From 59f55a3c598cfd36318551b5933ca656ef7092ce Mon Sep 17 00:00:00 2001 From: Martin Andersson Date: Fri, 12 Apr 2013 21:04:17 +0200 Subject: [PATCH] r600g: HACK: Fix nested loops on Cayman --- src/gallium/drivers/r600/r600_asm.c | 14 ++++++- src/gallium/drivers/r600/r600_shader.c | 69 ++++++++++++++++++++++++++++------ src/gallium/drivers/r600/r600d.h | 5 +++ 3 files changed, 75 insertions(+), 13 deletions(-) diff --git a/src/gallium/drivers/r600/r600_asm.c b/src/gallium/drivers/r600/r600_asm.c index 26a848a..2874adf 100644 --- a/src/gallium/drivers/r600/r600_asm.c +++ b/src/gallium/drivers/r600/r600_asm.c @@ -1985,6 +1985,7 @@ void r600_bytecode_disasm(struct r600_bytecode *bc) LIST_FOR_EACH_ENTRY(alu, &cf->alu, list) { const char *omod_str[] = {"","*2","*4","/2"}; const struct alu_op_info *aop = r600_isa_alu(alu->op); + bool cm_execmask_op = alu->execute_mask && bc->chip_class == CAYMAN; int o = 0; r600_bytecode_alu_nliterals(bc, alu, literal, &nliteral); @@ -1997,8 +1998,10 @@ void r600_bytecode_disasm(struct r600_bytecode *bc) alu->update_pred ? 'P':' ', alu->pred_sel ? alu->pred_sel==2 ? '0':'1':' '); - o += fprintf(stderr, "%s%s%s ", aop->name, - omod_str[alu->omod], alu->dst.clamp ? "_sat":""); + o += fprintf(stderr, "%s ", aop->name); + if (!cm_execmask_op) + o += fprintf(stderr, "%s ", omod_str[alu->omod]); + o += fprintf(stderr, "%s ", alu->dst.clamp ? "_sat":""); o += print_indent(o,60); o += print_dst(alu); @@ -2012,6 +2015,13 @@ void r600_bytecode_disasm(struct r600_bytecode *bc) o += fprintf(stderr, " BS:%d", alu->bank_swizzle); } + if (cm_execmask_op && alu->omod) { + static const char* cm_em_op_names[] = + {"BREAK", "CONTINUE", "KILL"}; + + fprintf(stderr, " %s", cm_em_op_names[alu->omod - 1]); + } + fprintf(stderr, "\n"); id += 2; diff --git a/src/gallium/drivers/r600/r600_shader.c b/src/gallium/drivers/r600/r600_shader.c index f801707..073d2c9 100644 --- a/src/gallium/drivers/r600/r600_shader.c +++ b/src/gallium/drivers/r600/r600_shader.c @@ -58,6 +58,9 @@ issued in the w slot as well. The compiler must issue the source argument to slots z, y, and x */ +static int hack_if; +static int hack_endif; + static int r600_shader_from_tgsi(struct r600_screen *rscreen, struct r600_pipe_shader *pipeshader, struct r600_shader_key key); @@ -5490,7 +5493,7 @@ static int tgsi_opdst(struct r600_shader_ctx *ctx) return 0; } -static int emit_logic_pred(struct r600_shader_ctx *ctx, int opcode) +static int emit_logic_pred(struct r600_shader_ctx *ctx, int opcode, int op_break) { struct r600_bytecode_alu alu; int r; @@ -5499,6 +5502,8 @@ static int emit_logic_pred(struct r600_shader_ctx *ctx, int opcode) alu.op = opcode; alu.execute_mask = 1; alu.update_pred = 1; + if (op_break) + alu.omod = SQ_ALU_EXECUTE_MASK_OP_BREAK; alu.dst.sel = ctx->temp_reg; alu.dst.write = 1; @@ -5510,7 +5515,10 @@ static int emit_logic_pred(struct r600_shader_ctx *ctx, int opcode) alu.last = 1; - r = r600_bytecode_add_alu_type(ctx->bc, &alu, CF_OP_ALU_PUSH_BEFORE); + if (op_break) + r = r600_bytecode_add_alu(ctx->bc, &alu); + else + r = r600_bytecode_add_alu_type(ctx->bc, &alu, CF_OP_ALU_PUSH_BEFORE); if (r) return r; return 0; @@ -5729,10 +5737,16 @@ static void break_loop_on_flag(struct r600_shader_ctx *ctx, unsigned fc_sp) #endif static int tgsi_if(struct r600_shader_ctx *ctx) -{ - emit_logic_pred(ctx, ALU_OP2_PRED_SETNE_INT); - - r600_bytecode_add_cfinst(ctx->bc, CF_OP_JUMP); +{ + if (hack_if) { + hack_if = 0; + emit_logic_pred(ctx, ALU_OP2_PRED_SETE_INT, 1); + r600_bytecode_add_cfinst(ctx->bc, CF_OP_NOP); + } + else { + emit_logic_pred(ctx, ALU_OP2_PRED_SETNE_INT, 0); + r600_bytecode_add_cfinst(ctx->bc, CF_OP_JUMP); + } fc_pushlevel(ctx, FC_IF); @@ -5752,7 +5766,15 @@ static int tgsi_else(struct r600_shader_ctx *ctx) static int tgsi_endif(struct r600_shader_ctx *ctx) { - pops(ctx, 1); + int pop = 0; + if (!hack_endif) { + pops(ctx, 1); + pop = 1; + } + else { + hack_endif = 0; + } + if (ctx->bc->fc_stack[ctx->bc->fc_sp].type != FC_IF) { R600_ERR("if/endif unbalanced in shader\n"); return -1; @@ -5760,7 +5782,7 @@ static int tgsi_endif(struct r600_shader_ctx *ctx) if (ctx->bc->fc_stack[ctx->bc->fc_sp].mid == NULL) { ctx->bc->fc_stack[ctx->bc->fc_sp].start->cf_addr = ctx->bc->cf_last->id + 2; - ctx->bc->fc_stack[ctx->bc->fc_sp].start->pop_count = 1; + ctx->bc->fc_stack[ctx->bc->fc_sp].start->pop_count = pop; } else { ctx->bc->fc_stack[ctx->bc->fc_sp].mid[0]->cf_addr = ctx->bc->cf_last->id + 2; } @@ -5772,6 +5794,9 @@ static int tgsi_endif(struct r600_shader_ctx *ctx) static int tgsi_bgnloop(struct r600_shader_ctx *ctx) { + hack_if = 1; + hack_endif = 1; + /* LOOP_START_DX10 ignores the LOOP_CONFIG* registers, so it is not * limited to 4096 iterations, like the other LOOP_* instructions. */ r600_bytecode_add_cfinst(ctx->bc, CF_OP_LOOP_START_DX10); @@ -5814,7 +5839,7 @@ static int tgsi_endloop(struct r600_shader_ctx *ctx) static int tgsi_loop_brk_cont(struct r600_shader_ctx *ctx) { - unsigned int fscp; +/* unsigned int fscp; for (fscp = ctx->bc->fc_sp; fscp > 0; fscp--) { @@ -5827,9 +5852,31 @@ static int tgsi_loop_brk_cont(struct r600_shader_ctx *ctx) return -EINVAL; } - r600_bytecode_add_cfinst(ctx->bc, ctx->inst_info->op); - fc_set_mid(ctx, fscp); + if (ctx->bc->chip_class == CAYMAN) { + struct r600_bytecode_alu alu = {}; + int r; + + alu.op = ALU_OP2_PRED_SETE; + alu.src[0].sel = V_SQ_ALU_SRC_0; + alu.src[1].sel = V_SQ_ALU_SRC_1; + + if (ctx->inst_info->op == CF_OP_LOOP_BREAK) + alu.omod = SQ_ALU_EXECUTE_MASK_OP_BREAK; + else + alu.omod = SQ_ALU_EXECUTE_MASK_OP_CONTINUE; + + alu.execute_mask = 1; + alu.last = 1; + + r = r600_bytecode_add_alu(ctx->bc, &alu); + if (r) + return r; + } else { + r600_bytecode_add_cfinst(ctx->bc, ctx->inst_info->op); + } + + fc_set_mid(ctx, fscp);*/ return 0; } diff --git a/src/gallium/drivers/r600/r600d.h b/src/gallium/drivers/r600/r600d.h index 9b31383..679dd81 100644 --- a/src/gallium/drivers/r600/r600d.h +++ b/src/gallium/drivers/r600/r600d.h @@ -3698,4 +3698,9 @@ #define DMA_PACKET_CONSTANT_FILL 0xd /* 7xx only */ #define DMA_PACKET_NOP 0xf +#define SQ_ALU_EXECUTE_MASK_OP_DEACTIVATE 0x0 +#define SQ_ALU_EXECUTE_MASK_OP_BREAK 0x1 +#define SQ_ALU_EXECUTE_MASK_OP_CONTINUE 0x2 +#define SQ_ALU_EXECUTE_MASK_OP_KILL 0x3 + #endif -- 1.8.2.1