The definition of op_data included opcode bits, not just
the size field of the various ldst instructions.
Signed-off-by: Richard Henderson <r...@twiddle.net>
---
tcg/aarch64/tcg-target.c | 111 +++++++++++++++++++++--------------------------
1 file changed, 49 insertions(+), 62 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 0746f2a..d6d3e86 100644
--- a/tcg/aarch64/tcg-target.c
+++ b/tcg/aarch64/tcg-target.c
@@ -242,13 +242,6 @@ static const enum aarch64_cond_code tcg_cond_to_aarch64[]
= {
[TCG_COND_LEU] = COND_LS,
};
-/* opcodes for LDR / STR instructions with base + simm9 addressing */
-enum aarch64_ldst_op_data { /* size of the data moved */
- LDST_8 = 0x38,
- LDST_16 = 0x78,
- LDST_32 = 0xb8,
- LDST_64 = 0xf8,
-};
enum aarch64_ldst_op_type { /* type of operation */
LDST_ST = 0x0, /* store */
LDST_LD = 0x4, /* load */
@@ -490,25 +483,23 @@ static void tcg_out_insn_3509(TCGContext *s, AArch64Insn
insn, TCGType ext,
}
-static inline void tcg_out_ldst_9(TCGContext *s,
- enum aarch64_ldst_op_data op_data,
+static inline void tcg_out_ldst_9(TCGContext *s, TCGMemOp size,
enum aarch64_ldst_op_type op_type,
TCGReg rd, TCGReg rn, intptr_t offset)
{
/* use LDUR with BASE register with 9bit signed unscaled offset */
- tcg_out32(s, op_data << 24 | op_type << 20
+ tcg_out32(s, 0x38000000 | size << 30 | op_type << 20
| (offset & 0x1ff) << 12 | rn << 5 | rd);
}
/* tcg_out_ldst_12 expects a scaled unsigned immediate offset */
-static inline void tcg_out_ldst_12(TCGContext *s,
- enum aarch64_ldst_op_data op_data,
+static inline void tcg_out_ldst_12(TCGContext *s, TCGMemOp size,
enum aarch64_ldst_op_type op_type,
TCGReg rd, TCGReg rn,
tcg_target_ulong scaled_uimm)
{
- tcg_out32(s, (op_data | 1) << 24
- | op_type << 20 | scaled_uimm << 10 | rn << 5 | rd);
+ tcg_out32(s, 0x39000000 | size << 30 | op_type << 20
+ | scaled_uimm << 10 | rn << 5 | rd);
}
/* Register to register move using ORR (shifted register with no shift). */
@@ -637,44 +628,40 @@ static void tcg_out_movi(TCGContext *s, TCGType type,
TCGReg rd,
}
}
-static inline void tcg_out_ldst_r(TCGContext *s,
- enum aarch64_ldst_op_data op_data,
+static inline void tcg_out_ldst_r(TCGContext *s, TCGMemOp size,
enum aarch64_ldst_op_type op_type,
TCGReg rd, TCGReg base, TCGReg regoff)
{
/* load from memory to register using base + 64bit register offset */
/* using f.e. STR Wt, [Xn, Xm] 0xb8600800|(regoff << 16)|(base << 5)|rd */
/* the 0x6000 is for the "no extend field" */
- tcg_out32(s, 0x00206800
- | op_data << 24 | op_type << 20 | regoff << 16 | base << 5 | rd);
+ tcg_out32(s, 0x38206800 | size << 30 | op_type << 20
+ | regoff << 16 | base << 5 | rd);
}
/* solve the whole ldst problem */
-static inline void tcg_out_ldst(TCGContext *s, enum aarch64_ldst_op_data data,
+static inline void tcg_out_ldst(TCGContext *s, TCGMemOp size,
enum aarch64_ldst_op_type type,
TCGReg rd, TCGReg rn, intptr_t offset)
{
if (offset >= -256 && offset < 256) {
- tcg_out_ldst_9(s, data, type, rd, rn, offset);
+ tcg_out_ldst_9(s, size, type, rd, rn, offset);
return;
}
- if (offset >= 256) {
- /* if the offset is naturally aligned and in range,
- then we can use the scaled uimm12 encoding */
- unsigned int s_bits = data >> 6;
- if (!(offset & ((1 << s_bits) - 1))) {
- tcg_target_ulong scaled_uimm = offset >> s_bits;
- if (scaled_uimm <= 0xfff) {
- tcg_out_ldst_12(s, data, type, rd, rn, scaled_uimm);
- return;
- }
+ /* If the offset is naturally aligned and in range, then we can
+ use the scaled uimm12 encoding */
+ if (offset >= 0 && !(offset & ((1 << size) - 1))) {
+ tcg_target_ulong scaled_uimm = offset >> size;
+ if (scaled_uimm <= 0xfff) {
+ tcg_out_ldst_12(s, size, type, rd, rn, scaled_uimm);
+ return;
}
}
- /* worst-case scenario, move offset to temp register, use reg offset */
+ /* Worst-case scenario, move offset to temp register, use reg offset. */
tcg_out_movi(s, TCG_TYPE_I64, TCG_REG_TMP, offset);
- tcg_out_ldst_r(s, data, type, rd, rn, TCG_REG_TMP);
+ tcg_out_ldst_r(s, size, type, rd, rn, TCG_REG_TMP);
}
static inline void tcg_out_mov(TCGContext *s,
@@ -688,14 +675,14 @@ static inline void tcg_out_mov(TCGContext *s,
static inline void tcg_out_ld(TCGContext *s, TCGType type, TCGReg arg,
TCGReg arg1, intptr_t arg2)
{
- tcg_out_ldst(s, (type == TCG_TYPE_I64) ? LDST_64 : LDST_32, LDST_LD,
+ tcg_out_ldst(s, type == TCG_TYPE_I64 ? MO_64 : MO_32, LDST_LD,
arg, arg1, arg2);
}
static inline void tcg_out_st(TCGContext *s, TCGType type, TCGReg arg,
TCGReg arg1, intptr_t arg2)
{
- tcg_out_ldst(s, (type == TCG_TYPE_I64) ? LDST_64 : LDST_32, LDST_ST,
+ tcg_out_ldst(s, type == TCG_TYPE_I64 ? MO_64 : MO_32, LDST_ST,
arg, arg1, arg2);
}
@@ -1095,12 +1082,12 @@ static void tcg_out_tlb_read(TCGContext *s, TCGReg
addr_reg, TCGMemOp s_bits,
/* Merge "low bits" from tlb offset, load the tlb comparator into X0.
X0 = load [X2 + (tlb_offset & 0x000fff)] */
- tcg_out_ldst(s, TARGET_LONG_BITS == 64 ? LDST_64 : LDST_32,
+ tcg_out_ldst(s, TARGET_LONG_BITS == 64 ? MO_64 : MO_32,
LDST_LD, TCG_REG_X0, TCG_REG_X2, tlb_offset & 0xfff);
/* Load the tlb addend. Do that early to avoid stalling.
X1 = load [X2 + (tlb_offset & 0xfff) + offsetof(addend)] */
- tcg_out_ldst(s, LDST_64, LDST_LD, TCG_REG_X1, TCG_REG_X2,
+ tcg_out_ldst(s, MO_64, LDST_LD, TCG_REG_X1, TCG_REG_X2,
(tlb_offset & 0xfff) + (offsetof(CPUTLBEntry, addend)) -
(is_read ? offsetof(CPUTLBEntry, addr_read)
: offsetof(CPUTLBEntry, addr_write)));
@@ -1122,43 +1109,43 @@ static void tcg_out_qemu_ld_direct(TCGContext *s,
TCGMemOp memop,
switch (memop & MO_SSIZE) {
case MO_UB:
- tcg_out_ldst_r(s, LDST_8, LDST_LD, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_8, LDST_LD, data_r, addr_r, off_r);
break;
case MO_SB:
- tcg_out_ldst_r(s, LDST_8, LDST_LD_S_X, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_8, LDST_LD_S_X, data_r, addr_r, off_r);
break;
case MO_UW:
- tcg_out_ldst_r(s, LDST_16, LDST_LD, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_16, LDST_LD, data_r, addr_r, off_r);
if (bswap) {
tcg_out_rev16(s, TCG_TYPE_I32, data_r, data_r);
}
break;
case MO_SW:
if (bswap) {
- tcg_out_ldst_r(s, LDST_16, LDST_LD, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_16, LDST_LD, data_r, addr_r, off_r);
tcg_out_rev16(s, TCG_TYPE_I32, data_r, data_r);
tcg_out_sxt(s, TCG_TYPE_I64, MO_16, data_r, data_r);
} else {
- tcg_out_ldst_r(s, LDST_16, LDST_LD_S_X, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_16, LDST_LD_S_X, data_r, addr_r, off_r);
}
break;
case MO_UL:
- tcg_out_ldst_r(s, LDST_32, LDST_LD, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_32, LDST_LD, data_r, addr_r, off_r);
if (bswap) {
tcg_out_rev(s, TCG_TYPE_I32, data_r, data_r);
}
break;
case MO_SL:
if (bswap) {
- tcg_out_ldst_r(s, LDST_32, LDST_LD, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_32, LDST_LD, data_r, addr_r, off_r);
tcg_out_rev(s, TCG_TYPE_I32, data_r, data_r);
tcg_out_sxt(s, TCG_TYPE_I64, MO_32, data_r, data_r);
} else {
- tcg_out_ldst_r(s, LDST_32, LDST_LD_S_X, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_32, LDST_LD_S_X, data_r, addr_r, off_r);
}
break;
case MO_Q:
- tcg_out_ldst_r(s, LDST_64, LDST_LD, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_64, LDST_LD, data_r, addr_r, off_r);
if (bswap) {
tcg_out_rev(s, TCG_TYPE_I64, data_r, data_r);
}
@@ -1175,28 +1162,28 @@ static void tcg_out_qemu_st_direct(TCGContext *s,
TCGMemOp memop,
switch (memop & MO_SIZE) {
case MO_8:
- tcg_out_ldst_r(s, LDST_8, LDST_ST, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_8, LDST_ST, data_r, addr_r, off_r);
break;
case MO_16:
if (bswap) {
tcg_out_rev16(s, TCG_TYPE_I32, TCG_REG_TMP, data_r);
data_r = TCG_REG_TMP;
}
- tcg_out_ldst_r(s, LDST_16, LDST_ST, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_16, LDST_ST, data_r, addr_r, off_r);
break;
case MO_32:
if (bswap) {
tcg_out_rev(s, TCG_TYPE_I32, TCG_REG_TMP, data_r);
data_r = TCG_REG_TMP;
}
- tcg_out_ldst_r(s, LDST_32, LDST_ST, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_32, LDST_ST, data_r, addr_r, off_r);
break;
case MO_64:
if (bswap) {
tcg_out_rev(s, TCG_TYPE_I64, TCG_REG_TMP, data_r);
data_r = TCG_REG_TMP;
}
- tcg_out_ldst_r(s, LDST_64, LDST_ST, data_r, addr_r, off_r);
+ tcg_out_ldst_r(s, MO_64, LDST_ST, data_r, addr_r, off_r);
break;
default:
tcg_abort();
@@ -1308,49 +1295,49 @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc,
case INDEX_op_ld8u_i32:
case INDEX_op_ld8u_i64:
- tcg_out_ldst(s, LDST_8, LDST_LD, a0, a1, a2);
+ tcg_out_ldst(s, MO_8, LDST_LD, a0, a1, a2);
break;
case INDEX_op_ld8s_i32:
- tcg_out_ldst(s, LDST_8, LDST_LD_S_W, a0, a1, a2);
+ tcg_out_ldst(s, MO_8, LDST_LD_S_W, a0, a1, a2);
break;
case INDEX_op_ld8s_i64:
- tcg_out_ldst(s, LDST_8, LDST_LD_S_X, a0, a1, a2);
+ tcg_out_ldst(s, MO_8, LDST_LD_S_X, a0, a1, a2);
break;
case INDEX_op_ld16u_i32:
case INDEX_op_ld16u_i64:
- tcg_out_ldst(s, LDST_16, LDST_LD, a0, a1, a2);
+ tcg_out_ldst(s, MO_16, LDST_LD, a0, a1, a2);
break;
case INDEX_op_ld16s_i32:
- tcg_out_ldst(s, LDST_16, LDST_LD_S_W, a0, a1, a2);
+ tcg_out_ldst(s, MO_16, LDST_LD_S_W, a0, a1, a2);
break;
case INDEX_op_ld16s_i64:
- tcg_out_ldst(s, LDST_16, LDST_LD_S_X, a0, a1, a2);
+ tcg_out_ldst(s, MO_16, LDST_LD_S_X, a0, a1, a2);
break;
case INDEX_op_ld_i32:
case INDEX_op_ld32u_i64:
- tcg_out_ldst(s, LDST_32, LDST_LD, a0, a1, a2);
+ tcg_out_ldst(s, MO_32, LDST_LD, a0, a1, a2);
break;
case INDEX_op_ld32s_i64:
- tcg_out_ldst(s, LDST_32, LDST_LD_S_X, a0, a1, a2);
+ tcg_out_ldst(s, MO_32, LDST_LD_S_X, a0, a1, a2);
break;
case INDEX_op_ld_i64:
- tcg_out_ldst(s, LDST_64, LDST_LD, a0, a1, a2);
+ tcg_out_ldst(s, MO_64, LDST_LD, a0, a1, a2);
break;
case INDEX_op_st8_i32:
case INDEX_op_st8_i64:
- tcg_out_ldst(s, LDST_8, LDST_ST, a0, a1, a2);
+ tcg_out_ldst(s, MO_8, LDST_ST, a0, a1, a2);
break;
case INDEX_op_st16_i32:
case INDEX_op_st16_i64:
- tcg_out_ldst(s, LDST_16, LDST_ST, a0, a1, a2);
+ tcg_out_ldst(s, MO_16, LDST_ST, a0, a1, a2);
break;
case INDEX_op_st_i32:
case INDEX_op_st32_i64:
- tcg_out_ldst(s, LDST_32, LDST_ST, a0, a1, a2);
+ tcg_out_ldst(s, MO_32, LDST_ST, a0, a1, a2);
break;
case INDEX_op_st_i64:
- tcg_out_ldst(s, LDST_64, LDST_ST, a0, a1, a2);
+ tcg_out_ldst(s, MO_64, LDST_ST, a0, a1, a2);
break;
case INDEX_op_add_i32:
--
1.8.5.3
