1 /**************************************************************************
3 * Copyright 2007 Tungsten Graphics, Inc., Cedar Park, Texas.
6 * Permission is hereby granted, free of charge, to any person obtaining a
7 * copy of this software and associated documentation files (the
8 * "Software"), to deal in the Software without restriction, including
9 * without limitation the rights to use, copy, modify, merge, publish,
10 * distribute, sub license, and/or sell copies of the Software, and to
11 * permit persons to whom the Software is furnished to do so, subject to
12 * the following conditions:
14 * The above copyright notice and this permission notice (including the
15 * next paragraph) shall be included in all copies or substantial portions
18 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
19 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
20 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
21 * IN NO EVENT SHALL TUNGSTEN GRAPHICS AND/OR ITS SUPPLIERS BE LIABLE FOR
22 * ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
23 * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
24 * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
26 **************************************************************************/
29 * Execute fragment shader using runtime SSE code generation.
32 #include "sp_context.h"
37 #include "pipe/p_state.h"
38 #include "pipe/p_defines.h"
39 #include "util/u_memory.h"
40 #include "tgsi/tgsi_exec.h"
41 #include "tgsi/tgsi_sse2.h"
44 #if defined(PIPE_ARCH_X86)
46 #include "rtasm/rtasm_x86sse.h"
51 * Subclass of sp_fragment_shader
53 struct sp_sse_fragment_shader
55 struct sp_fragment_shader base;
56 struct x86_function sse2_program;
57 tgsi_sse2_fs_function func;
58 float immediates[TGSI_EXEC_NUM_IMMEDIATES][4];
63 static INLINE struct sp_sse_fragment_shader *
64 sp_sse_fragment_shader(const struct sp_fragment_shader *base)
66 return (struct sp_sse_fragment_shader *) base;
71 fs_sse_prepare( const struct sp_fragment_shader *base,
72 struct tgsi_exec_machine *machine,
73 struct tgsi_sampler **samplers )
78 /* TODO: codegenerate the whole run function, skip this wrapper.
79 * TODO: break dependency on tgsi_exec_machine struct
80 * TODO: push Position calculation into the generated shader
81 * TODO: process >1 quad at a time
84 fs_sse_run( const struct sp_fragment_shader *base,
85 struct tgsi_exec_machine *machine,
86 struct quad_header *quad )
88 struct sp_sse_fragment_shader *shader = sp_sse_fragment_shader(base);
90 /* Compute X, Y, Z, W vals for this quad -- place in temp[0] for now */
91 sp_setup_pos_vector(quad->posCoef,
92 (float)quad->input.x0, (float)quad->input.y0,
96 tgsi_set_kill_mask(machine, 0x0);
97 tgsi_set_exec_mask(machine, 1, 1, 1, 1);
99 shader->func( machine,
101 (const float (*)[4])shader->immediates,
103 // , &machine->QuadPos
106 return ~(machine->Temps[TGSI_EXEC_TEMP_KILMASK_I].xyzw[TGSI_EXEC_TEMP_KILMASK_C].u[0]);
111 fs_sse_delete( struct sp_fragment_shader *base )
113 struct sp_sse_fragment_shader *shader = sp_sse_fragment_shader(base);
115 x86_release_func( &shader->sse2_program );
120 struct sp_fragment_shader *
121 softpipe_create_fs_sse(struct softpipe_context *softpipe,
122 const struct pipe_shader_state *templ)
124 struct sp_sse_fragment_shader *shader;
126 if (!softpipe->use_sse)
129 shader = CALLOC_STRUCT(sp_sse_fragment_shader);
133 x86_init_func( &shader->sse2_program );
135 if (!tgsi_emit_sse2( templ->tokens, &shader->sse2_program,
136 shader->immediates, FALSE )) {
141 shader->func = (tgsi_sse2_fs_function) x86_get_func( &shader->sse2_program );
143 x86_release_func( &shader->sse2_program );
148 shader->base.shader.tokens = NULL; /* don't hold reference to templ->tokens */
149 shader->base.prepare = fs_sse_prepare;
150 shader->base.run = fs_sse_run;
151 shader->base.delete = fs_sse_delete;
153 return &shader->base;
159 /* Maybe put this variant in the header file.
161 struct sp_fragment_shader *
162 softpipe_create_fs_sse(struct softpipe_context *softpipe,
163 const struct pipe_shader_state *templ)