tcg: Increase tcg_out_dupi_vec immediate to int64_t

While we don't store more than tcg_target_long in TCGTemp,
we shouldn't be limited to that for code generation.  We will
be able to use this for INDEX_op_dup2_vec with 2 constants.

Also pass along the minimal vece that may be said to apply
to the constant.  This allows some simplification in the
various backends.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This commit is contained in:
Richard Henderson 2020-03-31 01:02:08 -07:00
parent 0a6a8bc8eb
commit 4e18617555
4 changed files with 69 additions and 33 deletions

View file

@ -912,31 +912,41 @@ static void tcg_out_movi_int(TCGContext *s, TCGType type, TCGReg ret,
}
}
static void tcg_out_dupi_vec(TCGContext *s, TCGType type, TCGReg ret,
tcg_target_long val)
static void tcg_out_dupi_vec(TCGContext *s, TCGType type, unsigned vece,
TCGReg ret, int64_t val)
{
uint32_t load_insn;
int rel, low;
intptr_t add;
low = (int8_t)val;
if (low >= -16 && low < 16) {
if (val == (tcg_target_long)dup_const(MO_8, low)) {
switch (vece) {
case MO_8:
low = (int8_t)val;
if (low >= -16 && low < 16) {
tcg_out32(s, VSPLTISB | VRT(ret) | ((val & 31) << 16));
return;
}
if (val == (tcg_target_long)dup_const(MO_16, low)) {
if (have_isa_3_00) {
tcg_out32(s, XXSPLTIB | VRT(ret) | ((val & 0xff) << 11));
return;
}
break;
case MO_16:
low = (int16_t)val;
if (low >= -16 && low < 16) {
tcg_out32(s, VSPLTISH | VRT(ret) | ((val & 31) << 16));
return;
}
if (val == (tcg_target_long)dup_const(MO_32, low)) {
break;
case MO_32:
low = (int32_t)val;
if (low >= -16 && low < 16) {
tcg_out32(s, VSPLTISW | VRT(ret) | ((val & 31) << 16));
return;
}
}
if (have_isa_3_00 && val == (tcg_target_long)dup_const(MO_8, val)) {
tcg_out32(s, XXSPLTIB | VRT(ret) | ((val & 0xff) << 11));
return;
break;
}
/*
@ -956,14 +966,15 @@ static void tcg_out_dupi_vec(TCGContext *s, TCGType type, TCGReg ret,
if (TCG_TARGET_REG_BITS == 64) {
new_pool_label(s, val, rel, s->code_ptr, add);
} else {
new_pool_l2(s, rel, s->code_ptr, add, val, val);
new_pool_l2(s, rel, s->code_ptr, add, val >> 32, val);
}
} else {
load_insn = LVX | VRT(ret) | RB(TCG_REG_TMP1);
if (TCG_TARGET_REG_BITS == 64) {
new_pool_l2(s, rel, s->code_ptr, add, val, val);
} else {
new_pool_l4(s, rel, s->code_ptr, add, val, val, val, val);
new_pool_l4(s, rel, s->code_ptr, add,
val >> 32, val, val >> 32, val);
}
}