tcg: Mask TCGMemOp appropriately for indexing

The addition of MO_AMASK means that places that used inverted masks
need to be changed to use positive masks, and places that failed to
mask the intended bits need updating.

Backports commit 2b7ec66f025263a5331f37d5ad78a625496fd7bd from qemu
This commit is contained in:
Richard Henderson 2018-02-10 20:29:09 -05:00 committed by Lioncash
parent 336833c11e
commit c5a2a50c06
No known key found for this signature in database
GPG key ID: 4E3C3CC1031BA9C7
7 changed files with 30 additions and 30 deletions

View file

@ -1004,7 +1004,7 @@ static void tcg_out_qemu_ld_slow_path(TCGContext *s, TCGLabelQemuLdst *lb)
tcg_out_mov(s, TARGET_LONG_BITS == 64, TCG_REG_X1, lb->addrlo_reg);
tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_X2, oi);
tcg_out_adr(s, TCG_REG_X3, lb->raddr);
tcg_out_call(s, qemu_ld_helpers[opc & ~MO_SIGN]);
tcg_out_call(s, qemu_ld_helpers[opc & (MO_BSWAP | MO_SIZE)]);
if (opc & MO_SIGN) {
tcg_out_sxt(s, lb->type, size, lb->datalo_reg, TCG_REG_X0);
} else {
@ -1027,7 +1027,7 @@ static void tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *lb)
tcg_out_mov(s, size == MO_64, TCG_REG_X2, lb->datalo_reg);
tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_X3, oi);
tcg_out_adr(s, TCG_REG_X4, lb->raddr);
tcg_out_call(s, qemu_st_helpers[opc]);
tcg_out_call(s, qemu_st_helpers[opc & (MO_BSWAP | MO_SIZE)]);
tcg_out_goto(s, lb->raddr);
}

View file

@ -1260,9 +1260,9 @@ static void tcg_out_qemu_ld_slow_path(TCGContext *s, TCGLabelQemuLdst *lb)
icache usage. For pre-armv6, use the signed helpers since we do
not have a single insn sign-extend. */
if (use_armv6_instructions) {
func = qemu_ld_helpers[opc & ~MO_SIGN];
func = qemu_ld_helpers[opc & (MO_BSWAP | MO_SIZE)];
} else {
func = qemu_ld_helpers[opc];
func = qemu_ld_helpers[opc & (MO_BSWAP | MO_SSIZE)];
if (opc & MO_SIGN) {
opc = MO_UL;
}
@ -1337,7 +1337,7 @@ static void tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *lb)
argreg = tcg_out_arg_reg32(s, argreg, TCG_REG_R14);
/* Tail-call to the helper, which will return to the fast path. */
tcg_out_goto(s, COND_AL, qemu_st_helpers[opc]);
tcg_out_goto(s, COND_AL, qemu_st_helpers[opc & (MO_BSWAP | MO_SIZE)]);
}
#endif /* SOFTMMU */

View file

@ -1412,7 +1412,7 @@ static void tcg_out_qemu_ld_slow_path(TCGContext *s, TCGLabelQemuLdst *l)
(uintptr_t)l->raddr);
}
tcg_out_call(s, qemu_ld_helpers[opc & ~MO_SIGN]); // qq
tcg_out_call(s, qemu_ld_helpers[opc & (MO_BSWAP | MO_SIZE)]);
data_reg = l->datalo_reg;
switch (opc & MO_SSIZE) {
@ -1518,7 +1518,7 @@ static void tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l)
/* "Tail call" to the helper, with the return address back inline. */
tcg_out_push(s, retaddr);
tcg_out_jmp(s, qemu_st_helpers[opc]);
tcg_out_jmp(s, qemu_st_helpers[opc & (MO_BSWAP | MO_SIZE)]);
}
#elif defined(__x86_64__) && defined(__linux__)
# include <asm/prctl.h>

View file

@ -1031,7 +1031,7 @@ static void tcg_out_qemu_ld_slow_path(TCGContext *s, TCGLabelQemuLdst *l)
}
i = tcg_out_call_iarg_imm(s, i, oi);
i = tcg_out_call_iarg_imm(s, i, (intptr_t)l->raddr);
tcg_out_call_int(s, qemu_ld_helpers[opc], false);
tcg_out_call_int(s, qemu_ld_helpers[opc & (MO_BSWAP | MO_SSIZE)], false);
/* delay slot */
tcg_out_mov(s, TCG_TYPE_PTR, tcg_target_call_iarg_regs[0], TCG_AREG0);
@ -1094,7 +1094,7 @@ static void tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l)
computation to take place in the return address register. */
tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_RA, (intptr_t)l->raddr);
i = tcg_out_call_iarg_reg(s, i, TCG_REG_RA);
tcg_out_call_int(s, qemu_st_helpers[opc], true);
tcg_out_call_int(s, qemu_st_helpers[opc & (MO_BSWAP | MO_SIZE)], true);
/* delay slot */
tcg_out_mov(s, TCG_TYPE_PTR, tcg_target_call_iarg_regs[0], TCG_AREG0);
}

View file

@ -1495,7 +1495,7 @@ static void tcg_out_qemu_ld_slow_path(TCGContext *s, TCGLabelQemuLdst *lb)
tcg_out_movi(s, TCG_TYPE_I32, arg++, oi);
tcg_out32(s, MFSPR | RT(arg) | LR);
tcg_out_call(s, qemu_ld_helpers[opc & ~MO_SIGN]);
tcg_out_call(s, qemu_ld_helpers[opc & (MO_BSWAP | MO_SIZE)]);
lo = lb->datalo_reg;
hi = lb->datahi_reg;
@ -1565,7 +1565,7 @@ static void tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *lb)
tcg_out_movi(s, TCG_TYPE_I32, arg++, oi);
tcg_out32(s, MFSPR | RT(arg) | LR);
tcg_out_call(s, qemu_st_helpers[opc]);
tcg_out_call(s, qemu_st_helpers[opc & (MO_BSWAP | MO_SIZE)]);
tcg_out_b(s, 0, lb->raddr);
}
@ -1624,7 +1624,7 @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64)
tcg_out32(s, LWZ | TAI(datalo, addrlo, 4));
}
} else {
uint32_t insn = qemu_ldx_opc[opc];
uint32_t insn = qemu_ldx_opc[opc & (MO_BSWAP | MO_SSIZE)];
if (!HAVE_ISA_2_06 && insn == LDBRX) {
tcg_out32(s, ADDI | TAI(TCG_REG_R0, addrlo, 4));
tcg_out32(s, LWBRX | TAB(datalo, rbase, addrlo));
@ -1696,7 +1696,7 @@ static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is_64)
tcg_out32(s, STW | TAI(datalo, addrlo, 4));
}
} else {
uint32_t insn = qemu_stx_opc[opc];
uint32_t insn = qemu_stx_opc[opc & (MO_BSWAP | MO_SIZE)];
if (!HAVE_ISA_2_06 && insn == STDBRX) {
tcg_out32(s, STWBRX | SAB(datalo, rbase, addrlo));
tcg_out32(s, ADDI | TAI(TCG_REG_TMP1, addrlo, 4));

View file

@ -1573,7 +1573,7 @@ static void tcg_out_qemu_ld_slow_path(TCGContext *s, TCGLabelQemuLdst *lb)
}
tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_R4, oi);
tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_R5, (uintptr_t)lb->raddr);
tcg_out_call(s, qemu_ld_helpers[opc]);
tcg_out_call(s, qemu_ld_helpers[opc & (MO_BSWAP | MO_SSIZE)]);
tcg_out_mov(s, TCG_TYPE_I64, data_reg, TCG_REG_R2);
tgen_gotoi(s, S390_CC_ALWAYS, lb->raddr);
@ -1610,7 +1610,7 @@ static void tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *lb)
}
tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_R5, oi);
tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_R6, (uintptr_t)lb->raddr);
tcg_out_call(s, qemu_st_helpers[opc]);
tcg_out_call(s, qemu_st_helpers[opc & (MO_BSWAP | MO_SIZE)]);
tgen_gotoi(s, S390_CC_ALWAYS, lb->raddr);
}

View file

@ -1075,12 +1075,11 @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data, TCGReg addr,
TCGMemOp memop = get_memop(oi);
#ifdef CONFIG_SOFTMMU
unsigned memi = get_mmuidx(oi);
TCGMemOp s_bits = memop & MO_SIZE;
TCGReg addrz, param;
tcg_insn_unit *func;
tcg_insn_unit *label_ptr;
addrz = tcg_out_tlb_load(s, addr, memi, s_bits,
addrz = tcg_out_tlb_load(s, addr, memi, memop & MO_SIZE,
offsetof(CPUTLBEntry, addr_read));
/* The fast path is exactly one insn. Thus we can perform the
@ -1092,7 +1091,8 @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data, TCGReg addr,
tcg_out_bpcc0(s, COND_E, BPCC_A | BPCC_PT
| (TARGET_LONG_BITS == 64 ? BPCC_XCC : BPCC_ICC), 0);
/* delay slot */
tcg_out_ldst_rr(s, data, addrz, TCG_REG_O1, qemu_ld_opc[memop]);
tcg_out_ldst_rr(s, data, addrz, TCG_REG_O1,
qemu_ld_opc[memop & (MO_BSWAP | MO_SSIZE)]);
/* TLB Miss. */
@ -1105,10 +1105,10 @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data, TCGReg addr,
/* We use the helpers to extend SB and SW data, leaving the case
of SL needing explicit extending below. */
if ((memop & ~MO_BSWAP) == MO_SL) {
func = qemu_ld_trampoline[memop & ~MO_SIGN];
if ((memop & MO_SSIZE) == MO_SL) {
func = qemu_ld_trampoline[memop & (MO_BSWAP | MO_SIZE)];
} else {
func = qemu_ld_trampoline[memop];
func = qemu_ld_trampoline[memop & (MO_BSWAP | MO_SSIZE)];
}
assert(func != NULL);
tcg_out_call_nodelay(s, func);
@ -1119,13 +1119,13 @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data, TCGReg addr,
Which complicates things for sparcv8plus. */
if (SPARC64) {
/* We let the helper sign-extend SB and SW, but leave SL for here. */
if (is_64 && (memop & ~MO_BSWAP) == MO_SL) {
if (is_64 && (memop & MO_SSIZE) == MO_SL) {
tcg_out_arithi(s, data, TCG_REG_O0, 0, SHIFT_SRA);
} else {
tcg_out_mov(s, TCG_TYPE_REG, data, TCG_REG_O0);
}
} else {
if (s_bits == MO_64) {
if ((memop & MO_SIZE) == MO_64) {
tcg_out_arithi(s, TCG_REG_O0, TCG_REG_O0, 32, SHIFT_SLLX);
tcg_out_arithi(s, TCG_REG_O1, TCG_REG_O1, 0, SHIFT_SRL);
tcg_out_arith(s, data, TCG_REG_O0, TCG_REG_O1, ARITH_OR);
@ -1147,7 +1147,7 @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data, TCGReg addr,
}
tcg_out_ldst_rr(s, data, addr,
(GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_G0),
qemu_ld_opc[memop]);
qemu_ld_opc[memop & (MO_BSWAP | MO_SSIZE)]);
#endif /* CONFIG_SOFTMMU */
}
@ -1157,12 +1157,11 @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg data, TCGReg addr,
TCGMemOp memop = get_memop(oi);
#ifdef CONFIG_SOFTMMU
unsigned memi = get_mmuidx(oi);
TCGMemOp s_bits = memop & MO_SIZE;
TCGReg addrz, param;
tcg_insn_unit *func;
tcg_insn_unit *label_ptr;
addrz = tcg_out_tlb_load(s, addr, memi, s_bits,
addrz = tcg_out_tlb_load(s, addr, memi, memop & MO_SIZE,
offsetof(CPUTLBEntry, addr_write));
/* The fast path is exactly one insn. Thus we can perform the entire
@ -1172,7 +1171,8 @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg data, TCGReg addr,
tcg_out_bpcc0(s, COND_E, BPCC_A | BPCC_PT
| (TARGET_LONG_BITS == 64 ? BPCC_XCC : BPCC_ICC), 0);
/* delay slot */
tcg_out_ldst_rr(s, data, addrz, TCG_REG_O1, qemu_st_opc[memop]);
tcg_out_ldst_rr(s, data, addrz, TCG_REG_O1,
qemu_st_opc[memop & (MO_BSWAP | MO_SIZE)]);
/* TLB Miss. */
@ -1182,13 +1182,13 @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg data, TCGReg addr,
param++;
}
tcg_out_mov(s, TCG_TYPE_REG, param++, addr);
if (!SPARC64 && s_bits == MO_64) {
if (!SPARC64 && (memop & MO_SIZE) == MO_64) {
/* Skip the high-part; we'll perform the extract in the trampoline. */
param++;
}
tcg_out_mov(s, TCG_TYPE_REG, param++, data);
func = qemu_st_trampoline[memop];
func = qemu_st_trampoline[memop & (MO_BSWAP | MO_SIZE)];
assert(func != NULL);
tcg_out_call_nodelay(s, func);
/* delay slot */
@ -1202,7 +1202,7 @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg data, TCGReg addr,
}
tcg_out_ldst_rr(s, data, addr,
(GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_G0),
qemu_st_opc[memop]);
qemu_st_opc[memop & (MO_BSWAP | MO_SIZE)]);
#endif /* CONFIG_SOFTMMU */
}