx, we need to scale up integer coords for isaml based
55 * on LoD:
56 */
57 bool unminify_coords;
58
59 /* on a3xx do txf_ms w/ isaml and scaled coords: */
60 bool txf_ms_with_isaml;
61
62 /* on a4xx, for array textures we need to add 0.5 to the array
63 * index coordinate:
64 */
65 bool array_index_add_half;
66
67 /* on a6xx, rewrite samgp to sequence of samgq0-3 in vertex shaders:
68 */
69 bool samgq_workaround;
70 };
71
72 struct ir3_compiler * ir3_compiler_create(struct fd_device *dev, uint32_t gpu_id);
73
74 int ir3_compile_shader_nir(struct ir3_compiler *compiler,
75 struct ir3_shader_variant *so);
76
77 /* gpu pointer size in units of 32bit registers/slots */
78 static inline
79 unsigned ir3_pointer_size(struct ir3_compiler *compiler)
80 {
81 return (compiler->gpu_id >= 500) ? 2 : 1;
82 }
83
84 enum ir3_shader_debug {
85 IR3_DBG_SHADER_VS = 0x001,
86 IR3_DBG_SHADER_TCS = 0x002,
87 IR3_DBG_SHADER_TES = 0x004,
88 IR3_DBG_SHADER_GS = 0x008,
89 IR3_DBG_SHADER_FS = 0x010,
90 IR3_DBG_SHADER_CS = 0x020,
91 IR3_DBG_DISASM = 0x040,
92 IR3_DBG_OPTMSGS = 0x080,
93 IR3_DBG_FORCES2EN = 0x100,
94 IR3_DBG_NOUBOOPT = 0x200,
95 IR3_DBG_SCHEDMSGS = 0x400,
96 IR3_DBG_NOFP16 = 0x800,
97 };
98
99 extern enum ir3_shader_debug ir3_shader_debug;
100
101 static inline bool
102 shader_debug_enabled(gl_shader_stage type)
103 {
104 if (ir3_shader_debug & IR3_DBG_DISASM)
105 return true;
106
107 switch (type) {
108 case MESA_SHADER_VERTEX: return !!(ir3_shader_debug & IR3_DBG_SHADER_VS);
109 case MESA_SHADER_TESS_CTRL: return !!(ir3_shader_debug & IR3_DBG_SHADER_TCS);
110 case MESA_SHADER_TESS_EVAL: return !!(ir3_shader_debug & IR3_DBG_SHADER_TES);
111 case MESA_SHADER_GEOMETRY: return !!(ir3_shader_debug & IR3_DBG_SHADER_GS);
112 case MESA_SHADER_FRAGMENT: return !!(ir3_shader_debug & IR3_DBG_SHADER_FS);
113 case MESA_SHADER_COMPUTE: return !!(ir3_shader_debug & IR3_DBG_SHADER_CS);
114 default:
115 debug_assert(0);
116 return false;
117 }
118 }
119
120 static inline void
121 ir3_debug_print(struct ir3 *ir, const char *when)
122 {
123 if (ir3_shader_debug & IR3_DBG_OPTMSGS) {
124 printf("%s:\n", when);
125 ir3_print(ir);
126 }
127 }
128
129 #endif /* IR3_COMPILER_H_ */