tcg: allocate optimizer temps with tcg_malloc

Groundwork for supporting multiple TCG contexts.

While at it, also allocate temps_used directly as a bitmap of the
required size, instead of using a bitmap of TCG_MAX_TEMPS via
TCGTempSet.

Performance-wise we lose about 1.12% in a translation-heavy workload
such as booting+shutting down debian-arm:

Performance counter stats for 'taskset -c 0 arm-softmmu/qemu-system-arm \
-machine type=virt -nographic -smp 1 -m 4096 \
-netdev user,id=unet,hostfwd=tcp::2222-:22 \
-device virtio-net-device,netdev=unet \
-drive file=die-on-boot.qcow2,id=myblock,index=0,if=none \
-device virtio-blk-device,drive=myblock \
-kernel kernel.img -append console=ttyAMA0 root=/dev/vda1 \
-name arm,debug-threads=on -smp 1' (10 runs):

exec time (s) Relative slowdown wrt original (%)
---------------------------------------------------------------
original 20.213321616 0.
tcg_malloc 20.441130078 1.1270214
TCGContext 20.477846517 1.3086662
g_malloc 20.780527895 2.8061013

The other two alternatives shown in the table are:
- TCGContext: embed temps[TCG_MAX_TEMPS] and TCGTempSet used_temps
in TCGContext. This is simple enough but it isn't faster than using
tcg_malloc; moreover, it wastes memory.
- g_malloc: allocate/deallocate both temps and used_temps every time
tcg_optimize is executed.

Backports commit 34184b071817b4f9edbfd1aa2225c196f05a0947 from qemu
This commit is contained in:
Emilio G. Cota 2018-03-14 10:04:48 -04:00 committed by Lioncash
parent 1be7b55bb4
commit 5ad6116f20
No known key found for this signature in database
GPG key ID: 4E3C3CC1031BA9C7
3 changed files with 29 additions and 27 deletions

View file

@ -70,6 +70,17 @@ static inline unsigned long *bitmap_new(long nbits)
return ptr;
}
static inline void bitmap_zero(unsigned long *dst, long nbits)
{
if (small_nbits(nbits)) {
*dst = 0UL;
} else {
long len = BITS_TO_LONGS(nbits) * sizeof(unsigned long);
memset(dst, 0, len);
}
}
static inline long bitmap_count_one(const unsigned long *bitmap, long nbits)
{
if (small_nbits(nbits)) {

View file

@ -82,22 +82,13 @@ static void reset_temp(TCGContext *s, TCGArg arg)
reset_ts(arg_temp(arg));
}
/* Reset all temporaries, given that there are NB_TEMPS of them. */
static void reset_all_temps(TCGContext *s, int nb_temps)
{
long len = BITS_TO_LONGS(nb_temps) * sizeof(unsigned long);
memset(&s->temps2_used.l, 0, len);
}
/* Initialize and activate a temporary. */
static void init_ts_info(TCGContext *s, TCGTemp *ts)
static void init_ts_info(TCGContext *s, struct tcg_temp_info *infos,
TCGTempSet *temps_used, TCGTemp *ts)
{
struct tcg_temp_info *temps = s->temps2;
TCGTempSet *temps_used = &s->temps2_used;
size_t idx = temp_idx(s, ts);
if (!test_bit(idx, temps_used->l)) {
struct tcg_temp_info *ti = &temps[idx];
struct tcg_temp_info *ti = &infos[idx];
ts->state_ptr = ti;
ti->next_copy = ts;
@ -108,9 +99,10 @@ static void init_ts_info(TCGContext *s, TCGTemp *ts)
}
}
static void init_arg_info(TCGContext *s, TCGArg arg)
static void init_arg_info(TCGContext *s, struct tcg_temp_info *infos,
TCGTempSet *temps_used, TCGArg arg)
{
init_ts_info(s, arg_temp(arg));
init_ts_info(s, infos, temps_used, arg_temp(arg));
}
static TCGTemp *find_better_copy(TCGTemp *ts)
@ -594,6 +586,8 @@ void tcg_optimize(TCGContext *s)
{
int nb_temps, nb_globals;
TCGOp *op, *op_next, *prev_mb = NULL;
struct tcg_temp_info *infos;
TCGTempSet temps_used;
/* Array VALS has an element for each temp.
If this temp holds a constant then its value is kept in VALS' element.
@ -602,7 +596,8 @@ void tcg_optimize(TCGContext *s)
nb_temps = s->nb_temps;
nb_globals = s->nb_globals;
reset_all_temps(s, nb_temps);
bitmap_zero(temps_used.l, nb_temps);
infos = tcg_malloc(s, sizeof(struct tcg_temp_info) * nb_temps);
QTAILQ_FOREACH_SAFE(op, &s->ops, link, op_next) {
tcg_target_ulong mask, partmask, affected;
@ -619,14 +614,14 @@ void tcg_optimize(TCGContext *s)
for (i = 0; i < nb_oargs + nb_iargs; i++) {
TCGTemp *ts = arg_temp(op->args[i]);
if (ts) {
init_ts_info(s, ts);
init_ts_info(s, infos, &temps_used, ts);
}
}
} else {
nb_oargs = def->nb_oargs;
nb_iargs = def->nb_iargs;
for (i = 0; i < nb_oargs + nb_iargs; i++) {
init_arg_info(s, op->args[i]);
init_arg_info(s, infos, &temps_used, op->args[i]);
}
}
@ -1203,7 +1198,7 @@ void tcg_optimize(TCGContext *s)
op->args[1], op->args[2]);
if (tmp != 2) {
if (tmp) {
reset_all_temps(s, nb_temps);
bitmap_zero(temps_used.l, nb_temps);
op->opc = INDEX_op_br;
op->args[0] = op->args[3];
} else {
@ -1286,7 +1281,7 @@ void tcg_optimize(TCGContext *s)
if (tmp != 2) {
if (tmp) {
do_brcond_true:
reset_all_temps(s, nb_temps);
bitmap_zero(temps_used.l, nb_temps);
op->opc = INDEX_op_br;
op->args[0] = op->args[5];
} else {
@ -1302,7 +1297,7 @@ void tcg_optimize(TCGContext *s)
/* Simplify LT/GE comparisons vs zero to a single compare
vs the high word of the input. */
do_brcond_high:
reset_all_temps(s, nb_temps);
bitmap_zero(temps_used.l, nb_temps);
op->opc = INDEX_op_brcond_i32;
op->args[0] = op->args[1];
op->args[1] = op->args[3];
@ -1328,7 +1323,7 @@ void tcg_optimize(TCGContext *s)
goto do_default;
}
do_brcond_low:
reset_all_temps(s, nb_temps);
bitmap_zero(temps_used.l, nb_temps);
op->opc = INDEX_op_brcond_i32;
op->args[1] = op->args[2];
op->args[2] = op->args[4];
@ -1433,7 +1428,7 @@ void tcg_optimize(TCGContext *s)
if (!(op->args[nb_oargs + nb_iargs + 1]
& (TCG_CALL_NO_READ_GLOBALS | TCG_CALL_NO_WRITE_GLOBALS))) {
for (i = 0; i < nb_globals; i++) {
if (test_bit(i, s->temps2_used.l)) {
if (test_bit(i, temps_used.l)) {
reset_ts(&s->temps[i]);
}
}
@ -1448,7 +1443,7 @@ void tcg_optimize(TCGContext *s)
block, otherwise we only trash the output args. "mask" is
the non-zero bits mask for the first output arg. */
if (def->flags & TCG_OPF_BB_END) {
reset_all_temps(s, nb_temps);
bitmap_zero(temps_used.l, nb_temps);
} else {
do_reset_output:
for (i = 0; i < nb_oargs; i++) {

View file

@ -851,10 +851,6 @@ struct TCGContext {
int indirect_reg_alloc_order[50];
TCGOpDef *tcg_op_defs;
/* qemu/tcg/optimize.c */
struct tcg_temp_info temps2[TCG_MAX_TEMPS];
TCGTempSet temps2_used;
/* qemu/target-m68k/translate.c */
TCGv_i32 cpu_halted;
char cpu_reg_names[2 * 8 * 3 + 5 * 4];