ShuriZma
/
xemu
mirror of https://github.com/xemu-project/xemu.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

tcg/optimize: Add type to OptContext 

Compute the type of the operation early.

There are at least 4 places that used a def->flags ladder
to determine the type of the operation being optimized.

There were two places that assumed !TCG_OPF_64BIT means
TCG_TYPE_I32, and so could potentially compute incorrect
results for vector operations.

Reviewed-by: Luis Pires <luis.pires@eldorado.org.br>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This commit is contained in:
Richard Henderson 2021-08-25 08:00:20 -07:00
parent e8679955ec
commit 67f84c9621
1 changed files with 89 additions and 60 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

149
tcg/optimize.c
View File

@ -51,6 +51,7 @@ typedef struct OptContext {
/* In flight values from optimization. */ /* In flight values from optimization. */
uint64_t z_mask; uint64_t z_mask;
TCGType type;
} OptContext; } OptContext;
static inline TempOptInfo *ts_info(TCGTemp *ts) static inline TempOptInfo *ts_info(TCGTemp *ts)
@ -187,7 +188,6 @@ static bool tcg_opt_gen_mov(OptContext *ctx, TCGOp *op, TCGArg dst, TCGArg src)
{ {
TCGTemp *dst_ts = arg_temp(dst); TCGTemp *dst_ts = arg_temp(dst);
TCGTemp *src_ts = arg_temp(src); TCGTemp *src_ts = arg_temp(src);
const TCGOpDef *def;
TempOptInfo *di; TempOptInfo *di;
TempOptInfo *si; TempOptInfo *si;
uint64_t z_mask; uint64_t z_mask;
@ -201,16 +201,24 @@ static bool tcg_opt_gen_mov(OptContext *ctx, TCGOp *op, TCGArg dst, TCGArg src)
reset_ts(dst_ts); reset_ts(dst_ts);
di = ts_info(dst_ts); di = ts_info(dst_ts);
si = ts_info(src_ts); si = ts_info(src_ts);
def = &tcg_op_defs[op->opc];
if (def->flags & TCG_OPF_VECTOR) { switch (ctx->type) {
new_op = INDEX_op_mov_vec; case TCG_TYPE_I32:
} else if (def->flags & TCG_OPF_64BIT) {
new_op = INDEX_op_mov_i64;
} else {
new_op = INDEX_op_mov_i32; new_op = INDEX_op_mov_i32;
break;
case TCG_TYPE_I64:
new_op = INDEX_op_mov_i64;
break;
case TCG_TYPE_V64:
case TCG_TYPE_V128:
case TCG_TYPE_V256:
/* TCGOP_VECL and TCGOP_VECE remain unchanged. */
new_op = INDEX_op_mov_vec;
break;
default:
g_assert_not_reached();
} }
op->opc = new_op; op->opc = new_op;
/* TCGOP_VECL and TCGOP_VECE remain unchanged. */
op->args[0] = dst; op->args[0] = dst;
op->args[1] = src; op->args[1] = src;
@ -237,20 +245,9 @@ static bool tcg_opt_gen_mov(OptContext *ctx, TCGOp *op, TCGArg dst, TCGArg src)
static bool tcg_opt_gen_movi(OptContext *ctx, TCGOp *op, static bool tcg_opt_gen_movi(OptContext *ctx, TCGOp *op,
TCGArg dst, uint64_t val) TCGArg dst, uint64_t val)
{ {
const TCGOpDef *def = &tcg_op_defs[op->opc];
TCGType type;
TCGTemp *tv;
if (def->flags & TCG_OPF_VECTOR) {
type = TCGOP_VECL(op) + TCG_TYPE_V64;
} else if (def->flags & TCG_OPF_64BIT) {
type = TCG_TYPE_I64;
} else {
type = TCG_TYPE_I32;
}
/* Convert movi to mov with constant temp. */ /* Convert movi to mov with constant temp. */
tv = tcg_constant_internal(type, val); TCGTemp *tv = tcg_constant_internal(ctx->type, val);
init_ts_info(ctx, tv); init_ts_info(ctx, tv);
return tcg_opt_gen_mov(ctx, op, dst, temp_arg(tv)); return tcg_opt_gen_mov(ctx, op, dst, temp_arg(tv));
} }
@ -420,11 +417,11 @@ static uint64_t do_constant_folding_2(TCGOpcode op, uint64_t x, uint64_t y)
} }
} }
static uint64_t do_constant_folding(TCGOpcode op, uint64_t x, uint64_t y) static uint64_t do_constant_folding(TCGOpcode op, TCGType type,
uint64_t x, uint64_t y)
{ {
const TCGOpDef *def = &tcg_op_defs[op];
uint64_t res = do_constant_folding_2(op, x, y); uint64_t res = do_constant_folding_2(op, x, y);
if (!(def->flags & TCG_OPF_64BIT)) { if (type == TCG_TYPE_I32) {
res = (int32_t)res; res = (int32_t)res;
} }
return res; return res;
@ -510,19 +507,21 @@ static bool do_constant_folding_cond_eq(TCGCond c)
* Return -1 if the condition can't be simplified, * Return -1 if the condition can't be simplified,
* and the result of the condition (0 or 1) if it can. * and the result of the condition (0 or 1) if it can.
*/ */
static int do_constant_folding_cond(TCGOpcode op, TCGArg x, static int do_constant_folding_cond(TCGType type, TCGArg x,
TCGArg y, TCGCond c) TCGArg y, TCGCond c)
{ {
uint64_t xv = arg_info(x)->val; uint64_t xv = arg_info(x)->val;
uint64_t yv = arg_info(y)->val; uint64_t yv = arg_info(y)->val;
if (arg_is_const(x) && arg_is_const(y)) { if (arg_is_const(x) && arg_is_const(y)) {
const TCGOpDef *def = &tcg_op_defs[op]; switch (type) {
tcg_debug_assert(!(def->flags & TCG_OPF_VECTOR)); case TCG_TYPE_I32:
if (def->flags & TCG_OPF_64BIT) {
return do_constant_folding_cond_64(xv, yv, c);
} else {
return do_constant_folding_cond_32(xv, yv, c); return do_constant_folding_cond_32(xv, yv, c);
case TCG_TYPE_I64:
return do_constant_folding_cond_64(xv, yv, c);
default:
/* Only scalar comparisons are optimizable */
return -1;
} }
} else if (args_are_copies(x, y)) { } else if (args_are_copies(x, y)) {
return do_constant_folding_cond_eq(c); return do_constant_folding_cond_eq(c);
@ -677,7 +676,7 @@ static bool fold_const1(OptContext *ctx, TCGOp *op)
uint64_t t; uint64_t t;
t = arg_info(op->args[1])->val; t = arg_info(op->args[1])->val;
t = do_constant_folding(op->opc, t, 0); t = do_constant_folding(op->opc, ctx->type, t, 0);
return tcg_opt_gen_movi(ctx, op, op->args[0], t); return tcg_opt_gen_movi(ctx, op, op->args[0], t);
} }
return false; return false;
@ -689,7 +688,7 @@ static bool fold_const2(OptContext *ctx, TCGOp *op)
uint64_t t1 = arg_info(op->args[1])->val; uint64_t t1 = arg_info(op->args[1])->val;
uint64_t t2 = arg_info(op->args[2])->val; uint64_t t2 = arg_info(op->args[2])->val;
t1 = do_constant_folding(op->opc, t1, t2); t1 = do_constant_folding(op->opc, ctx->type, t1, t2);
return tcg_opt_gen_movi(ctx, op, op->args[0], t1); return tcg_opt_gen_movi(ctx, op, op->args[0], t1);
} }
return false; return false;
@ -791,7 +790,7 @@ static bool fold_andc(OptContext *ctx, TCGOp *op)
static bool fold_brcond(OptContext *ctx, TCGOp *op) static bool fold_brcond(OptContext *ctx, TCGOp *op)
{ {
TCGCond cond = op->args[2]; TCGCond cond = op->args[2];
int i = do_constant_folding_cond(op->opc, op->args[0], op->args[1], cond); int i = do_constant_folding_cond(ctx->type, op->args[0], op->args[1], cond);
if (i == 0) { if (i == 0) {
tcg_op_remove(ctx->tcg, op); tcg_op_remove(ctx->tcg, op);
@ -836,7 +835,7 @@ static bool fold_brcond2(OptContext *ctx, TCGOp *op)
* Simplify EQ/NE comparisons where one of the pairs * Simplify EQ/NE comparisons where one of the pairs
* can be simplified. * can be simplified.
*/ */
i = do_constant_folding_cond(INDEX_op_brcond_i32, op->args[0], i = do_constant_folding_cond(TCG_TYPE_I32, op->args[0],
op->args[2], cond); op->args[2], cond);
switch (i ^ inv) { switch (i ^ inv) {
case 0: case 0:
@ -845,7 +844,7 @@ static bool fold_brcond2(OptContext *ctx, TCGOp *op)
goto do_brcond_high; goto do_brcond_high;
} }
i = do_constant_folding_cond(INDEX_op_brcond_i32, op->args[1], i = do_constant_folding_cond(TCG_TYPE_I32, op->args[1],
op->args[3], cond); op->args[3], cond);
switch (i ^ inv) { switch (i ^ inv) {
case 0: case 0:
@ -887,7 +886,7 @@ static bool fold_bswap(OptContext *ctx, TCGOp *op)
if (arg_is_const(op->args[1])) { if (arg_is_const(op->args[1])) {
uint64_t t = arg_info(op->args[1])->val; uint64_t t = arg_info(op->args[1])->val;
t = do_constant_folding(op->opc, t, op->args[2]); t = do_constant_folding(op->opc, ctx->type, t, op->args[2]);
return tcg_opt_gen_movi(ctx, op, op->args[0], t); return tcg_opt_gen_movi(ctx, op, op->args[0], t);
} }
return false; return false;
@ -931,7 +930,7 @@ static bool fold_count_zeros(OptContext *ctx, TCGOp *op)
uint64_t t = arg_info(op->args[1])->val; uint64_t t = arg_info(op->args[1])->val;
if (t != 0) { if (t != 0) {
t = do_constant_folding(op->opc, t, 0); t = do_constant_folding(op->opc, ctx->type, t, 0);
return tcg_opt_gen_movi(ctx, op, op->args[0], t); return tcg_opt_gen_movi(ctx, op, op->args[0], t);
} }
return tcg_opt_gen_mov(ctx, op, op->args[0], op->args[2]); return tcg_opt_gen_mov(ctx, op, op->args[0], op->args[2]);
@ -1063,9 +1062,8 @@ static bool fold_mov(OptContext *ctx, TCGOp *op)
static bool fold_movcond(OptContext *ctx, TCGOp *op) static bool fold_movcond(OptContext *ctx, TCGOp *op)
{ {
TCGOpcode opc = op->opc;
TCGCond cond = op->args[5]; TCGCond cond = op->args[5];
int i = do_constant_folding_cond(opc, op->args[1], op->args[2], cond); int i = do_constant_folding_cond(ctx->type, op->args[1], op->args[2], cond);
if (i >= 0) { if (i >= 0) {
return tcg_opt_gen_mov(ctx, op, op->args[0], op->args[4 - i]); return tcg_opt_gen_mov(ctx, op, op->args[0], op->args[4 - i]);
@ -1074,9 +1072,18 @@ static bool fold_movcond(OptContext *ctx, TCGOp *op)
if (arg_is_const(op->args[3]) && arg_is_const(op->args[4])) { if (arg_is_const(op->args[3]) && arg_is_const(op->args[4])) {
uint64_t tv = arg_info(op->args[3])->val; uint64_t tv = arg_info(op->args[3])->val;
uint64_t fv = arg_info(op->args[4])->val; uint64_t fv = arg_info(op->args[4])->val;
TCGOpcode opc;
opc = (opc == INDEX_op_movcond_i32 switch (ctx->type) {
? INDEX_op_setcond_i32 : INDEX_op_setcond_i64); case TCG_TYPE_I32:
opc = INDEX_op_setcond_i32;
break;
case TCG_TYPE_I64:
opc = INDEX_op_setcond_i64;
break;
default:
g_assert_not_reached();
}
if (tv == 1 && fv == 0) { if (tv == 1 && fv == 0) {
op->opc = opc; op->opc = opc;
@ -1181,7 +1188,7 @@ static bool fold_remainder(OptContext *ctx, TCGOp *op)
static bool fold_setcond(OptContext *ctx, TCGOp *op) static bool fold_setcond(OptContext *ctx, TCGOp *op)
{ {
TCGCond cond = op->args[3]; TCGCond cond = op->args[3];
int i = do_constant_folding_cond(op->opc, op->args[1], op->args[2], cond); int i = do_constant_folding_cond(ctx->type, op->args[1], op->args[2], cond);
if (i >= 0) { if (i >= 0) {
return tcg_opt_gen_movi(ctx, op, op->args[0], i); return tcg_opt_gen_movi(ctx, op, op->args[0], i);
@ -1220,7 +1227,7 @@ static bool fold_setcond2(OptContext *ctx, TCGOp *op)
* Simplify EQ/NE comparisons where one of the pairs * Simplify EQ/NE comparisons where one of the pairs
* can be simplified. * can be simplified.
*/ */
i = do_constant_folding_cond(INDEX_op_setcond_i32, op->args[1], i = do_constant_folding_cond(TCG_TYPE_I32, op->args[1],
op->args[3], cond); op->args[3], cond);
switch (i ^ inv) { switch (i ^ inv) {
case 0: case 0:
@ -1229,7 +1236,7 @@ static bool fold_setcond2(OptContext *ctx, TCGOp *op)
goto do_setcond_high; goto do_setcond_high;
} }
i = do_constant_folding_cond(INDEX_op_setcond_i32, op->args[2], i = do_constant_folding_cond(TCG_TYPE_I32, op->args[2],
op->args[4], cond); op->args[4], cond);
switch (i ^ inv) { switch (i ^ inv) {
case 0: case 0:
@ -1331,6 +1338,15 @@ void tcg_optimize(TCGContext *s)
init_arguments(&ctx, op, def->nb_oargs + def->nb_iargs); init_arguments(&ctx, op, def->nb_oargs + def->nb_iargs);
copy_propagate(&ctx, op, def->nb_oargs, def->nb_iargs); copy_propagate(&ctx, op, def->nb_oargs, def->nb_iargs);
/* Pre-compute the type of the operation. */
if (def->flags & TCG_OPF_VECTOR) {
ctx.type = TCG_TYPE_V64 + TCGOP_VECL(op);
} else if (def->flags & TCG_OPF_64BIT) {
ctx.type = TCG_TYPE_I64;
} else {
ctx.type = TCG_TYPE_I32;
}
/* For commutative operations make constant second argument */ /* For commutative operations make constant second argument */
switch (opc) { switch (opc) {
CASE_OP_32_64_VEC(add): CASE_OP_32_64_VEC(add):
@ -1411,19 +1427,24 @@ void tcg_optimize(TCGContext *s)
/* Proceed with possible constant folding. */ /* Proceed with possible constant folding. */
break; break;
} }
if (opc == INDEX_op_sub_i32) { switch (ctx.type) {
case TCG_TYPE_I32:
neg_op = INDEX_op_neg_i32; neg_op = INDEX_op_neg_i32;
have_neg = TCG_TARGET_HAS_neg_i32; have_neg = TCG_TARGET_HAS_neg_i32;
} else if (opc == INDEX_op_sub_i64) { break;
case TCG_TYPE_I64:
neg_op = INDEX_op_neg_i64; neg_op = INDEX_op_neg_i64;
have_neg = TCG_TARGET_HAS_neg_i64; have_neg = TCG_TARGET_HAS_neg_i64;
} else if (TCG_TARGET_HAS_neg_vec) {
TCGType type = TCGOP_VECL(op) + TCG_TYPE_V64;
unsigned vece = TCGOP_VECE(op);
neg_op = INDEX_op_neg_vec;
have_neg = tcg_can_emit_vec_op(neg_op, type, vece) > 0;
} else {
break; break;
case TCG_TYPE_V64:
case TCG_TYPE_V128:
case TCG_TYPE_V256:
neg_op = INDEX_op_neg_vec;
have_neg = tcg_can_emit_vec_op(neg_op, ctx.type,
TCGOP_VECE(op)) > 0;
break;
default:
g_assert_not_reached();
} }
if (!have_neg) { if (!have_neg) {
break; break;
@ -1476,15 +1497,23 @@ void tcg_optimize(TCGContext *s)
TCGOpcode not_op; TCGOpcode not_op;
bool have_not; bool have_not;
if (def->flags & TCG_OPF_VECTOR) { switch (ctx.type) {
not_op = INDEX_op_not_vec; case TCG_TYPE_I32:
have_not = TCG_TARGET_HAS_not_vec;
} else if (def->flags & TCG_OPF_64BIT) {
not_op = INDEX_op_not_i64;
have_not = TCG_TARGET_HAS_not_i64;
} else {
not_op = INDEX_op_not_i32; not_op = INDEX_op_not_i32;
have_not = TCG_TARGET_HAS_not_i32; have_not = TCG_TARGET_HAS_not_i32;
break;
case TCG_TYPE_I64:
not_op = INDEX_op_not_i64;
have_not = TCG_TARGET_HAS_not_i64;
break;
case TCG_TYPE_V64:
case TCG_TYPE_V128:
case TCG_TYPE_V256:
not_op = INDEX_op_not_vec;
have_not = TCG_TARGET_HAS_not_vec;
break;
default:
g_assert_not_reached();
} }
if (!have_not) { if (!have_not) {
break; break;
@ -1755,7 +1784,7 @@ void tcg_optimize(TCGContext *s)
below, we can ignore high bits, but for further optimizations we below, we can ignore high bits, but for further optimizations we
need to record that the high bits contain garbage. */ need to record that the high bits contain garbage. */
partmask = z_mask; partmask = z_mask;
if (!(def->flags & TCG_OPF_64BIT)) { if (ctx.type == TCG_TYPE_I32) {
z_mask |= ~(tcg_target_ulong)0xffffffffu; z_mask |= ~(tcg_target_ulong)0xffffffffu;
partmask &= 0xffffffffu; partmask &= 0xffffffffu;
affected &= 0xffffffffu; affected &= 0xffffffffu;