diff options
Diffstat (limited to 'src/mesa/state_tracker')
| -rw-r--r-- | src/mesa/state_tracker/st_atifs_to_tgsi.c | 845 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_atifs_to_tgsi.h | 67 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_atom_constbuf.c | 15 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_atom_sampler.c | 7 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_atom_shader.c | 65 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_cb_drawpixels.c | 1 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_cb_program.c | 31 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_cb_texture.c | 7 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_cb_xformfb.c | 4 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_draw.c | 37 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_extensions.c | 39 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_gen_mipmap.c | 102 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 4 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_program.c | 34 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_program.h | 8 | ||||
| -rw-r--r-- | src/mesa/state_tracker/st_vdpau.c | 181 |
16 files changed, 1261 insertions, 186 deletions
diff --git a/src/mesa/state_tracker/st_atifs_to_tgsi.c b/src/mesa/state_tracker/st_atifs_to_tgsi.c new file mode 100644 index 00000000000..66f442aee5a --- /dev/null +++ b/src/mesa/state_tracker/st_atifs_to_tgsi.c @@ -0,0 +1,845 @@ +/* + * Copyright (C) 2016 Miklós Máté + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#include "main/mtypes.h" +#include "main/atifragshader.h" +#include "main/errors.h" +#include "program/prog_parameter.h" + +#include "tgsi/tgsi_ureg.h" +#include "tgsi/tgsi_scan.h" +#include "tgsi/tgsi_transform.h" + +#include "st_program.h" +#include "st_atifs_to_tgsi.h" + +/** + * Intermediate state used during shader translation. + */ +struct st_translate { + struct ureg_program *ureg; + struct ati_fragment_shader *atifs; + + struct ureg_dst temps[MAX_PROGRAM_TEMPS]; + struct ureg_src *constants; + struct ureg_dst outputs[PIPE_MAX_SHADER_OUTPUTS]; + struct ureg_src inputs[PIPE_MAX_SHADER_INPUTS]; + struct ureg_src samplers[PIPE_MAX_SAMPLERS]; + + const GLuint *inputMapping; + const GLuint *outputMapping; + + unsigned current_pass; + + bool regs_written[MAX_NUM_PASSES_ATI][MAX_NUM_FRAGMENT_REGISTERS_ATI]; + + boolean error; +}; + +struct instruction_desc { + unsigned TGSI_opcode; + const char *name; + unsigned char arg_count; +}; + +static const struct instruction_desc inst_desc[] = { + {TGSI_OPCODE_MOV, "MOV", 1}, + {TGSI_OPCODE_NOP, "UND", 0}, /* unused */ + {TGSI_OPCODE_ADD, "ADD", 2}, + {TGSI_OPCODE_MUL, "MUL", 2}, + {TGSI_OPCODE_SUB, "SUB", 2}, + {TGSI_OPCODE_DP3, "DOT3", 2}, + {TGSI_OPCODE_DP4, "DOT4", 2}, + {TGSI_OPCODE_MAD, "MAD", 3}, + {TGSI_OPCODE_LRP, "LERP", 3}, + {TGSI_OPCODE_NOP, "CND", 3}, + {TGSI_OPCODE_NOP, "CND0", 3}, + {TGSI_OPCODE_NOP, "DOT2_ADD", 3} +}; + +static struct ureg_dst +get_temp(struct st_translate *t, unsigned index) +{ + if (ureg_dst_is_undef(t->temps[index])) + t->temps[index] = ureg_DECL_temporary(t->ureg); + return t->temps[index]; +} + +static struct ureg_src +apply_swizzle(struct st_translate *t, + struct ureg_src src, GLuint swizzle) +{ + if (swizzle == GL_SWIZZLE_STR_ATI) { + return src; + } else if (swizzle == GL_SWIZZLE_STQ_ATI) { + return ureg_swizzle(src, + TGSI_SWIZZLE_X, + TGSI_SWIZZLE_Y, + TGSI_SWIZZLE_W, + TGSI_SWIZZLE_Z); + } else { + struct ureg_dst tmp[2]; + struct ureg_src imm[3]; + + tmp[0] = get_temp(t, MAX_NUM_FRAGMENT_REGISTERS_ATI); + tmp[1] = get_temp(t, MAX_NUM_FRAGMENT_REGISTERS_ATI + 1); + imm[0] = src; + imm[1] = ureg_imm4f(t->ureg, 1.0f, 1.0f, 0.0f, 0.0f); + imm[2] = ureg_imm4f(t->ureg, 0.0f, 0.0f, 1.0f, 1.0f); + ureg_insn(t->ureg, TGSI_OPCODE_MAD, &tmp[0], 1, imm, 3); + + if (swizzle == GL_SWIZZLE_STR_DR_ATI) { + imm[0] = ureg_scalar(src, TGSI_SWIZZLE_Z); + } else { + imm[0] = ureg_scalar(src, TGSI_SWIZZLE_W); + } + ureg_insn(t->ureg, TGSI_OPCODE_RCP, &tmp[1], 1, &imm[0], 1); + + imm[0] = ureg_src(tmp[0]); + imm[1] = ureg_src(tmp[1]); + ureg_insn(t->ureg, TGSI_OPCODE_MUL, &tmp[0], 1, imm, 2); + + return ureg_src(tmp[0]); + } +} + +static struct ureg_src +get_source(struct st_translate *t, GLuint src_type) +{ + if (src_type >= GL_REG_0_ATI && src_type <= GL_REG_5_ATI) { + if (t->regs_written[t->current_pass][src_type - GL_REG_0_ATI]) { + return ureg_src(get_temp(t, src_type - GL_REG_0_ATI)); + } else { + return ureg_imm1f(t->ureg, 0.0f); + } + } else if (src_type >= GL_CON_0_ATI && src_type <= GL_CON_7_ATI) { + return t->constants[src_type - GL_CON_0_ATI]; + } else if (src_type == GL_ZERO) { + return ureg_imm1f(t->ureg, 0.0f); + } else if (src_type == GL_ONE) { + return ureg_imm1f(t->ureg, 1.0f); + } else if (src_type == GL_PRIMARY_COLOR_ARB) { + return t->inputs[t->inputMapping[VARYING_SLOT_COL0]]; + } else if (src_type == GL_SECONDARY_INTERPOLATOR_ATI) { + return t->inputs[t->inputMapping[VARYING_SLOT_COL1]]; + } else { + /* frontend prevents this */ + unreachable("unknown source"); + } +} + +static struct ureg_src +prepare_argument(struct st_translate *t, const unsigned argId, + const struct atifragshader_src_register *srcReg) +{ + struct ureg_src src = get_source(t, srcReg->Index); + struct ureg_dst arg = get_temp(t, MAX_NUM_FRAGMENT_REGISTERS_ATI + argId); + + switch (srcReg->argRep) { + case GL_NONE: + break; + case GL_RED: + src = ureg_scalar(src, TGSI_SWIZZLE_X); + break; + case GL_GREEN: + src = ureg_scalar(src, TGSI_SWIZZLE_Y); + break; + case GL_BLUE: + src = ureg_scalar(src, TGSI_SWIZZLE_Z); + break; + case GL_ALPHA: + src = ureg_scalar(src, TGSI_SWIZZLE_W); + break; + } + ureg_insn(t->ureg, TGSI_OPCODE_MOV, &arg, 1, &src, 1); + + if (srcReg->argMod & GL_COMP_BIT_ATI) { + struct ureg_src modsrc[2]; + modsrc[0] = ureg_imm1f(t->ureg, 1.0f); + modsrc[1] = ureg_src(arg); + + ureg_insn(t->ureg, TGSI_OPCODE_SUB, &arg, 1, modsrc, 2); + } + if (srcReg->argMod & GL_BIAS_BIT_ATI) { + struct ureg_src modsrc[2]; + modsrc[0] = ureg_src(arg); + modsrc[1] = ureg_imm1f(t->ureg, 0.5f); + + ureg_insn(t->ureg, TGSI_OPCODE_SUB, &arg, 1, modsrc, 2); + } + if (srcReg->argMod & GL_2X_BIT_ATI) { + struct ureg_src modsrc[2]; + modsrc[0] = ureg_src(arg); + modsrc[1] = ureg_src(arg); + + ureg_insn(t->ureg, TGSI_OPCODE_ADD, &arg, 1, modsrc, 2); + } + if (srcReg->argMod & GL_NEGATE_BIT_ATI) { + struct ureg_src modsrc[2]; + modsrc[0] = ureg_src(arg); + modsrc[1] = ureg_imm1f(t->ureg, -1.0f); + + ureg_insn(t->ureg, TGSI_OPCODE_MUL, &arg, 1, modsrc, 2); + } + return ureg_src(arg); +} + +/* These instructions need special treatment */ +static void +emit_special_inst(struct st_translate *t, const struct instruction_desc *desc, + struct ureg_dst *dst, struct ureg_src *args, unsigned argcount) +{ + struct ureg_dst tmp[1]; + struct ureg_src src[3]; + + if (!strcmp(desc->name, "CND")) { + tmp[0] = get_temp(t, MAX_NUM_FRAGMENT_REGISTERS_ATI + 2); /* re-purpose a3 */ + src[0] = ureg_imm1f(t->ureg, 0.5f); + src[1] = args[2]; + ureg_insn(t->ureg, TGSI_OPCODE_SUB, tmp, 1, src, 2); + src[0] = ureg_src(tmp[0]); + src[1] = args[0]; + src[2] = args[1]; + ureg_insn(t->ureg, TGSI_OPCODE_CMP, dst, 1, src, 3); + } else if (!strcmp(desc->name, "CND0")) { + src[0] = args[2]; + src[1] = args[1]; + src[2] = args[0]; + ureg_insn(t->ureg, TGSI_OPCODE_CMP, dst, 1, src, 3); + } else if (!strcmp(desc->name, "DOT2_ADD")) { + /* note: DP2A is not implemented in most pipe drivers */ + tmp[0] = get_temp(t, MAX_NUM_FRAGMENT_REGISTERS_ATI); /* re-purpose a1 */ + src[0] = args[0]; + src[1] = args[1]; + ureg_insn(t->ureg, TGSI_OPCODE_DP2, tmp, 1, src, 2); + src[0] = ureg_src(tmp[0]); + src[1] = ureg_scalar(args[2], TGSI_SWIZZLE_Z); + ureg_insn(t->ureg, TGSI_OPCODE_ADD, dst, 1, src, 2); + } +} + +static void +emit_arith_inst(struct st_translate *t, + const struct instruction_desc *desc, + struct ureg_dst *dst, struct ureg_src *args, unsigned argcount) +{ + if (desc->TGSI_opcode == TGSI_OPCODE_NOP) { + return emit_special_inst(t, desc, dst, args, argcount); + } + + ureg_insn(t->ureg, desc->TGSI_opcode, dst, 1, args, argcount); +} + +static void +emit_dstmod(struct st_translate *t, + struct ureg_dst dst, GLuint dstMod) +{ + float imm; + struct ureg_src src[3]; + GLuint scale = dstMod & ~GL_SATURATE_BIT_ATI; + + if (dstMod == GL_NONE) { + return; + } + + switch (scale) { + case GL_2X_BIT_ATI: + imm = 2.0f; + break; + case GL_4X_BIT_ATI: + imm = 4.0f; + break; + case GL_8X_BIT_ATI: + imm = 8.0f; + break; + case GL_HALF_BIT_ATI: + imm = 0.5f; + break; + case GL_QUARTER_BIT_ATI: + imm = 0.25f; + break; + case GL_EIGHTH_BIT_ATI: + imm = 0.125f; + break; + default: + imm = 1.0f; + } + + src[0] = ureg_src(dst); + src[1] = ureg_imm1f(t->ureg, imm); + if (dstMod & GL_SATURATE_BIT_ATI) { + dst = ureg_saturate(dst); + } + ureg_insn(t->ureg, TGSI_OPCODE_MUL, &dst, 1, src, 2); +} + +/** + * Compile one setup instruction to TGSI instructions. + */ +static void +compile_setupinst(struct st_translate *t, + const unsigned r, + const struct atifs_setupinst *texinst) +{ + struct ureg_dst dst[1]; + struct ureg_src src[2]; + + if (!texinst->Opcode) + return; + + dst[0] = get_temp(t, r); + + GLuint pass_tex = texinst->src; + + if (pass_tex >= GL_TEXTURE0_ARB && pass_tex <= GL_TEXTURE7_ARB) { + unsigned attr = pass_tex - GL_TEXTURE0_ARB + VARYING_SLOT_TEX0; + + src[0] = t->inputs[t->inputMapping[attr]]; + } else if (pass_tex >= GL_REG_0_ATI && pass_tex <= GL_REG_5_ATI) { + unsigned reg = pass_tex - GL_REG_0_ATI; + + /* the frontend already validated that REG is only allowed in second pass */ + if (t->regs_written[0][reg]) { + src[0] = ureg_src(t->temps[reg]); + } else { + src[0] = ureg_imm1f(t->ureg, 0.0f); + } + } + src[0] = apply_swizzle(t, src[0], texinst->swizzle); + + if (texinst->Opcode == ATI_FRAGMENT_SHADER_SAMPLE_OP) { + /* by default texture and sampler indexes are the same */ + src[1] = t->samplers[r]; + /* the texture target is still unknown, it will be fixed in the draw call */ + ureg_tex_insn(t->ureg, TGSI_OPCODE_TEX, dst, 1, TGSI_TEXTURE_2D, + NULL, 0, src, 2); + } else if (texinst->Opcode == ATI_FRAGMENT_SHADER_PASS_OP) { + ureg_insn(t->ureg, TGSI_OPCODE_MOV, dst, 1, src, 1); + } + + t->regs_written[t->current_pass][r] = true; +} + +/** + * Compile one arithmetic operation COLOR&ALPHA pair into TGSI instructions. + */ +static void +compile_instruction(struct st_translate *t, + const struct atifs_instruction *inst) +{ + unsigned optype; + + for (optype = 0; optype < 2; optype++) { /* color, alpha */ + const struct instruction_desc *desc; + struct ureg_dst dst[1]; + struct ureg_src args[3]; /* arguments for the main operation */ + unsigned arg; + unsigned dstreg = inst->DstReg[optype].Index - GL_REG_0_ATI; + + if (!inst->Opcode[optype]) + continue; + + desc = &inst_desc[inst->Opcode[optype] - GL_MOV_ATI]; + + /* prepare the arguments */ + for (arg = 0; arg < desc->arg_count; arg++) { + if (arg >= inst->ArgCount[optype]) { + _mesa_warning(0, "Using 0 for missing argument %d of %s\n", + arg, desc->name); + args[arg] = ureg_imm1f(t->ureg, 0.0f); + } else { + args[arg] = prepare_argument(t, arg, + &inst->SrcReg[optype][arg]); + } + } + + /* prepare dst */ + dst[0] = get_temp(t, dstreg); + + if (optype) { + dst[0] = ureg_writemask(dst[0], TGSI_WRITEMASK_W); + } else { + GLuint dstMask = inst->DstReg[optype].dstMask; + if (dstMask == GL_NONE) { + dst[0] = ureg_writemask(dst[0], TGSI_WRITEMASK_XYZ); + } else { + dst[0] = ureg_writemask(dst[0], dstMask); /* the enum values match */ + } + } + + /* emit the main instruction */ + emit_arith_inst(t, desc, dst, args, arg); + + emit_dstmod(t, *dst, inst->DstReg[optype].dstMod); + + t->regs_written[t->current_pass][dstreg] = true; + } +} + +static void +finalize_shader(struct st_translate *t, unsigned numPasses) +{ + struct ureg_dst dst[1] = { { 0 } }; + struct ureg_src src[1] = { { 0 } }; + + if (t->regs_written[numPasses-1][0]) { + /* copy the result into the OUT slot */ + dst[0] = t->outputs[t->outputMapping[FRAG_RESULT_COLOR]]; + src[0] = ureg_src(t->temps[0]); + ureg_insn(t->ureg, TGSI_OPCODE_MOV, dst, 1, src, 1); + } + + /* signal the end of the program */ + ureg_insn(t->ureg, TGSI_OPCODE_END, dst, 0, src, 0); +} + +/** + * Called when a new variant is needed, we need to translate + * the ATI fragment shader to TGSI + */ +enum pipe_error +st_translate_atifs_program( + struct ureg_program *ureg, + struct ati_fragment_shader *atifs, + struct gl_program *program, + GLuint numInputs, + const GLuint inputMapping[], + const ubyte inputSemanticName[], + const ubyte inputSemanticIndex[], + const GLuint interpMode[], + GLuint numOutputs, + const GLuint outputMapping[], + const ubyte outputSemanticName[], + const ubyte outputSemanticIndex[]) +{ + enum pipe_error ret = PIPE_OK; + + unsigned pass, i, r; + + struct st_translate translate, *t; + t = &translate; + memset(t, 0, sizeof *t); + + t->inputMapping = inputMapping; + t->outputMapping = outputMapping; + t->ureg = ureg; + t->atifs = atifs; + + /* + * Declare input attributes. + */ + for (i = 0; i < numInputs; i++) { + t->inputs[i] = ureg_DECL_fs_input(ureg, + inputSemanticName[i], + inputSemanticIndex[i], + interpMode[i]); + } + + /* + * Declare output attributes: + * we always have numOutputs=1 and it's FRAG_RESULT_COLOR + */ + t->outputs[0] = ureg_DECL_output(ureg, + TGSI_SEMANTIC_COLOR, + outputSemanticIndex[0]); + + /* Emit constants and immediates. Mesa uses a single index space + * for these, so we put all the translated regs in t->constants. + */ + if (program->Parameters) { + t->constants = calloc(program->Parameters->NumParameters, + sizeof t->constants[0]); + if (t->constants == NULL) { + ret = PIPE_ERROR_OUT_OF_MEMORY; + goto out; + } + + for (i = 0; i < program->Parameters->NumParameters; i++) { + switch (program->Parameters->Parameters[i].Type) { + case PROGRAM_STATE_VAR: + case PROGRAM_UNIFORM: + t->constants[i] = ureg_DECL_constant(ureg, i); + break; + case PROGRAM_CONSTANT: + t->constants[i] = + ureg_DECL_immediate(ureg, + (const float*)program->Parameters->ParameterValues[i], + 4); + break; + default: + break; + } + } + } + + /* texture samplers */ + for (i = 0; i < MAX_NUM_FRAGMENT_REGISTERS_ATI; i++) { + if (program->SamplersUsed & (1 << i)) { + t->samplers[i] = ureg_DECL_sampler(ureg, i); + /* the texture target is still unknown, it will be fixed in the draw call */ + ureg_DECL_sampler_view(ureg, i, TGSI_TEXTURE_2D, + TGSI_RETURN_TYPE_FLOAT, + TGSI_RETURN_TYPE_FLOAT, + TGSI_RETURN_TYPE_FLOAT, + TGSI_RETURN_TYPE_FLOAT); + } + } + + /* emit instructions */ + for (pass = 0; pass < atifs->NumPasses; pass++) { + t->current_pass = pass; + for (r = 0; r < MAX_NUM_FRAGMENT_REGISTERS_ATI; r++) { + struct atifs_setupinst *texinst = &atifs->SetupInst[pass][r]; + compile_setupinst(t, r, texinst); + } + for (i = 0; i < atifs->numArithInstr[pass]; i++) { + struct atifs_instruction *inst = &atifs->Instructions[pass][i]; + compile_instruction(t, inst); + } + } + + finalize_shader(t, atifs->NumPasses); + +out: + free(t->constants); + + if (t->error) { + debug_printf("%s: translate error flag set\n", __func__); + } + + return ret; +} + +/** + * Called in ProgramStringNotify, we need to fill the metadata of the + * gl_program attached to the ati_fragment_shader + */ +void +st_init_atifs_prog(struct gl_context *ctx, struct gl_program *prog) +{ + /* we know this is st_fragment_program, because of st_new_ati_fs() */ + struct st_fragment_program *stfp = (struct st_fragment_program *) prog; + struct ati_fragment_shader *atifs = stfp->ati_fs; + + unsigned pass, i, r, optype, arg; + + static const gl_state_index fog_params_state[STATE_LENGTH] = + {STATE_INTERNAL, STATE_FOG_PARAMS_OPTIMIZED, 0, 0, 0}; + static const gl_state_index fog_color[STATE_LENGTH] = + {STATE_FOG_COLOR, 0, 0, 0, 0}; + + prog->InputsRead = 0; + prog->OutputsWritten = BITFIELD64_BIT(FRAG_RESULT_COLOR); + prog->SamplersUsed = 0; + prog->Parameters = _mesa_new_parameter_list(); + + /* fill in InputsRead, SamplersUsed, TexturesUsed */ + for (pass = 0; pass < atifs->NumPasses; pass++) { + for (r = 0; r < MAX_NUM_FRAGMENT_REGISTERS_ATI; r++) { + struct atifs_setupinst *texinst = &atifs->SetupInst[pass][r]; + GLuint pass_tex = texinst->src; + + if (texinst->Opcode == ATI_FRAGMENT_SHADER_SAMPLE_OP) { + /* mark which texcoords are used */ + prog->InputsRead |= BITFIELD64_BIT(VARYING_SLOT_TEX0 + pass_tex - GL_TEXTURE0_ARB); + /* by default there is 1:1 mapping between samplers and textures */ + prog->SamplersUsed |= (1 << r); + /* the target is unknown here, it will be fixed in the draw call */ + prog->TexturesUsed[r] = TEXTURE_2D_BIT; + } else if (texinst->Opcode == ATI_FRAGMENT_SHADER_PASS_OP) { + if (pass_tex >= GL_TEXTURE0_ARB && pass_tex <= GL_TEXTURE7_ARB) { + prog->InputsRead |= BITFIELD64_BIT(VARYING_SLOT_TEX0 + pass_tex - GL_TEXTURE0_ARB); + } + } + } + } + for (pass = 0; pass < atifs->NumPasses; pass++) { + for (i = 0; i < atifs->numArithInstr[pass]; i++) { + struct atifs_instruction *inst = &atifs->Instructions[pass][i]; + + for (optype = 0; optype < 2; optype++) { /* color, alpha */ + if (inst->Opcode[optype]) { + for (arg = 0; arg < inst->ArgCount[optype]; arg++) { + GLint index = inst->SrcReg[optype][arg].Index; + if (index == GL_PRIMARY_COLOR_EXT) { + prog->InputsRead |= BITFIELD64_BIT(VARYING_SLOT_COL0); + } else if (index == GL_SECONDARY_INTERPOLATOR_ATI) { + /* note: ATI_fragment_shader.txt never specifies what + * GL_SECONDARY_INTERPOLATOR_ATI is, swrast uses + * VARYING_SLOT_COL1 for this input */ + prog->InputsRead |= BITFIELD64_BIT(VARYING_SLOT_COL1); + } + } + } + } + } + } + /* we may need fog */ + prog->InputsRead |= BITFIELD64_BIT(VARYING_SLOT_FOGC); + + /* we always have the ATI_fs constants, and the fog params */ + for (i = 0; i < MAX_NUM_FRAGMENT_CONSTANTS_ATI; i++) { + _mesa_add_parameter(prog->Parameters, PROGRAM_UNIFORM, + NULL, 4, GL_FLOAT, NULL, NULL); + } + _mesa_add_state_reference(prog->Parameters, fog_params_state); + _mesa_add_state_reference(prog->Parameters, fog_color); + + prog->NumInstructions = 0; + prog->NumTemporaries = MAX_NUM_FRAGMENT_REGISTERS_ATI + 3; /* 3 input temps for arith ops */ + prog->NumParameters = MAX_NUM_FRAGMENT_CONSTANTS_ATI + 2; /* 2 state variables for fog */ +} + + +struct tgsi_atifs_transform { + struct tgsi_transform_context base; + struct tgsi_shader_info info; + const struct st_fp_variant_key *key; + bool first_instruction_emitted; + unsigned fog_factor_temp; + unsigned fog_clamp_imm; +}; + +static inline struct tgsi_atifs_transform * +tgsi_atifs_transform(struct tgsi_transform_context *tctx) +{ + return (struct tgsi_atifs_transform *)tctx; +} + +/* copied from st_cb_drawpixels_shader.c */ +static void +set_src(struct tgsi_full_instruction *inst, unsigned i, unsigned file, unsigned index, + unsigned x, unsigned y, unsigned z, unsigned w) +{ + inst->Src[i].Register.File = file; + inst->Src[i].Register.Index = index; + inst->Src[i].Register.SwizzleX = x; + inst->Src[i].Register.SwizzleY = y; + inst->Src[i].Register.SwizzleZ = z; + inst->Src[i].Register.SwizzleW = w; +} + +#define SET_SRC(inst, i, file, index, x, y, z, w) \ + set_src(inst, i, file, index, TGSI_SWIZZLE_##x, TGSI_SWIZZLE_##y, \ + TGSI_SWIZZLE_##z, TGSI_SWIZZLE_##w) + +static void +transform_decl(struct tgsi_transform_context *tctx, + struct tgsi_full_declaration *decl) +{ + struct tgsi_atifs_transform *ctx = tgsi_atifs_transform(tctx); + + if (decl->Declaration.File == TGSI_FILE_SAMPLER_VIEW) { + /* fix texture target */ + unsigned newtarget = ctx->key->texture_targets[decl->Range.First]; + if (newtarget) + decl->SamplerView.Resource = newtarget; + } + + tctx->emit_declaration(tctx, decl); +} + +static void +transform_instr(struct tgsi_transform_context *tctx, + struct tgsi_full_instruction *current_inst) +{ + struct tgsi_atifs_transform *ctx = tgsi_atifs_transform(tctx); + + if (ctx->first_instruction_emitted) + goto transform_inst; + + ctx->first_instruction_emitted = true; + + if (ctx->key->fog) { + /* add a new temp for the fog factor */ + ctx->fog_factor_temp = ctx->info.file_max[TGSI_FILE_TEMPORARY] + 1; + tgsi_transform_temp_decl(tctx, ctx->fog_factor_temp); + + /* add immediates for clamp */ + ctx->fog_clamp_imm = ctx->info.immediate_count; + tgsi_transform_immediate_decl(tctx, 1.0f, 0.0f, 0.0f, 0.0f); + } + +transform_inst: + if (current_inst->Instruction.Opcode == TGSI_OPCODE_TEX) { + /* fix texture target */ + unsigned newtarget = ctx->key->texture_targets[current_inst->Src[1].Register.Index]; + if (newtarget) + current_inst->Texture.Texture = newtarget; + + } else if (ctx->key->fog && current_inst->Instruction.Opcode == TGSI_OPCODE_MOV && + current_inst->Dst[0].Register.File == TGSI_FILE_OUTPUT) { + struct tgsi_full_instruction inst; + unsigned i; + int fogc_index = -1; + + /* find FOGC input */ + for (i = 0; i < ctx->info.num_inputs; i++) { + if (ctx->info.input_semantic_name[i] == TGSI_SEMANTIC_FOG) { + fogc_index = i; + break; + } + } + if (fogc_index < 0) { + /* should never be reached, because fog coord input is always declared */ + tctx->emit_instruction(tctx, current_inst); + return; + } + + /* compute the 1 component fog factor f */ + if (ctx->key->fog == 1) { + /* LINEAR formula: f = (end - z) / (end - start) + * with optimized parameters: + * f = MAD(fogcoord, oparams.x, oparams.y) + */ + inst = tgsi_default_full_instruction(); + inst.Instruction.Opcode = TGSI_OPCODE_MAD; + inst.Instruction.NumDstRegs = 1; + inst.Dst[0].Register.File = TGSI_FILE_TEMPORARY; + inst.Dst[0].Register.Index = ctx->fog_factor_temp; + inst.Dst[0].Register.WriteMask = TGSI_WRITEMASK_XYZW; + inst.Instruction.NumSrcRegs = 3; + SET_SRC(&inst, 0, TGSI_FILE_INPUT, fogc_index, X, Y, Z, W); + SET_SRC(&inst, 1, TGSI_FILE_CONSTANT, MAX_NUM_FRAGMENT_CONSTANTS_ATI, X, X, X, X); + SET_SRC(&inst, 2, TGSI_FILE_CONSTANT, MAX_NUM_FRAGMENT_CONSTANTS_ATI, Y, Y, Y, Y); + tctx->emit_instruction(tctx, &inst); + } else if (ctx->key->fog == 2) { + /* EXP formula: f = exp(-dens * z) + * with optimized parameters: + * f = MUL(fogcoord, oparams.z); f= EX2(-f) + */ + inst = tgsi_default_full_instruction(); + inst.Instruction.Opcode = TGSI_OPCODE_MUL; + inst.Instruction.NumDstRegs = 1; + inst.Dst[0].Register.File = TGSI_FILE_TEMPORARY; + inst.Dst[0].Register.Index = ctx->fog_factor_temp; + inst.Dst[0].Register.WriteMask = TGSI_WRITEMASK_XYZW; + inst.Instruction.NumSrcRegs = 2; + SET_SRC(&inst, 0, TGSI_FILE_INPUT, fogc_index, X, Y, Z, W); + SET_SRC(&inst, 1, TGSI_FILE_CONSTANT, MAX_NUM_FRAGMENT_CONSTANTS_ATI, Z, Z, Z, Z); + tctx->emit_instruction(tctx, &inst); + + inst = tgsi_default_full_instruction(); + inst.Instruction.Opcode = TGSI_OPCODE_EX2; + inst.Instruction.NumDstRegs = 1; + inst.Dst[0].Register.File = TGSI_FILE_TEMPORARY; + inst.Dst[0].Register.Index = ctx->fog_factor_temp; + inst.Dst[0].Register.WriteMask = TGSI_WRITEMASK_XYZW; + inst.Instruction.NumSrcRegs = 1; + SET_SRC(&inst, 0, TGSI_FILE_TEMPORARY, ctx->fog_factor_temp, X, Y, Z, W); + inst.Src[0].Register.Negate = 1; + tctx->emit_instruction(tctx, &inst); + } else if (ctx->key->fog == 3) { + /* EXP2 formula: f = exp(-(dens * z)^2) + * with optimized parameters: + * f = MUL(fogcoord, oparams.w); f=MUL(f, f); f= EX2(-f) + */ + inst = tgsi_default_full_instruction(); + inst.Instruction.Opcode = TGSI_OPCODE_MUL; + inst.Instruction.NumDstRegs = 1; + inst.Dst[0].Register.File = TGSI_FILE_TEMPORARY; + inst.Dst[0].Register.Index = ctx->fog_factor_temp; + inst.Dst[0].Register.WriteMask = TGSI_WRITEMASK_XYZW; + inst.Instruction.NumSrcRegs = 2; + SET_SRC(&inst, 0, TGSI_FILE_INPUT, fogc_index, X, Y, Z, W); + SET_SRC(&inst, 1, TGSI_FILE_CONSTANT, MAX_NUM_FRAGMENT_CONSTANTS_ATI, W, W, W, W); + tctx->emit_instruction(tctx, &inst); + + inst = tgsi_default_full_instruction(); + inst.Instruction.Opcode = TGSI_OPCODE_MUL; + inst.Instruction.NumDstRegs = 1; + inst.Dst[0].Register.File = TGSI_FILE_TEMPORARY; + inst.Dst[0].Register.Index = ctx->fog_factor_temp; + inst.Dst[0].Register.WriteMask = TGSI_WRITEMASK_XYZW; + inst.Instruction.NumSrcRegs = 2; + SET_SRC(&inst, 0, TGSI_FILE_TEMPORARY, ctx->fog_factor_temp, X, Y, Z, W); + SET_SRC(&inst, 1, TGSI_FILE_TEMPORARY, ctx->fog_factor_temp, X, Y, Z, W); + tctx->emit_instruction(tctx, &inst); + + inst = tgsi_default_full_instruction(); + inst.Instruction.Opcode = TGSI_OPCODE_EX2; + inst.Instruction.NumDstRegs = 1; + inst.Dst[0].Register.File = TGSI_FILE_TEMPORARY; + inst.Dst[0].Register.Index = ctx->fog_factor_temp; + inst.Dst[0].Register.WriteMask = TGSI_WRITEMASK_XYZW; + inst.Instruction.NumSrcRegs = 1; + SET_SRC(&inst, 0, TGSI_FILE_TEMPORARY, ctx->fog_factor_temp, X, Y, Z, W); + inst.Src[0].Register.Negate ^= 1; + tctx->emit_instruction(tctx, &inst); + } + /* f = CLAMP(f, 0.0, 1.0) */ + inst = tgsi_default_full_instruction(); + inst.Instruction.Opcode = TGSI_OPCODE_CLAMP; + inst.Instruction.NumDstRegs = 1; + inst.Dst[0].Register.File = TGSI_FILE_TEMPORARY; + inst.Dst[0].Register.Index = ctx->fog_factor_temp; + inst.Dst[0].Register.WriteMask = TGSI_WRITEMASK_XYZW; + inst.Instruction.NumSrcRegs = 3; + SET_SRC(&inst, 0, TGSI_FILE_TEMPORARY, ctx->fog_factor_temp, X, Y, Z, W); + SET_SRC(&inst, 1, TGSI_FILE_IMMEDIATE, ctx->fog_clamp_imm, Y, Y, Y, Y); // 0.0 + SET_SRC(&inst, 2, TGSI_FILE_IMMEDIATE, ctx->fog_clamp_imm, X, X, X, X); // 1.0 + tctx->emit_instruction(tctx, &inst); + + /* REG0 = LRP(f, REG0, fogcolor) */ + inst = tgsi_default_full_instruction(); + inst.Instruction.Opcode = TGSI_OPCODE_LRP; + inst.Instruction.NumDstRegs = 1; + inst.Dst[0].Register.File = TGSI_FILE_TEMPORARY; + inst.Dst[0].Register.Index = 0; + inst.Dst[0].Register.WriteMask = TGSI_WRITEMASK_XYZW; + inst.Instruction.NumSrcRegs = 3; + SET_SRC(&inst, 0, TGSI_FILE_TEMPORARY, ctx->fog_factor_temp, X, X, X, Y); + SET_SRC(&inst, 1, TGSI_FILE_TEMPORARY, 0, X, Y, Z, W); + SET_SRC(&inst, 2, TGSI_FILE_CONSTANT, MAX_NUM_FRAGMENT_CONSTANTS_ATI + 1, X, Y, Z, W); + tctx->emit_instruction(tctx, &inst); + } + + tctx->emit_instruction(tctx, current_inst); +} + +/* + * A post-process step in the draw call to fix texture targets and + * insert code for fog. + */ +const struct tgsi_token * +st_fixup_atifs(const struct tgsi_token *tokens, + const struct st_fp_variant_key *key) +{ + struct tgsi_atifs_transform ctx; + struct tgsi_token *newtoks; + int newlen; + + memset(&ctx, 0, sizeof(ctx)); + ctx.base.transform_declaration = transform_decl; + ctx.base.transform_instruction = transform_instr; + ctx.key = key; + tgsi_scan_shader(tokens, &ctx.info); + + newlen = tgsi_num_tokens(tokens) + 30; + newtoks = tgsi_alloc_tokens(newlen); + if (!newtoks) + return NULL; + + tgsi_transform_shader(tokens, newtoks, newlen, &ctx.base); + return newtoks; +} + diff --git a/src/mesa/state_tracker/st_atifs_to_tgsi.h b/src/mesa/state_tracker/st_atifs_to_tgsi.h new file mode 100644 index 00000000000..c1b6758ba02 --- /dev/null +++ b/src/mesa/state_tracker/st_atifs_to_tgsi.h @@ -0,0 +1,67 @@ +/* + * Copyright (C) 2016 Miklós Máté + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#ifndef ST_ATIFS_TO_TGSI_H +#define ST_ATIFS_TO_TGSI_H + +#if defined __cplusplus +extern "C" { +#endif + +#include "main/glheader.h" +#include "pipe/p_defines.h" + +struct gl_context; +struct gl_program; +struct ureg_program; +struct tgsi_token; +struct ati_fragment_shader; +struct st_fp_variant_key; + +enum pipe_error +st_translate_atifs_program( + struct ureg_program *ureg, + struct ati_fragment_shader *atifs, + struct gl_program *program, + GLuint numInputs, + const GLuint inputMapping[], + const ubyte inputSemanticName[], + const ubyte inputSemanticIndex[], + const GLuint interpMode[], + GLuint numOutputs, + const GLuint outputMapping[], + const ubyte outputSemanticName[], + const ubyte outputSemanticIndex[]); + + +void +st_init_atifs_prog(struct gl_context *ctx, struct gl_program *prog); + +const struct tgsi_token * +st_fixup_atifs(const struct tgsi_token *tokens, + const struct st_fp_variant_key *key); + +#if defined __cplusplus +} /* extern "C" */ +#endif + +#endif /* ST_ATIFS_TO_TGSI_H */ diff --git a/src/mesa/state_tracker/st_atom_constbuf.c b/src/mesa/state_tracker/st_atom_constbuf.c index 407dfd31c80..a980dbedac5 100644 --- a/src/mesa/state_tracker/st_atom_constbuf.c +++ b/src/mesa/state_tracker/st_atom_constbuf.c @@ -64,6 +64,21 @@ void st_upload_constants( struct st_context *st, shader_type == PIPE_SHADER_TESS_EVAL || shader_type == PIPE_SHADER_COMPUTE); + /* update the ATI constants before rendering */ + if (shader_type == PIPE_SHADER_FRAGMENT && st->fp->ati_fs) { + struct ati_fragment_shader *ati_fs = st->fp->ati_fs; + unsigned c; + + for (c = 0; c < MAX_NUM_FRAGMENT_CONSTANTS_ATI; c++) { + if (ati_fs->LocalConstDef & (1 << c)) + memcpy(params->ParameterValues[c], + ati_fs->Constants[c], sizeof(GLfloat) * 4); + else + memcpy(params->ParameterValues[c], + st->ctx->ATIFragmentShader.GlobalConstants[c], sizeof(GLfloat) * 4); + } + } + /* update constants */ if (params && params->NumParameters) { struct pipe_constant_buffer cb; diff --git a/src/mesa/state_tracker/st_atom_sampler.c b/src/mesa/state_tracker/st_atom_sampler.c index 82dcf5ee0ca..a1cfa1c34c5 100644 --- a/src/mesa/state_tracker/st_atom_sampler.c +++ b/src/mesa/state_tracker/st_atom_sampler.c @@ -133,18 +133,19 @@ convert_sampler(struct st_context *st, { const struct gl_texture_object *texobj; struct gl_context *ctx = st->ctx; - struct gl_sampler_object *msamp; + const struct gl_sampler_object *msamp; GLenum texBaseFormat; texobj = ctx->Texture.Unit[texUnit]._Current; if (!texobj) { texobj = _mesa_get_fallback_texture(ctx, TEXTURE_2D_INDEX); + msamp = &texobj->Sampler; + } else { + msamp = _mesa_get_samplerobj(ctx, texUnit); } texBaseFormat = _mesa_texture_base_format(texobj); - msamp = _mesa_get_samplerobj(ctx, texUnit); - memset(sampler, 0, sizeof(*sampler)); sampler->wrap_s = gl_wrap_xlate(msamp->WrapS); sampler->wrap_t = gl_wrap_xlate(msamp->WrapT); diff --git a/src/mesa/state_tracker/st_atom_shader.c b/src/mesa/state_tracker/st_atom_shader.c index 709f0cbcb91..d0c2429dcef 100644 --- a/src/mesa/state_tracker/st_atom_shader.c +++ b/src/mesa/state_tracker/st_atom_shader.c @@ -38,18 +38,69 @@ #include "main/imports.h" #include "main/mtypes.h" #include "main/framebuffer.h" +#include "main/texobj.h" +#include "main/texstate.h" #include "program/program.h" #include "pipe/p_context.h" #include "pipe/p_shader_tokens.h" #include "util/u_simple_shaders.h" #include "cso_cache/cso_context.h" +#include "util/u_debug.h" #include "st_context.h" #include "st_atom.h" #include "st_program.h" +/** Compress the fog function enums into a 2-bit value */ +static GLuint +translate_fog_mode(GLenum mode) +{ + switch (mode) { + case GL_LINEAR: return 1; + case GL_EXP: return 2; + case GL_EXP2: return 3; + default: + return 0; + } +} + +static unsigned +get_texture_target(struct gl_context *ctx, const unsigned unit) +{ + struct gl_texture_object *texObj = _mesa_get_tex_unit(ctx, unit)->_Current; + gl_texture_index index; + + if (texObj) { + index = _mesa_tex_target_to_index(ctx, texObj->Target); + } else { + /* fallback for missing texture */ + index = TEXTURE_2D_INDEX; + } + + /* Map mesa texture target to TGSI texture target. + * Copied from st_mesa_to_tgsi.c, the shadow part is omitted */ + switch(index) { + case TEXTURE_2D_MULTISAMPLE_INDEX: return TGSI_TEXTURE_2D_MSAA; + case TEXTURE_2D_MULTISAMPLE_ARRAY_INDEX: return TGSI_TEXTURE_2D_ARRAY_MSAA; + case TEXTURE_BUFFER_INDEX: return TGSI_TEXTURE_BUFFER; + case TEXTURE_1D_INDEX: return TGSI_TEXTURE_1D; + case TEXTURE_2D_INDEX: return TGSI_TEXTURE_2D; + case TEXTURE_3D_INDEX: return TGSI_TEXTURE_3D; + case TEXTURE_CUBE_INDEX: return TGSI_TEXTURE_CUBE; + case TEXTURE_CUBE_ARRAY_INDEX: return TGSI_TEXTURE_CUBE_ARRAY; + case TEXTURE_RECT_INDEX: return TGSI_TEXTURE_RECT; + case TEXTURE_1D_ARRAY_INDEX: return TGSI_TEXTURE_1D_ARRAY; + case TEXTURE_2D_ARRAY_INDEX: return TGSI_TEXTURE_2D_ARRAY; + case TEXTURE_EXTERNAL_INDEX: return TGSI_TEXTURE_2D; + default: + debug_assert(0); + return TGSI_TEXTURE_1D; + } +} + + /** * Update fragment program state/atom. This involves translating the * Mesa fragment program into a gallium fragment program and binding it. @@ -79,6 +130,18 @@ update_fp( struct st_context *st ) st->ctx->Multisample.MinSampleShadingValue * _mesa_geometric_samples(st->ctx->DrawBuffer) > 1; + if (stfp->ati_fs) { + unsigned u; + + if (st->ctx->Fog.Enabled) { + key.fog = translate_fog_mode(st->ctx->Fog.Mode); + } + + for (u = 0; u < MAX_NUM_FRAGMENT_REGISTERS_ATI; u++) { + key.texture_targets[u] = get_texture_target(st->ctx, u); + } + } + st->fp_variant = st_get_fp_variant(st, stfp, &key); st_reference_fragprog(st, &st->fp, stfp); @@ -91,7 +154,7 @@ update_fp( struct st_context *st ) const struct st_tracked_state st_update_fp = { "st_update_fp", /* name */ { /* dirty */ - _NEW_BUFFERS | _NEW_MULTISAMPLE, /* mesa */ + _NEW_BUFFERS | _NEW_MULTISAMPLE | _NEW_FOG, /* mesa */ ST_NEW_FRAGMENT_PROGRAM /* st */ }, update_fp /* update */ diff --git a/src/mesa/state_tracker/st_cb_drawpixels.c b/src/mesa/state_tracker/st_cb_drawpixels.c index 09f4d8e00d1..01ed5441d11 100644 --- a/src/mesa/state_tracker/st_cb_drawpixels.c +++ b/src/mesa/state_tracker/st_cb_drawpixels.c @@ -1302,6 +1302,7 @@ blit_copy_pixels(struct gl_context *ctx, GLint srcx, GLint srcy, !ctx->FragmentProgram.Enabled && !ctx->VertexProgram.Enabled && !ctx->_Shader->CurrentProgram[MESA_SHADER_FRAGMENT] && + !ctx->ATIFragmentShader._Enabled && ctx->DrawBuffer->_NumColorDrawBuffers == 1 && !ctx->Query.CondRenderQuery && !ctx->Query.CurrentOcclusionObject) { diff --git a/src/mesa/state_tracker/st_cb_program.c b/src/mesa/state_tracker/st_cb_program.c index 27cc0f3d154..d79cfe239e4 100644 --- a/src/mesa/state_tracker/st_cb_program.c +++ b/src/mesa/state_tracker/st_cb_program.c @@ -46,6 +46,7 @@ #include "st_mesa_to_tgsi.h" #include "st_cb_program.h" #include "st_glsl_to_tgsi.h" +#include "st_atifs_to_tgsi.h" @@ -302,6 +303,22 @@ st_program_string_notify( struct gl_context *ctx, if (st->cp == stcp) st->dirty_cp.st |= ST_NEW_COMPUTE_PROGRAM; } + else if (target == GL_FRAGMENT_SHADER_ATI) { + assert(prog); + + struct st_fragment_program *stfp = (struct st_fragment_program *) prog; + assert(stfp->ati_fs); + assert(stfp->ati_fs->Program == prog); + + st_init_atifs_prog(ctx, prog); + + st_release_fp_variants(st, stfp); + if (!st_translate_fragment_program(st, stfp)) + return false; + + if (st->fp == stfp) + st->dirty.st |= ST_NEW_FRAGMENT_PROGRAM; + } if (ST_DEBUG & DEBUG_PRECOMPILE || st->shader_has_one_variant[stage]) @@ -310,6 +327,19 @@ st_program_string_notify( struct gl_context *ctx, return GL_TRUE; } +/** + * Called via ctx->Driver.NewATIfs() + * Called in glEndFragmentShaderATI() + */ +static struct gl_program * +st_new_ati_fs(struct gl_context *ctx, struct ati_fragment_shader *curProg) +{ + struct gl_program *prog = ctx->Driver.NewProgram(ctx, GL_FRAGMENT_PROGRAM_ARB, + curProg->Id); + struct st_fragment_program *stfp = (struct st_fragment_program *)prog; + stfp->ati_fs = curProg; + return prog; +} /** * Plug in the program and shader-related device driver functions. @@ -322,6 +352,7 @@ st_init_program_functions(struct dd_function_table *functions) functions->NewProgram = st_new_program; functions->DeleteProgram = st_delete_program; functions->ProgramStringNotify = st_program_string_notify; + functions->NewATIfs = st_new_ati_fs; functions->LinkShader = st_link_shader; } diff --git a/src/mesa/state_tracker/st_cb_texture.c b/src/mesa/state_tracker/st_cb_texture.c index 460c1790663..3980f5d2f51 100644 --- a/src/mesa/state_tracker/st_cb_texture.c +++ b/src/mesa/state_tracker/st_cb_texture.c @@ -2886,12 +2886,17 @@ st_finalize_texture(struct gl_context *ctx, /* Need to import images in main memory or held in other textures. */ if (stImage && stObj->pt != stImage->pt) { + GLuint height = stObj->height0; GLuint depth = stObj->depth0; + + if (stObj->base.Target != GL_TEXTURE_1D_ARRAY) + height = u_minify(height, level); if (stObj->base.Target == GL_TEXTURE_3D) depth = u_minify(depth, level); + if (level == 0 || (stImage->base.Width == u_minify(stObj->width0, level) && - stImage->base.Height == u_minify(stObj->height0, level) && + stImage->base.Height == height && stImage->base.Depth == depth)) { /* src image fits expected dest mipmap level size */ copy_image_data_to_texture(st, stObj, level, stImage); diff --git a/src/mesa/state_tracker/st_cb_xformfb.c b/src/mesa/state_tracker/st_cb_xformfb.c index 0c01cd5ab78..a5cf3dfd5a9 100644 --- a/src/mesa/state_tracker/st_cb_xformfb.c +++ b/src/mesa/state_tracker/st_cb_xformfb.c @@ -125,7 +125,7 @@ st_begin_transform_feedback(struct gl_context *ctx, GLenum mode, if (bo && bo->buffer) { unsigned stream = - obj->shader_program->LinkedTransformFeedback.BufferStream[i]; + obj->shader_program->LinkedTransformFeedback.Buffers[i].Stream; /* Check whether we need to recreate the target. */ if (!sobj->targets[i] || @@ -204,7 +204,7 @@ st_end_transform_feedback(struct gl_context *ctx, for (i = 0; i < ARRAY_SIZE(sobj->targets); i++) { unsigned stream = - obj->shader_program->LinkedTransformFeedback.BufferStream[i]; + obj->shader_program->LinkedTransformFeedback.Buffers[i].Stream; /* Is it not bound or already set for this stream? */ if (!sobj->targets[i] || sobj->draw_count[stream]) diff --git a/src/mesa/state_tracker/st_draw.c b/src/mesa/state_tracker/st_draw.c index fdd59a383a9..3db5749725e 100644 --- a/src/mesa/state_tracker/st_draw.c +++ b/src/mesa/state_tracker/st_draw.c @@ -127,35 +127,6 @@ setup_index_buffer(struct st_context *st, /** - * Prior to drawing, check that any uniforms referenced by the - * current shader have been set. If a uniform has not been set, - * issue a warning. - */ -static void -check_uniforms(struct gl_context *ctx) -{ - struct gl_shader_program **shProg = ctx->_Shader->CurrentProgram; - unsigned j; - - for (j = 0; j < 3; j++) { - unsigned i; - - if (shProg[j] == NULL || !shProg[j]->LinkStatus) - continue; - - for (i = 0; i < shProg[j]->NumUniformStorage; i++) { - const struct gl_uniform_storage *u = &shProg[j]->UniformStorage[i]; - if (!u->initialized) { - _mesa_warning(ctx, - "Using shader with uninitialized uniform: %s", - u->name); - } - } - } -} - - -/** * Translate OpenGL primtive type (GL_POINTS, GL_TRIANGLE_STRIP, etc) to * the corresponding Gallium type. */ @@ -203,14 +174,6 @@ st_draw_vbo(struct gl_context *ctx, /* Validate state. */ if (st->dirty.st || st->dirty.mesa || ctx->NewDriverState) { st_validate_state(st, ST_PIPELINE_RENDER); - -#if 0 - if (MESA_VERBOSE & VERBOSE_GLSL) { - check_uniforms(ctx); - } -#else - (void) check_uniforms; -#endif } if (st->vertex_array_out_of_memory) { diff --git a/src/mesa/state_tracker/st_extensions.c b/src/mesa/state_tracker/st_extensions.c index 2fdaba073a2..8748ab5c876 100644 --- a/src/mesa/state_tracker/st_extensions.c +++ b/src/mesa/state_tracker/st_extensions.c @@ -552,7 +552,6 @@ void st_init_extensions(struct pipe_screen *screen, boolean has_lib_dxtc) { unsigned i; - int glsl_feature_level; GLboolean *extension_table = (GLboolean *) extensions; static const struct st_extension_cap_mapping cap_mapping[] = { @@ -811,6 +810,7 @@ void st_init_extensions(struct pipe_screen *screen, extensions->EXT_texture_env_dot3 = GL_TRUE; extensions->EXT_vertex_array_bgra = GL_TRUE; + extensions->ATI_fragment_shader = GL_TRUE; extensions->ATI_texture_env_combine3 = GL_TRUE; extensions->MESA_pack_invert = GL_TRUE; @@ -844,12 +844,8 @@ void st_init_extensions(struct pipe_screen *screen, ARRAY_SIZE(vertex_mapping), PIPE_BUFFER, PIPE_BIND_VERTEX_BUFFER); - /* Figure out GLSL support. */ - glsl_feature_level = screen->get_param(screen, PIPE_CAP_GLSL_FEATURE_LEVEL); - - consts->GLSLVersion = glsl_feature_level; - if (glsl_feature_level >= 410) - consts->GLSLVersion = 410; + /* Figure out GLSL support and set GLSLVersion to it. */ + consts->GLSLVersion = screen->get_param(screen, PIPE_CAP_GLSL_FEATURE_LEVEL); _mesa_override_glsl_version(consts); @@ -858,9 +854,9 @@ void st_init_extensions(struct pipe_screen *screen, consts->ForceGLSLVersion = options->force_glsl_version; } - if (glsl_feature_level >= 400) + if (consts->GLSLVersion >= 400) extensions->ARB_gpu_shader5 = GL_TRUE; - if (glsl_feature_level >= 410) + if (consts->GLSLVersion >= 410) extensions->ARB_shader_precision = GL_TRUE; /* This extension needs full OpenGL 3.2, but we don't know if that's @@ -925,6 +921,23 @@ void st_init_extensions(struct pipe_screen *screen, extensions->ARB_sync = GL_TRUE; } + /* Needs PIPE_CAP_SAMPLE_SHADING + all the sample-related bits of + * ARB_gpu_shader5. This enables all the per-sample shading ES extensions. + */ + extensions->OES_sample_variables = extensions->ARB_sample_shading && + extensions->ARB_gpu_shader5; + + /* If we don't have native ETC2 support, we don't keep track of the + * original ETC2 data. This is necessary to be able to copy images between + * compatible view classes. + */ + if (extensions->ARB_copy_image && screen->is_format_supported( + screen, PIPE_FORMAT_ETC2_RGB8, + PIPE_TEXTURE_2D, 0, + PIPE_BIND_SAMPLER_VIEW)) { + extensions->OES_copy_image = GL_TRUE; + } + /* Maximum sample count. */ { enum pipe_format color_formats[] = { @@ -1020,6 +1033,12 @@ void st_init_extensions(struct pipe_screen *screen, PIPE_BIND_SAMPLER_VIEW); } + extensions->OES_texture_buffer = + extensions->ARB_texture_buffer_object && + extensions->ARB_texture_buffer_range && + extensions->ARB_texture_buffer_object_rgb32 && + extensions->ARB_shader_image_load_store; + /* Unpacking a varying in the fragment shader costs 1 texture indirection. * If the number of available texture indirections is very limited, then we * prefer to disable varying packing rather than run the risk of varying @@ -1036,7 +1055,7 @@ void st_init_extensions(struct pipe_screen *screen, consts->MaxViewports = screen->get_param(screen, PIPE_CAP_MAX_VIEWPORTS); if (consts->MaxViewports >= 16) { - if (glsl_feature_level >= 400) { + if (consts->GLSLVersion >= 400) { consts->ViewportBounds.Min = -32768.0; consts->ViewportBounds.Max = 32767.0; } else { diff --git a/src/mesa/state_tracker/st_gen_mipmap.c b/src/mesa/state_tracker/st_gen_mipmap.c index c4b3492b0d3..a14bbfabaa3 100644 --- a/src/mesa/state_tracker/st_gen_mipmap.c +++ b/src/mesa/state_tracker/st_gen_mipmap.c @@ -82,7 +82,6 @@ st_generate_mipmap(struct gl_context *ctx, GLenum target, const uint baseLevel = texObj->BaseLevel; enum pipe_format format; uint lastLevel, first_layer, last_layer; - uint dstLevel; if (!pt) return; @@ -103,42 +102,33 @@ st_generate_mipmap(struct gl_context *ctx, GLenum target, stObj->lastLevel = lastLevel; if (!texObj->Immutable) { - if (pt->last_level < lastLevel) { - /* The current gallium texture doesn't have space for all the - * mipmap levels we need to generate. So allocate a new texture. - */ - struct pipe_resource *oldTex = stObj->pt; - - /* create new texture with space for more levels */ - stObj->pt = st_texture_create(st, - oldTex->target, - oldTex->format, - lastLevel, - oldTex->width0, - oldTex->height0, - oldTex->depth0, - oldTex->array_size, - 0, - oldTex->bind); - - /* This will copy the old texture's base image into the new texture - * which we just allocated. - */ - st_finalize_texture(ctx, st->pipe, texObj); - - /* release the old tex (will likely be freed too) */ - pipe_resource_reference(&oldTex, NULL); - st_texture_release_all_sampler_views(st, stObj); - } - else { - /* Make sure that the base texture image data is present in the - * texture buffer. - */ - st_finalize_texture(ctx, st->pipe, texObj); - } + const GLboolean genSave = texObj->GenerateMipmap; + + /* Temporarily set GenerateMipmap to true so that allocate_full_mipmap() + * makes the right decision about full mipmap allocation. + */ + texObj->GenerateMipmap = GL_TRUE; + + _mesa_prepare_mipmap_levels(ctx, texObj, baseLevel, lastLevel); + + texObj->GenerateMipmap = genSave; + + /* At this point, memory for all the texture levels has been + * allocated. However, the base level image may be in one resource + * while the subsequent/smaller levels may be in another resource. + * Finalizing the texture will copy the base images from the former + * resource to the latter. + * + * After this, we'll have all mipmap levels in one resource. + */ + st_finalize_texture(ctx, st->pipe, texObj); } pt = stObj->pt; + if (!pt) { + _mesa_error(ctx, GL_OUT_OF_MEMORY, "mipmap generation"); + return; + } assert(pt->last_level >= lastLevel); @@ -169,48 +159,4 @@ st_generate_mipmap(struct gl_context *ctx, GLenum target, _mesa_generate_mipmap(ctx, target, texObj); } } - - /* Fill in the Mesa gl_texture_image fields */ - for (dstLevel = baseLevel + 1; dstLevel <= lastLevel; dstLevel++) { - const uint srcLevel = dstLevel - 1; - const struct gl_texture_image *srcImage - = _mesa_get_tex_image(ctx, texObj, target, srcLevel); - struct gl_texture_image *dstImage; - struct st_texture_image *stImage; - uint border = srcImage->Border; - uint dstWidth, dstHeight, dstDepth; - - dstWidth = u_minify(pt->width0, dstLevel); - if (texObj->Target == GL_TEXTURE_1D_ARRAY) { - dstHeight = pt->array_size; - } - else { - dstHeight = u_minify(pt->height0, dstLevel); - } - if (texObj->Target == GL_TEXTURE_2D_ARRAY || - texObj->Target == GL_TEXTURE_CUBE_MAP_ARRAY) { - dstDepth = pt->array_size; - } - else { - dstDepth = u_minify(pt->depth0, dstLevel); - } - - dstImage = _mesa_get_tex_image(ctx, texObj, target, dstLevel); - if (!dstImage) { - _mesa_error(ctx, GL_OUT_OF_MEMORY, "generating mipmaps"); - return; - } - - /* Free old image data */ - ctx->Driver.FreeTextureImageBuffer(ctx, dstImage); - - /* initialize new image */ - _mesa_init_teximage_fields(ctx, dstImage, dstWidth, dstHeight, - dstDepth, border, srcImage->InternalFormat, - srcImage->TexFormat); - - stImage = st_texture_image(dstImage); - - pipe_resource_reference(&stImage->pt, pt); - } } diff --git a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp index 06b4bb41a9b..23786b85529 100644 --- a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp +++ b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp @@ -6811,7 +6811,7 @@ st_link_shader(struct gl_context *ctx, struct gl_shader_program *prog) validate_ir_tree(ir); } - build_program_resource_list(prog); + build_program_resource_list(ctx, prog); for (unsigned i = 0; i < MESA_SHADER_STAGES; i++) { struct gl_program *linked_prog; @@ -6861,7 +6861,7 @@ st_translate_stream_output_info(glsl_to_tgsi_visitor *glsl_to_tgsi, } for (i = 0; i < PIPE_MAX_SO_BUFFERS; i++) { - so->stride[i] = info->BufferStride[i]; + so->stride[i] = info->Buffers[i].Stride; } so->num_outputs = info->NumOutputs; } diff --git a/src/mesa/state_tracker/st_program.c b/src/mesa/state_tracker/st_program.c index 80dcfd82743..94dc48971ec 100644 --- a/src/mesa/state_tracker/st_program.c +++ b/src/mesa/state_tracker/st_program.c @@ -53,6 +53,7 @@ #include "st_context.h" #include "st_program.h" #include "st_mesa_to_tgsi.h" +#include "st_atifs_to_tgsi.h" #include "cso_cache/cso_context.h" @@ -811,7 +812,22 @@ st_translate_fragment_program(struct st_context *st, free_glsl_to_tgsi_visitor(stfp->glsl_to_tgsi); stfp->glsl_to_tgsi = NULL; - } else + } else if (stfp->ati_fs) + st_translate_atifs_program(ureg, + stfp->ati_fs, + &stfp->Base.Base, + /* inputs */ + fs_num_inputs, + inputMapping, + input_semantic_name, + input_semantic_index, + interpMode, + /* outputs */ + fs_num_outputs, + outputMapping, + fs_output_semantic_name, + fs_output_semantic_index); + else st_translate_mesa_program(st->ctx, TGSI_PROCESSOR_FRAGMENT, ureg, @@ -849,6 +865,16 @@ st_create_fp_variant(struct st_context *st, assert(!(key->bitmap && key->drawpixels)); + /* Fix texture targets and add fog for ATI_fs */ + if (stfp->ati_fs) { + const struct tgsi_token *tokens = st_fixup_atifs(tgsi.tokens, key); + + if (tokens) + tgsi.tokens = tokens; + else + fprintf(stderr, "mesa: cannot post-process ATI_fs\n"); + } + /* Emulate features. */ if (key->clamp_color || key->persample_shading) { const struct tgsi_token *tokens; @@ -858,9 +884,11 @@ st_create_fp_variant(struct st_context *st, tokens = tgsi_emulate(tgsi.tokens, flags); - if (tokens) + if (tokens) { + if (tgsi.tokens != stfp->tgsi.tokens) + tgsi_free_tokens(tgsi.tokens); tgsi.tokens = tokens; - else + } else fprintf(stderr, "mesa: cannot emulate deprecated features\n"); } diff --git a/src/mesa/state_tracker/st_program.h b/src/mesa/state_tracker/st_program.h index 028fba99a74..7c90fd74e14 100644 --- a/src/mesa/state_tracker/st_program.h +++ b/src/mesa/state_tracker/st_program.h @@ -35,6 +35,7 @@ #define ST_PROGRAM_H #include "main/mtypes.h" +#include "main/atifragshader.h" #include "program/program.h" #include "pipe/p_state.h" #include "st_context.h" @@ -65,6 +66,12 @@ struct st_fp_variant_key /** for ARB_sample_shading */ GLuint persample_shading:1; + + /** needed for ATI_fragment_shader */ + GLuint fog:2; + + /** needed for ATI_fragment_shader */ + char texture_targets[MAX_NUM_FRAGMENT_REGISTERS_ATI]; }; @@ -99,6 +106,7 @@ struct st_fragment_program struct gl_fragment_program Base; struct pipe_shader_state tgsi; struct glsl_to_tgsi_visitor* glsl_to_tgsi; + struct ati_fragment_shader *ati_fs; struct st_fp_variant *variants; }; diff --git a/src/mesa/state_tracker/st_vdpau.c b/src/mesa/state_tracker/st_vdpau.c index 71dd15bc4fe..b9abebfc7bf 100644 --- a/src/mesa/state_tracker/st_vdpau.c +++ b/src/mesa/state_tracker/st_vdpau.c @@ -39,8 +39,6 @@ #include "pipe/p_state.h" #include "pipe/p_video_codec.h" -#include "state_tracker/vdpau_interop.h" - #include "util/u_inlines.h" #include "st_vdpau.h" @@ -51,70 +49,155 @@ #ifdef HAVE_ST_VDPAU +#include "state_tracker/vdpau_interop.h" +#include "state_tracker/vdpau_dmabuf.h" +#include "state_tracker/vdpau_funcs.h" +#include "state_tracker/drm_driver.h" + +static struct pipe_resource * +st_vdpau_video_surface_gallium(struct gl_context *ctx, const GLvoid *vdpSurface, + GLuint index) +{ + int (*getProcAddr)(uint32_t device, uint32_t id, void **ptr); + uint32_t device = (uintptr_t)ctx->vdpDevice; + struct pipe_sampler_view *sv; + VdpVideoSurfaceGallium *f; + + struct pipe_video_buffer *buffer; + struct pipe_sampler_view **samplers; + + getProcAddr = (void *)ctx->vdpGetProcAddress; + if (getProcAddr(device, VDP_FUNC_ID_VIDEO_SURFACE_GALLIUM, (void**)&f)) + return NULL; + + buffer = f((uintptr_t)vdpSurface); + if (!buffer) + return NULL; + + samplers = buffer->get_sampler_view_planes(buffer); + if (!samplers) + return NULL; + + sv = samplers[index >> 1]; + if (!sv) + return NULL; + + return sv->texture; +} + +static struct pipe_resource * +st_vdpau_output_surface_gallium(struct gl_context *ctx, const GLvoid *vdpSurface) +{ + int (*getProcAddr)(uint32_t device, uint32_t id, void **ptr); + uint32_t device = (uintptr_t)ctx->vdpDevice; + VdpOutputSurfaceGallium *f; + + getProcAddr = (void *)ctx->vdpGetProcAddress; + if (getProcAddr(device, VDP_FUNC_ID_OUTPUT_SURFACE_GALLIUM, (void**)&f)) + return NULL; + + return f((uintptr_t)vdpSurface); +} + +static struct pipe_resource * +st_vdpau_resource_from_description(struct gl_context *ctx, + const struct VdpSurfaceDMABufDesc *desc) +{ + struct st_context *st = st_context(ctx); + struct pipe_resource templ, *res; + struct winsys_handle whandle; + + if (desc->handle == -1) + return NULL; + + memset(&templ, 0, sizeof(templ)); + templ.target = PIPE_TEXTURE_2D; + templ.last_level = 0; + templ.depth0 = 1; + templ.array_size = 1; + templ.width0 = desc->width; + templ.height0 = desc->height; + templ.format = VdpFormatRGBAToPipe(desc->format); + templ.bind = PIPE_BIND_SAMPLER_VIEW | PIPE_BIND_RENDER_TARGET; + templ.usage = PIPE_USAGE_DEFAULT; + + memset(&whandle, 0, sizeof(whandle)); + whandle.type = DRM_API_HANDLE_TYPE_FD; + whandle.handle = desc->handle; + whandle.offset = desc->offset; + whandle.stride = desc->stride; + + res = st->pipe->screen->resource_from_handle(st->pipe->screen, &templ, &whandle, + PIPE_HANDLE_USAGE_READ_WRITE); + close(desc->handle); + + return res; +} + +static struct pipe_resource * +st_vdpau_output_surface_dma_buf(struct gl_context *ctx, const GLvoid *vdpSurface) +{ + int (*getProcAddr)(uint32_t device, uint32_t id, void **ptr); + uint32_t device = (uintptr_t)ctx->vdpDevice; + + struct VdpSurfaceDMABufDesc desc; + VdpOutputSurfaceDMABuf *f; + + getProcAddr = (void *)ctx->vdpGetProcAddress; + if (getProcAddr(device, VDP_FUNC_ID_OUTPUT_SURFACE_DMA_BUF, (void**)&f)) + return NULL; + + if (f((uintptr_t)vdpSurface, &desc) != VDP_STATUS_OK) + return NULL; + + return st_vdpau_resource_from_description(ctx, &desc); +} + +static struct pipe_resource * +st_vdpau_video_surface_dma_buf(struct gl_context *ctx, const GLvoid *vdpSurface, + GLuint index) +{ + int (*getProcAddr)(uint32_t device, uint32_t id, void **ptr); + uint32_t device = (uintptr_t)ctx->vdpDevice; + + struct VdpSurfaceDMABufDesc desc; + VdpVideoSurfaceDMABuf *f; + + getProcAddr = (void *)ctx->vdpGetProcAddress; + if (getProcAddr(device, VDP_FUNC_ID_VIDEO_SURFACE_DMA_BUF, (void**)&f)) + return NULL; + + if (f((uintptr_t)vdpSurface, index, &desc) != VDP_STATUS_OK) + return NULL; + + return st_vdpau_resource_from_description(ctx, &desc); +} + static void st_vdpau_map_surface(struct gl_context *ctx, GLenum target, GLenum access, GLboolean output, struct gl_texture_object *texObj, struct gl_texture_image *texImage, const GLvoid *vdpSurface, GLuint index) { - int (*getProcAddr)(uint32_t device, uint32_t id, void **ptr); - uint32_t device = (uintptr_t)ctx->vdpDevice; - struct st_context *st = st_context(ctx); struct st_texture_object *stObj = st_texture_object(texObj); struct st_texture_image *stImage = st_texture_image(texImage); - + struct pipe_resource *res; struct pipe_sampler_view templ, **sampler_view; mesa_format texFormat; - getProcAddr = (void *)ctx->vdpGetProcAddress; if (output) { - VdpOutputSurfaceGallium *f; - - if (getProcAddr(device, VDP_FUNC_ID_OUTPUT_SURFACE_GALLIUM, (void**)&f)) { - _mesa_error(ctx, GL_INVALID_OPERATION, "VDPAUMapSurfacesNV"); - return; - } - - res = f((uintptr_t)vdpSurface); + res = st_vdpau_output_surface_dma_buf(ctx, vdpSurface); - if (!res) { - _mesa_error(ctx, GL_INVALID_OPERATION, "VDPAUMapSurfacesNV"); - return; - } + if (!res) + res = st_vdpau_output_surface_gallium(ctx, vdpSurface); } else { - struct pipe_sampler_view *sv; - VdpVideoSurfaceGallium *f; - - struct pipe_video_buffer *buffer; - struct pipe_sampler_view **samplers; - - if (getProcAddr(device, VDP_FUNC_ID_VIDEO_SURFACE_GALLIUM, (void**)&f)) { - _mesa_error(ctx, GL_INVALID_OPERATION, "VDPAUMapSurfacesNV"); - return; - } - - buffer = f((uintptr_t)vdpSurface); - if (!buffer) { - _mesa_error(ctx, GL_INVALID_OPERATION, "VDPAUMapSurfacesNV"); - return; - } - - samplers = buffer->get_sampler_view_planes(buffer); - if (!samplers) { - _mesa_error(ctx, GL_INVALID_OPERATION, "VDPAUMapSurfacesNV"); - return; - } - - sv = samplers[index >> 1]; - if (!sv) { - _mesa_error(ctx, GL_INVALID_OPERATION, "VDPAUMapSurfacesNV"); - return; - } - - res = sv->texture; + res = st_vdpau_video_surface_dma_buf(ctx, vdpSurface, index); + + if (!res) + res = st_vdpau_video_surface_gallium(ctx, vdpSurface, index); } if (!res) { |