/* * Copyright © 2011 Intel Corporation * * 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 (including the next * paragraph) 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 "brw_vec4.h" #include "brw_cfg.h" extern "C" { #include "main/macros.h" #include "main/shaderobj.h" #include "program/prog_print.h" #include "program/prog_parameter.h" } #define MAX_INSTRUCTION (1 << 30) using namespace brw; namespace brw { /** * Common helper for constructing swizzles. When only a subset of * channels of a vec4 are used, we don't want to reference the other * channels, as that will tell optimization passes that those other * channels are used. */ unsigned swizzle_for_size(int size) { static const unsigned size_swizzles[4] = { BRW_SWIZZLE4(SWIZZLE_X, SWIZZLE_X, SWIZZLE_X, SWIZZLE_X), BRW_SWIZZLE4(SWIZZLE_X, SWIZZLE_Y, SWIZZLE_Y, SWIZZLE_Y), BRW_SWIZZLE4(SWIZZLE_X, SWIZZLE_Y, SWIZZLE_Z, SWIZZLE_Z), BRW_SWIZZLE4(SWIZZLE_X, SWIZZLE_Y, SWIZZLE_Z, SWIZZLE_W), }; assert((size >= 1) && (size <= 4)); return size_swizzles[size - 1]; } void src_reg::init() { memset(this, 0, sizeof(*this)); this->file = BAD_FILE; } src_reg::src_reg(register_file file, int reg, const glsl_type *type) { init(); this->file = file; this->reg = reg; if (type && (type->is_scalar() || type->is_vector() || type->is_matrix())) this->swizzle = swizzle_for_size(type->vector_elements); else this->swizzle = SWIZZLE_XYZW; } /** Generic unset register constructor. */ src_reg::src_reg() { init(); } src_reg::src_reg(float f) { init(); this->file = IMM; this->type = BRW_REGISTER_TYPE_F; this->imm.f = f; } src_reg::src_reg(uint32_t u) { init(); this->file = IMM; this->type = BRW_REGISTER_TYPE_UD; this->imm.u = u; } src_reg::src_reg(int32_t i) { init(); this->file = IMM; this->type = BRW_REGISTER_TYPE_D; this->imm.i = i; } src_reg::src_reg(dst_reg reg) { init(); this->file = reg.file; this->reg = reg.reg; this->reg_offset = reg.reg_offset; this->type = reg.type; this->reladdr = reg.reladdr; this->fixed_hw_reg = reg.fixed_hw_reg; int swizzles[4]; int next_chan = 0; int last = 0; for (int i = 0; i < 4; i++) { if (!(reg.writemask & (1 << i))) continue; swizzles[next_chan++] = last = i; } for (; next_chan < 4; next_chan++) { swizzles[next_chan] = last; } this->swizzle = BRW_SWIZZLE4(swizzles[0], swizzles[1], swizzles[2], swizzles[3]); } void dst_reg::init() { memset(this, 0, sizeof(*this)); this->file = BAD_FILE; this->writemask = WRITEMASK_XYZW; } dst_reg::dst_reg() { init(); } dst_reg::dst_reg(register_file file, int reg) { init(); this->file = file; this->reg = reg; } dst_reg::dst_reg(register_file file, int reg, const glsl_type *type, int writemask) { init(); this->file = file; this->reg = reg; this->type = brw_type_for_base_type(type); this->writemask = writemask; } dst_reg::dst_reg(struct brw_reg reg) { init(); this->file = HW_REG; this->fixed_hw_reg = reg; } dst_reg::dst_reg(src_reg reg) { init(); this->file = reg.file; this->reg = reg.reg; this->reg_offset = reg.reg_offset; this->type = reg.type; /* How should we do writemasking when converting from a src_reg? It seems * pretty obvious that for src.xxxx the caller wants to write to src.x, but * what about for src.wx? Just special-case src.xxxx for now. */ if (reg.swizzle == BRW_SWIZZLE_XXXX) this->writemask = WRITEMASK_X; else this->writemask = WRITEMASK_XYZW; this->reladdr = reg.reladdr; this->fixed_hw_reg = reg.fixed_hw_reg; } bool vec4_instruction::is_send_from_grf() { switch (opcode) { case SHADER_OPCODE_SHADER_TIME_ADD: case VS_OPCODE_PULL_CONSTANT_LOAD_GEN7: return true; default: return false; } } bool vec4_visitor::can_do_source_mods(vec4_instruction *inst) { if (brw->gen == 6 && inst->is_math()) return false; if (inst->is_send_from_grf()) return false; return true; } /** * Returns how many MRFs an opcode will write over. * * Note that this is not the 0 or 1 implied writes in an actual gen * instruction -- the generate_* functions generate additional MOVs * for setup. */ int vec4_visitor::implied_mrf_writes(vec4_instruction *inst) { if (inst->mlen == 0) return 0; switch (inst->opcode) { case SHADER_OPCODE_RCP: case SHADER_OPCODE_RSQ: case SHADER_OPCODE_SQRT: case SHADER_OPCODE_EXP2: case SHADER_OPCODE_LOG2: case SHADER_OPCODE_SIN: case SHADER_OPCODE_COS: return 1; case SHADER_OPCODE_INT_QUOTIENT: case SHADER_OPCODE_INT_REMAINDER: case SHADER_OPCODE_POW: return 2; case VS_OPCODE_URB_WRITE: return 1; case VS_OPCODE_PULL_CONSTANT_LOAD: return 2; case VS_OPCODE_SCRATCH_READ: return 2; case VS_OPCODE_SCRATCH_WRITE: return 3; case SHADER_OPCODE_SHADER_TIME_ADD: return 0; case SHADER_OPCODE_TEX: case SHADER_OPCODE_TXL: case SHADER_OPCODE_TXD: case SHADER_OPCODE_TXF: case SHADER_OPCODE_TXF_MS: case SHADER_OPCODE_TXS: return inst->header_present ? 1 : 0; default: assert(!"not reached"); return inst->mlen; } } bool src_reg::equals(src_reg *r) { return (file == r->file && reg == r->reg && reg_offset == r->reg_offset && type == r->type && negate == r->negate && abs == r->abs && swizzle == r->swizzle && !reladdr && !r->reladdr && memcmp(&fixed_hw_reg, &r->fixed_hw_reg, sizeof(fixed_hw_reg)) == 0 && imm.u == r->imm.u); } /** * Must be called after calculate_live_intervales() to remove unused * writes to registers -- register allocation will fail otherwise * because something deffed but not used won't be considered to * interfere with other regs. */ bool vec4_visitor::dead_code_eliminate() { bool progress = false; int pc = 0; calculate_live_intervals(); foreach_list_safe(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; if (inst->dst.file == GRF) { assert(this->virtual_grf_end[inst->dst.reg] >= pc); if (this->virtual_grf_end[inst->dst.reg] == pc) { inst->remove(); progress = true; } } pc++; } if (progress) live_intervals_valid = false; return progress; } void vec4_visitor::split_uniform_registers() { /* Prior to this, uniforms have been in an array sized according to * the number of vector uniforms present, sparsely filled (so an * aggregate results in reg indices being skipped over). Now we're * going to cut those aggregates up so each .reg index is one * vector. The goal is to make elimination of unused uniform * components easier later. */ foreach_list(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; for (int i = 0 ; i < 3; i++) { if (inst->src[i].file != UNIFORM) continue; assert(!inst->src[i].reladdr); inst->src[i].reg += inst->src[i].reg_offset; inst->src[i].reg_offset = 0; } } /* Update that everything is now vector-sized. */ for (int i = 0; i < this->uniforms; i++) { this->uniform_size[i] = 1; } } void vec4_visitor::pack_uniform_registers() { bool uniform_used[this->uniforms]; int new_loc[this->uniforms]; int new_chan[this->uniforms]; memset(uniform_used, 0, sizeof(uniform_used)); memset(new_loc, 0, sizeof(new_loc)); memset(new_chan, 0, sizeof(new_chan)); /* Find which uniform vectors are actually used by the program. We * expect unused vector elements when we've moved array access out * to pull constants, and from some GLSL code generators like wine. */ foreach_list(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; for (int i = 0 ; i < 3; i++) { if (inst->src[i].file != UNIFORM) continue; uniform_used[inst->src[i].reg] = true; } } int new_uniform_count = 0; /* Now, figure out a packing of the live uniform vectors into our * push constants. */ for (int src = 0; src < uniforms; src++) { int size = this->uniform_vector_size[src]; if (!uniform_used[src]) { this->uniform_vector_size[src] = 0; continue; } int dst; /* Find the lowest place we can slot this uniform in. */ for (dst = 0; dst < src; dst++) { if (this->uniform_vector_size[dst] + size <= 4) break; } if (src == dst) { new_loc[src] = dst; new_chan[src] = 0; } else { new_loc[src] = dst; new_chan[src] = this->uniform_vector_size[dst]; /* Move the references to the data */ for (int j = 0; j < size; j++) { prog_data->param[dst * 4 + new_chan[src] + j] = prog_data->param[src * 4 + j]; } this->uniform_vector_size[dst] += size; this->uniform_vector_size[src] = 0; } new_uniform_count = MAX2(new_uniform_count, dst + 1); } this->uniforms = new_uniform_count; /* Now, update the instructions for our repacked uniforms. */ foreach_list(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; for (int i = 0 ; i < 3; i++) { int src = inst->src[i].reg; if (inst->src[i].file != UNIFORM) continue; inst->src[i].reg = new_loc[src]; int sx = BRW_GET_SWZ(inst->src[i].swizzle, 0) + new_chan[src]; int sy = BRW_GET_SWZ(inst->src[i].swizzle, 1) + new_chan[src]; int sz = BRW_GET_SWZ(inst->src[i].swizzle, 2) + new_chan[src]; int sw = BRW_GET_SWZ(inst->src[i].swizzle, 3) + new_chan[src]; inst->src[i].swizzle = BRW_SWIZZLE4(sx, sy, sz, sw); } } } bool src_reg::is_zero() const { if (file != IMM) return false; if (type == BRW_REGISTER_TYPE_F) { return imm.f == 0.0; } else { return imm.i == 0; } } bool src_reg::is_one() const { if (file != IMM) return false; if (type == BRW_REGISTER_TYPE_F) { return imm.f == 1.0; } else { return imm.i == 1; } } /** * Does algebraic optimizations (0 * a = 0, 1 * a = a, a + 0 = a). * * While GLSL IR also performs this optimization, we end up with it in * our instruction stream for a couple of reasons. One is that we * sometimes generate silly instructions, for example in array access * where we'll generate "ADD offset, index, base" even if base is 0. * The other is that GLSL IR's constant propagation doesn't track the * components of aggregates, so some VS patterns (initialize matrix to * 0, accumulate in vertex blending factors) end up breaking down to * instructions involving 0. */ bool vec4_visitor::opt_algebraic() { bool progress = false; foreach_list(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; switch (inst->opcode) { case BRW_OPCODE_ADD: if (inst->src[1].is_zero()) { inst->opcode = BRW_OPCODE_MOV; inst->src[1] = src_reg(); progress = true; } break; case BRW_OPCODE_MUL: if (inst->src[1].is_zero()) { inst->opcode = BRW_OPCODE_MOV; switch (inst->src[0].type) { case BRW_REGISTER_TYPE_F: inst->src[0] = src_reg(0.0f); break; case BRW_REGISTER_TYPE_D: inst->src[0] = src_reg(0); break; case BRW_REGISTER_TYPE_UD: inst->src[0] = src_reg(0u); break; default: assert(!"not reached"); inst->src[0] = src_reg(0.0f); break; } inst->src[1] = src_reg(); progress = true; } else if (inst->src[1].is_one()) { inst->opcode = BRW_OPCODE_MOV; inst->src[1] = src_reg(); progress = true; } break; default: break; } } if (progress) this->live_intervals_valid = false; return progress; } /** * Only a limited number of hardware registers may be used for push * constants, so this turns access to the overflowed constants into * pull constants. */ void vec4_visitor::move_push_constants_to_pull_constants() { int pull_constant_loc[this->uniforms]; /* Only allow 32 registers (256 uniform components) as push constants, * which is the limit on gen6. */ int max_uniform_components = 32 * 8; if (this->uniforms * 4 <= max_uniform_components) return; /* Make some sort of choice as to which uniforms get sent to pull * constants. We could potentially do something clever here like * look for the most infrequently used uniform vec4s, but leave * that for later. */ for (int i = 0; i < this->uniforms * 4; i += 4) { pull_constant_loc[i / 4] = -1; if (i >= max_uniform_components) { const float **values = &prog_data->param[i]; /* Try to find an existing copy of this uniform in the pull * constants if it was part of an array access already. */ for (unsigned int j = 0; j < prog_data->nr_pull_params; j += 4) { int matches; for (matches = 0; matches < 4; matches++) { if (prog_data->pull_param[j + matches] != values[matches]) break; } if (matches == 4) { pull_constant_loc[i / 4] = j / 4; break; } } if (pull_constant_loc[i / 4] == -1) { assert(prog_data->nr_pull_params % 4 == 0); pull_constant_loc[i / 4] = prog_data->nr_pull_params / 4; for (int j = 0; j < 4; j++) { prog_data->pull_param[prog_data->nr_pull_params++] = values[j]; } } } } /* Now actually rewrite usage of the things we've moved to pull * constants. */ foreach_list_safe(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; for (int i = 0 ; i < 3; i++) { if (inst->src[i].file != UNIFORM || pull_constant_loc[inst->src[i].reg] == -1) continue; int uniform = inst->src[i].reg; dst_reg temp = dst_reg(this, glsl_type::vec4_type); emit_pull_constant_load(inst, temp, inst->src[i], pull_constant_loc[uniform]); inst->src[i].file = temp.file; inst->src[i].reg = temp.reg; inst->src[i].reg_offset = temp.reg_offset; inst->src[i].reladdr = NULL; } } /* Repack push constants to remove the now-unused ones. */ pack_uniform_registers(); } /** * Sets the dependency control fields on instructions after register * allocation and before the generator is run. * * When you have a sequence of instructions like: * * DP4 temp.x vertex uniform[0] * DP4 temp.y vertex uniform[0] * DP4 temp.z vertex uniform[0] * DP4 temp.w vertex uniform[0] * * The hardware doesn't know that it can actually run the later instructions * while the previous ones are in flight, producing stalls. However, we have * manual fields we can set in the instructions that let it do so. */ void vec4_visitor::opt_set_dependency_control() { vec4_instruction *last_grf_write[BRW_MAX_GRF]; uint8_t grf_channels_written[BRW_MAX_GRF]; vec4_instruction *last_mrf_write[BRW_MAX_GRF]; uint8_t mrf_channels_written[BRW_MAX_GRF]; cfg_t cfg(this); assert(prog_data->total_grf || !"Must be called after register allocation"); for (int i = 0; i < cfg.num_blocks; i++) { bblock_t *bblock = cfg.blocks[i]; vec4_instruction *inst; memset(last_grf_write, 0, sizeof(last_grf_write)); memset(last_mrf_write, 0, sizeof(last_mrf_write)); for (inst = (vec4_instruction *)bblock->start; inst != (vec4_instruction *)bblock->end->next; inst = (vec4_instruction *)inst->next) { /* If we read from a register that we were doing dependency control * on, don't do dependency control across the read. */ for (int i = 0; i < 3; i++) { int reg = inst->src[i].reg + inst->src[i].reg_offset; if (inst->src[i].file == GRF) { last_grf_write[reg] = NULL; } else if (inst->src[i].file == HW_REG) { memset(last_grf_write, 0, sizeof(last_grf_write)); break; } assert(inst->src[i].file != MRF); } /* In the presence of send messages, totally interrupt dependency * control. They're long enough that the chance of dependency * control around them just doesn't matter. */ if (inst->mlen) { memset(last_grf_write, 0, sizeof(last_grf_write)); memset(last_mrf_write, 0, sizeof(last_mrf_write)); continue; } /* It looks like setting dependency control on a predicated * instruction hangs the GPU. */ if (inst->predicate) { memset(last_grf_write, 0, sizeof(last_grf_write)); memset(last_mrf_write, 0, sizeof(last_mrf_write)); continue; } /* Now, see if we can do dependency control for this instruction * against a previous one writing to its destination. */ int reg = inst->dst.reg + inst->dst.reg_offset; if (inst->dst.file == GRF) { if (last_grf_write[reg] && !(inst->dst.writemask & grf_channels_written[reg])) { last_grf_write[reg]->no_dd_clear = true; inst->no_dd_check = true; } else { grf_channels_written[reg] = 0; } last_grf_write[reg] = inst; grf_channels_written[reg] |= inst->dst.writemask; } else if (inst->dst.file == MRF) { if (last_mrf_write[reg] && !(inst->dst.writemask & mrf_channels_written[reg])) { last_mrf_write[reg]->no_dd_clear = true; inst->no_dd_check = true; } else { mrf_channels_written[reg] = 0; } last_mrf_write[reg] = inst; mrf_channels_written[reg] |= inst->dst.writemask; } else if (inst->dst.reg == HW_REG) { if (inst->dst.fixed_hw_reg.file == BRW_GENERAL_REGISTER_FILE) memset(last_grf_write, 0, sizeof(last_grf_write)); if (inst->dst.fixed_hw_reg.file == BRW_MESSAGE_REGISTER_FILE) memset(last_mrf_write, 0, sizeof(last_mrf_write)); } } } } bool vec4_instruction::can_reswizzle_dst(int dst_writemask, int swizzle, int swizzle_mask) { /* If this instruction sets anything not referenced by swizzle, then we'd * totally break it when we reswizzle. */ if (dst.writemask & ~swizzle_mask) return false; switch (opcode) { case BRW_OPCODE_DP4: case BRW_OPCODE_DP3: case BRW_OPCODE_DP2: return true; default: /* Check if there happens to be no reswizzling required. */ for (int c = 0; c < 4; c++) { int bit = 1 << BRW_GET_SWZ(swizzle, c); /* Skip components of the swizzle not used by the dst. */ if (!(dst_writemask & (1 << c))) continue; /* We don't do the reswizzling yet, so just sanity check that we * don't have to. */ if (bit != (1 << c)) return false; } return true; } } /** * For any channels in the swizzle's source that were populated by this * instruction, rewrite the instruction to put the appropriate result directly * in those channels. * * e.g. for swizzle=yywx, MUL a.xy b c -> MUL a.yy_x b.yy z.yy_x */ void vec4_instruction::reswizzle_dst(int dst_writemask, int swizzle) { int new_writemask = 0; switch (opcode) { case BRW_OPCODE_DP4: case BRW_OPCODE_DP3: case BRW_OPCODE_DP2: for (int c = 0; c < 4; c++) { int bit = 1 << BRW_GET_SWZ(swizzle, c); /* Skip components of the swizzle not used by the dst. */ if (!(dst_writemask & (1 << c))) continue; /* If we were populating this component, then populate the * corresponding channel of the new dst. */ if (dst.writemask & bit) new_writemask |= (1 << c); } dst.writemask = new_writemask; break; default: for (int c = 0; c < 4; c++) { /* Skip components of the swizzle not used by the dst. */ if (!(dst_writemask & (1 << c))) continue; /* We don't do the reswizzling yet, so just sanity check that we * don't have to. */ assert((1 << BRW_GET_SWZ(swizzle, c)) == (1 << c)); } break; } } /* * Tries to reduce extra MOV instructions by taking temporary GRFs that get * just written and then MOVed into another reg and making the original write * of the GRF write directly to the final destination instead. */ bool vec4_visitor::opt_register_coalesce() { bool progress = false; int next_ip = 0; calculate_live_intervals(); foreach_list_safe(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; int ip = next_ip; next_ip++; if (inst->opcode != BRW_OPCODE_MOV || (inst->dst.file != GRF && inst->dst.file != MRF) || inst->predicate || inst->src[0].file != GRF || inst->dst.type != inst->src[0].type || inst->src[0].abs || inst->src[0].negate || inst->src[0].reladdr) continue; bool to_mrf = (inst->dst.file == MRF); /* Can't coalesce this GRF if someone else was going to * read it later. */ if (this->virtual_grf_end[inst->src[0].reg] > ip) continue; /* We need to check interference with the final destination between this * instruction and the earliest instruction involved in writing the GRF * we're eliminating. To do that, keep track of which of our source * channels we've seen initialized. */ bool chans_needed[4] = {false, false, false, false}; int chans_remaining = 0; int swizzle_mask = 0; for (int i = 0; i < 4; i++) { int chan = BRW_GET_SWZ(inst->src[0].swizzle, i); if (!(inst->dst.writemask & (1 << i))) continue; swizzle_mask |= (1 << chan); if (!chans_needed[chan]) { chans_needed[chan] = true; chans_remaining++; } } /* Now walk up the instruction stream trying to see if we can rewrite * everything writing to the temporary to write into the destination * instead. */ vec4_instruction *scan_inst; for (scan_inst = (vec4_instruction *)inst->prev; scan_inst->prev != NULL; scan_inst = (vec4_instruction *)scan_inst->prev) { if (scan_inst->dst.file == GRF && scan_inst->dst.reg == inst->src[0].reg && scan_inst->dst.reg_offset == inst->src[0].reg_offset) { /* Found something writing to the reg we want to coalesce away. */ if (to_mrf) { /* SEND instructions can't have MRF as a destination. */ if (scan_inst->mlen) break; if (brw->gen == 6) { /* gen6 math instructions must have the destination be * GRF, so no compute-to-MRF for them. */ if (scan_inst->is_math()) { break; } } } /* If we can't handle the swizzle, bail. */ if (!scan_inst->can_reswizzle_dst(inst->dst.writemask, inst->src[0].swizzle, swizzle_mask)) { break; } /* Mark which channels we found unconditional writes for. */ if (!scan_inst->predicate) { for (int i = 0; i < 4; i++) { if (scan_inst->dst.writemask & (1 << i) && chans_needed[i]) { chans_needed[i] = false; chans_remaining--; } } } if (chans_remaining == 0) break; } /* We don't handle flow control here. Most computation of values * that could be coalesced happens just before their use. */ if (scan_inst->opcode == BRW_OPCODE_DO || scan_inst->opcode == BRW_OPCODE_WHILE || scan_inst->opcode == BRW_OPCODE_ELSE || scan_inst->opcode == BRW_OPCODE_ENDIF) { break; } /* You can't read from an MRF, so if someone else reads our MRF's * source GRF that we wanted to rewrite, that stops us. If it's a * GRF we're trying to coalesce to, we don't actually handle * rewriting sources so bail in that case as well. */ bool interfered = false; for (int i = 0; i < 3; i++) { if (scan_inst->src[i].file == GRF && scan_inst->src[i].reg == inst->src[0].reg && scan_inst->src[i].reg_offset == inst->src[0].reg_offset) { interfered = true; } } if (interfered) break; /* If somebody else writes our destination here, we can't coalesce * before that. */ if (scan_inst->dst.file == inst->dst.file && scan_inst->dst.reg == inst->dst.reg) { break; } /* Check for reads of the register we're trying to coalesce into. We * can't go rewriting instructions above that to put some other value * in the register instead. */ if (to_mrf && scan_inst->mlen > 0) { if (inst->dst.reg >= scan_inst->base_mrf && inst->dst.reg < scan_inst->base_mrf + scan_inst->mlen) { break; } } else { for (int i = 0; i < 3; i++) { if (scan_inst->src[i].file == inst->dst.file && scan_inst->src[i].reg == inst->dst.reg && scan_inst->src[i].reg_offset == inst->src[0].reg_offset) { interfered = true; } } if (interfered) break; } } if (chans_remaining == 0) { /* If we've made it here, we have an MOV we want to coalesce out, and * a scan_inst pointing to the earliest instruction involved in * computing the value. Now go rewrite the instruction stream * between the two. */ while (scan_inst != inst) { if (scan_inst->dst.file == GRF && scan_inst->dst.reg == inst->src[0].reg && scan_inst->dst.reg_offset == inst->src[0].reg_offset) { scan_inst->reswizzle_dst(inst->dst.writemask, inst->src[0].swizzle); scan_inst->dst.file = inst->dst.file; scan_inst->dst.reg = inst->dst.reg; scan_inst->dst.reg_offset = inst->dst.reg_offset; scan_inst->saturate |= inst->saturate; } scan_inst = (vec4_instruction *)scan_inst->next; } inst->remove(); progress = true; } } if (progress) live_intervals_valid = false; return progress; } /** * Splits virtual GRFs requesting more than one contiguous physical register. * * We initially create large virtual GRFs for temporary structures, arrays, * and matrices, so that the dereference visitor functions can add reg_offsets * to work their way down to the actual member being accessed. But when it * comes to optimization, we'd like to treat each register as individual * storage if possible. * * So far, the only thing that might prevent splitting is a send message from * a GRF on IVB. */ void vec4_visitor::split_virtual_grfs() { int num_vars = this->virtual_grf_count; int new_virtual_grf[num_vars]; bool split_grf[num_vars]; memset(new_virtual_grf, 0, sizeof(new_virtual_grf)); /* Try to split anything > 0 sized. */ for (int i = 0; i < num_vars; i++) { split_grf[i] = this->virtual_grf_sizes[i] != 1; } /* Check that the instructions are compatible with the registers we're trying * to split. */ foreach_list(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; /* If there's a SEND message loading from a GRF on gen7+, it needs to be * contiguous. Assume that the GRF for the SEND is always in src[0]. */ if (inst->is_send_from_grf()) { split_grf[inst->src[0].reg] = false; } } /* Allocate new space for split regs. Note that the virtual * numbers will be contiguous. */ for (int i = 0; i < num_vars; i++) { if (!split_grf[i]) continue; new_virtual_grf[i] = virtual_grf_alloc(1); for (int j = 2; j < this->virtual_grf_sizes[i]; j++) { int reg = virtual_grf_alloc(1); assert(reg == new_virtual_grf[i] + j - 1); (void) reg; } this->virtual_grf_sizes[i] = 1; } foreach_list(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; if (inst->dst.file == GRF && split_grf[inst->dst.reg] && inst->dst.reg_offset != 0) { inst->dst.reg = (new_virtual_grf[inst->dst.reg] + inst->dst.reg_offset - 1); inst->dst.reg_offset = 0; } for (int i = 0; i < 3; i++) { if (inst->src[i].file == GRF && split_grf[inst->src[i].reg] && inst->src[i].reg_offset != 0) { inst->src[i].reg = (new_virtual_grf[inst->src[i].reg] + inst->src[i].reg_offset - 1); inst->src[i].reg_offset = 0; } } } this->live_intervals_valid = false; } void vec4_visitor::dump_instruction(backend_instruction *be_inst) { vec4_instruction *inst = (vec4_instruction *)be_inst; printf("%s ", brw_instruction_name(inst->opcode)); switch (inst->dst.file) { case GRF: printf("vgrf%d.%d", inst->dst.reg, inst->dst.reg_offset); break; case MRF: printf("m%d", inst->dst.reg); break; case BAD_FILE: printf("(null)"); break; default: printf("???"); break; } if (inst->dst.writemask != WRITEMASK_XYZW) { printf("."); if (inst->dst.writemask & 1) printf("x"); if (inst->dst.writemask & 2) printf("y"); if (inst->dst.writemask & 4) printf("z"); if (inst->dst.writemask & 8) printf("w"); } printf(", "); for (int i = 0; i < 3; i++) { switch (inst->src[i].file) { case GRF: printf("vgrf%d", inst->src[i].reg); break; case ATTR: printf("attr%d", inst->src[i].reg); break; case UNIFORM: printf("u%d", inst->src[i].reg); break; case IMM: switch (inst->src[i].type) { case BRW_REGISTER_TYPE_F: printf("%fF", inst->src[i].imm.f); break; case BRW_REGISTER_TYPE_D: printf("%dD", inst->src[i].imm.i); break; case BRW_REGISTER_TYPE_UD: printf("%uU", inst->src[i].imm.u); break; default: printf("???"); break; } break; case BAD_FILE: printf("(null)"); break; default: printf("???"); break; } if (inst->src[i].reg_offset) printf(".%d", inst->src[i].reg_offset); static const char *chans[4] = {"x", "y", "z", "w"}; printf("."); for (int c = 0; c < 4; c++) { printf("%s", chans[BRW_GET_SWZ(inst->src[i].swizzle, c)]); } if (i < 3) printf(", "); } printf("\n"); } /** * Replace each register of type ATTR in this->instructions with a reference * to a fixed HW register. */ void vec4_visitor::lower_attributes_to_hw_regs(const int *attribute_map) { foreach_list(node, &this->instructions) { vec4_instruction *inst = (vec4_instruction *)node; /* We have to support ATTR as a destination for GL_FIXED fixup. */ if (inst->dst.file == ATTR) { int grf = attribute_map[inst->dst.reg + inst->dst.reg_offset]; /* All attributes used in the shader need to have been assigned a * hardware register by the caller */ assert(grf != 0); struct brw_reg reg = brw_vec8_grf(grf, 0); reg.type = inst->dst.type; reg.dw1.bits.writemask = inst->dst.writemask; inst->dst.file = HW_REG; inst->dst.fixed_hw_reg = reg; } for (int i = 0; i < 3; i++) { if (inst->src[i].file != ATTR) continue; int grf = attribute_map[inst->src[i].reg + inst->src[i].reg_offset]; /* All attributes used in the shader need to have been assigned a * hardware register by the caller */ assert(grf != 0); struct brw_reg reg = brw_vec8_grf(grf, 0); reg.dw1.bits.swizzle = inst->src[i].swizzle; reg.type = inst->src[i].type; if (inst->src[i].abs) reg = brw_abs(reg); if (inst->src[i].negate) reg = negate(reg); inst->src[i].file = HW_REG; inst->src[i].fixed_hw_reg = reg; } } } int vec4_vs_visitor::setup_attributes(int payload_reg) { int nr_attributes; int attribute_map[VERT_ATTRIB_MAX + 1]; memset(attribute_map, 0, sizeof(attribute_map)); nr_attributes = 0; for (int i = 0; i < VERT_ATTRIB_MAX; i++) { if (vs_prog_data->inputs_read & BITFIELD64_BIT(i)) { attribute_map[i] = payload_reg + nr_attributes; nr_attributes++; } } /* VertexID is stored by the VF as the last vertex element, but we * don't represent it with a flag in inputs_read, so we call it * VERT_ATTRIB_MAX. */ if (vs_prog_data->uses_vertexid) { attribute_map[VERT_ATTRIB_MAX] = payload_reg + nr_attributes; nr_attributes++; } lower_attributes_to_hw_regs(attribute_map); /* The BSpec says we always have to read at least one thing from * the VF, and it appears that the hardware wedges otherwise. */ if (nr_attributes == 0) nr_attributes = 1; prog_data->urb_read_length = (nr_attributes + 1) / 2; unsigned vue_entries = MAX2(nr_attributes, prog_data->vue_map.num_slots); if (brw->gen == 6) prog_data->urb_entry_size = ALIGN(vue_entries, 8) / 8; else prog_data->urb_entry_size = ALIGN(vue_entries, 4) / 4; return payload_reg + nr_attributes; } int vec4_visitor::setup_uniforms(int reg) { /* The pre-gen6 VS requires that some push constants get loaded no * matter what, or the GPU would hang. */ if (brw->gen < 6 && this->uniforms == 0) { this->uniform_vector_size[this->uniforms] = 1; for (unsigned int i = 0; i < 4; i++) { unsigned int slot = this->uniforms * 4 + i; static float zero = 0.0; prog_data->param[slot] = &zero; } this->uniforms++; reg++; } else { reg += ALIGN(uniforms, 2) / 2; } prog_data->nr_params = this->uniforms * 4; prog_data->curb_read_length = reg - 1; return reg; } void vec4_visitor::setup_payload(void) { int reg = 0; /* The payload always contains important data in g0, which contains * the URB handles that are passed on to the URB write at the end * of the thread. So, we always start push constants at g1. */ reg++; reg = setup_uniforms(reg); reg = setup_attributes(reg); this->first_non_payload_grf = reg; } src_reg vec4_visitor::get_timestamp() { assert(brw->gen >= 7); src_reg ts = src_reg(brw_reg(BRW_ARCHITECTURE_REGISTER_FILE, BRW_ARF_TIMESTAMP, 0, BRW_REGISTER_TYPE_UD, BRW_VERTICAL_STRIDE_0, BRW_WIDTH_4, BRW_HORIZONTAL_STRIDE_4, BRW_SWIZZLE_XYZW, WRITEMASK_XYZW)); dst_reg dst = dst_reg(this, glsl_type::uvec4_type); vec4_instruction *mov = emit(MOV(dst, ts)); /* We want to read the 3 fields we care about (mostly field 0, but also 2) * even if it's not enabled in the dispatch. */ mov->force_writemask_all = true; return src_reg(dst); } void vec4_visitor::emit_shader_time_begin() { current_annotation = "shader time start"; shader_start_time = get_timestamp(); } void vec4_visitor::emit_shader_time_end() { current_annotation = "shader time end"; src_reg shader_end_time = get_timestamp(); /* Check that there weren't any timestamp reset events (assuming these * were the only two timestamp reads that happened). */ src_reg reset_end = shader_end_time; reset_end.swizzle = BRW_SWIZZLE_ZZZZ; vec4_instruction *test = emit(AND(dst_null_d(), reset_end, src_reg(1u))); test->conditional_mod = BRW_CONDITIONAL_Z; emit(IF(BRW_PREDICATE_NORMAL)); /* Take the current timestamp and get the delta. */ shader_start_time.negate = true; dst_reg diff = dst_reg(this, glsl_type::uint_type); emit(ADD(diff, shader_start_time, shader_end_time)); /* If there were no instructions between the two timestamp gets, the diff * is 2 cycles. Remove that overhead, so I can forget about that when * trying to determine the time taken for single instructions. */ emit(ADD(diff, src_reg(diff), src_reg(-2u))); emit_shader_time_write(ST_VS, src_reg(diff)); emit_shader_time_write(ST_VS_WRITTEN, src_reg(1u)); emit(BRW_OPCODE_ELSE); emit_shader_time_write(ST_VS_RESET, src_reg(1u)); emit(BRW_OPCODE_ENDIF); } void vec4_visitor::emit_shader_time_write(enum shader_time_shader_type type, src_reg value) { int shader_time_index = brw_get_shader_time_index(brw, shader_prog, prog, type); dst_reg dst = dst_reg(this, glsl_type::get_array_instance(glsl_type::vec4_type, 2)); dst_reg offset = dst; dst_reg time = dst; time.reg_offset++; offset.type = BRW_REGISTER_TYPE_UD; emit(MOV(offset, src_reg(shader_time_index * SHADER_TIME_STRIDE))); time.type = BRW_REGISTER_TYPE_UD; emit(MOV(time, src_reg(value))); emit(SHADER_OPCODE_SHADER_TIME_ADD, dst_reg(), src_reg(dst)); } bool vec4_visitor::run() { sanity_param_count = prog->Parameters->NumParameters; if (INTEL_DEBUG & DEBUG_SHADER_TIME) emit_shader_time_begin(); emit_prolog(); /* Generate VS IR for main(). (the visitor only descends into * functions called "main"). */ if (shader) { visit_instructions(shader->ir); } else { emit_program_code(); } base_ir = NULL; if (key->userclip_active && !key->uses_clip_distance) setup_uniform_clipplane_values(); emit_thread_end(); /* Before any optimization, push array accesses out to scratch * space where we need them to be. This pass may allocate new * virtual GRFs, so we want to do it early. It also makes sure * that we have reladdr computations available for CSE, since we'll * often do repeated subexpressions for those. */ if (shader) { move_grf_array_access_to_scratch(); move_uniform_array_access_to_pull_constants(); } else { /* The ARB_vertex_program frontend emits pull constant loads directly * rather than using reladdr, so we don't need to walk through all the * instructions looking for things to move. There isn't anything. * * We do still need to split things to vec4 size. */ split_uniform_registers(); } pack_uniform_registers(); move_push_constants_to_pull_constants(); split_virtual_grfs(); bool progress; do { progress = false; progress = dead_code_eliminate() || progress; progress = opt_copy_propagation() || progress; progress = opt_algebraic() || progress; progress = opt_register_coalesce() || progress; } while (progress); if (failed) return false; setup_payload(); if (false) { /* Debug of register spilling: Go spill everything. */ const int grf_count = virtual_grf_count; float spill_costs[virtual_grf_count]; bool no_spill[virtual_grf_count]; evaluate_spill_costs(spill_costs, no_spill); for (int i = 0; i < grf_count; i++) { if (no_spill[i]) continue; spill_reg(i); } } while (!reg_allocate()) { if (failed) break; } opt_schedule_instructions(); opt_set_dependency_control(); /* If any state parameters were appended, then ParameterValues could have * been realloced, in which case the driver uniform storage set up by * _mesa_associate_uniform_storage() would point to freed memory. Make * sure that didn't happen. */ assert(sanity_param_count == prog->Parameters->NumParameters); return !failed; } } /* namespace brw */ extern "C" { /** * Compile a vertex shader. * * Returns the final assembly and the program's size. */ const unsigned * brw_vs_emit(struct brw_context *brw, struct gl_shader_program *prog, struct brw_vs_compile *c, struct brw_vs_prog_data *prog_data, void *mem_ctx, unsigned *final_assembly_size) { bool start_busy = false; float start_time = 0; if (unlikely(brw->perf_debug)) { start_busy = (brw->batch.last_bo && drm_intel_bo_busy(brw->batch.last_bo)); start_time = get_time(); } struct brw_shader *shader = NULL; if (prog) shader = (brw_shader *) prog->_LinkedShaders[MESA_SHADER_VERTEX]; if (unlikely(INTEL_DEBUG & DEBUG_VS)) { if (prog) { printf("GLSL IR for native vertex shader %d:\n", prog->Name); _mesa_print_ir(shader->ir, NULL); printf("\n\n"); } else { printf("ARB_vertex_program %d for native vertex shader\n", c->vp->program.Base.Id); _mesa_print_program(&c->vp->program.Base); } } vec4_vs_visitor v(brw, c, prog_data, prog, shader, mem_ctx); if (!v.run()) { if (prog) { prog->LinkStatus = false; ralloc_strcat(&prog->InfoLog, v.fail_msg); } _mesa_problem(NULL, "Failed to compile vertex shader: %s\n", v.fail_msg); return NULL; } vec4_generator g(brw, prog, &c->vp->program.Base, mem_ctx, INTEL_DEBUG & DEBUG_VS); const unsigned *generated =g.generate_assembly(&v.instructions, final_assembly_size); if (unlikely(brw->perf_debug) && shader) { if (shader->compiled_once) { brw_vs_debug_recompile(brw, prog, &c->key); } if (start_busy && !drm_intel_bo_busy(brw->batch.last_bo)) { perf_debug("VS compile took %.03f ms and stalled the GPU\n", (get_time() - start_time) * 1000); } shader->compiled_once = true; } return generated; } } /* extern "C" */