return &arch_cpu->neg;
}
+/**
+ * env_tlb(env)
+ * @env: The architecture environment
+ *
+ * Return the CPUTLB state associated with the environment.
+ */
+static inline CPUTLB *env_tlb(CPUArchState *env)
+{
+ return &env_neg(env)->tlb;
+}
+
#endif /* CPU_ALL_H */
/*
* Data elements that are per MMU mode, accessed by the fast path.
+ * The structure is aligned to aid loading the pair with one insn.
*/
typedef struct CPUTLBDescFast {
/* Contains (n_entries - 1) << CPU_TLB_ENTRY_BITS */
uintptr_t mask;
/* The array of tlb entries itself. */
CPUTLBEntry *table;
-} CPUTLBDescFast;
+} CPUTLBDescFast QEMU_ALIGNED(2 * sizeof(void *));
/*
* Data elements that are shared between all MMU modes.
/*
* The entire softmmu tlb, for all MMU modes.
* The meaning of each of the MMU modes is defined in the target code.
+ * Since this is placed within CPUNegativeOffsetState, the smallest
+ * negative offsets are at the end of the struct.
*/
typedef struct CPUTLB {
- CPUTLBDescFast f[NB_MMU_MODES];
- CPUTLBDesc d[NB_MMU_MODES];
CPUTLBCommon c;
+ CPUTLBDesc d[NB_MMU_MODES];
+ CPUTLBDescFast f[NB_MMU_MODES];
} CPUTLB;
-/* There are target-specific members named "tlb". This is temporary. */
-#define CPU_COMMON CPUTLB tlb_;
-#define env_tlb(ENV) (&(ENV)->tlb_)
+/* This will be used by TCG backends to compute offsets. */
+#define TLB_MASK_TABLE_OFS(IDX) \
+ ((int)offsetof(ArchCPU, neg.tlb.f[IDX]) - (int)offsetof(ArchCPU, env))
#else
-#define CPU_COMMON /* Nothing */
+typedef struct CPUTLB { } CPUTLB;
#endif /* !CONFIG_USER_ONLY && CONFIG_TCG */
+#define CPU_COMMON /* Nothing */
+
/*
* This structure must be placed in ArchCPU immedately
* before CPUArchState, as a field named "neg".
*/
typedef struct CPUNegativeOffsetState {
+ CPUTLB tlb;
IcountDecr icount_decr;
} CPUNegativeOffsetState;
label->label_ptr[0] = label_ptr;
}
-/* We expect to use a 24-bit unsigned offset from ENV. */
-QEMU_BUILD_BUG_ON(offsetof(CPUArchState, tlb_.f[NB_MMU_MODES - 1].table)
- > 0xffffff);
+/* We expect to use a 7-bit scaled negative offset from ENV. */
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) > 0);
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) < -512);
/* Load and compare a TLB entry, emitting the conditional jump to the
slow path for the failure case, which will be patched later when finalizing
tcg_insn_unit **label_ptr, int mem_index,
bool is_read)
{
- int mask_ofs = offsetof(CPUArchState, tlb_.f[mem_index].mask);
- int table_ofs = offsetof(CPUArchState, tlb_.f[mem_index].table);
+ int fast_ofs = TLB_MASK_TABLE_OFS(mem_index);
+ int mask_ofs = fast_ofs + offsetof(CPUTLBDescFast, mask);
+ int table_ofs = fast_ofs + offsetof(CPUTLBDescFast, table);
unsigned a_bits = get_alignment_bits(opc);
unsigned s_bits = opc & MO_SIZE;
unsigned a_mask = (1u << a_bits) - 1;
TCGType mask_type;
uint64_t compare_mask;
- if (table_ofs > 0xfff) {
- int table_hi = table_ofs & ~0xfff;
- int mask_hi = mask_ofs & ~0xfff;
-
- table_base = TCG_REG_X1;
- if (mask_hi == table_hi) {
- mask_base = table_base;
- } else if (mask_hi) {
- mask_base = TCG_REG_X0;
- tcg_out_insn(s, 3401, ADDI, TCG_TYPE_I64,
- mask_base, TCG_AREG0, mask_hi);
- }
- tcg_out_insn(s, 3401, ADDI, TCG_TYPE_I64,
- table_base, TCG_AREG0, table_hi);
- mask_ofs -= mask_hi;
- table_ofs -= table_hi;
- }
-
mask_type = (TARGET_PAGE_BITS + CPU_TLB_DYN_MAX_BITS > 32
? TCG_TYPE_I64 : TCG_TYPE_I32);
#define TLB_SHIFT (CPU_TLB_ENTRY_BITS + CPU_TLB_BITS)
-/* We expect to use a 20-bit unsigned offset from ENV. */
-QEMU_BUILD_BUG_ON(offsetof(CPUArchState, tlb_.f[NB_MMU_MODES - 1].table)
- > 0xfffff);
+/* We expect to use an 9-bit sign-magnitude negative offset from ENV. */
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) > 0);
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) < -256);
/* Load and compare a TLB entry, leaving the flags set. Returns the register
containing the addend of the tlb entry. Clobbers R0, R1, R2, TMP. */
{
int cmp_off = (is_load ? offsetof(CPUTLBEntry, addr_read)
: offsetof(CPUTLBEntry, addr_write));
- int mask_off = offsetof(CPUArchState, tlb_.f[mem_index].mask);
- int table_off = offsetof(CPUArchState, tlb_.f[mem_index].table);
- TCGReg mask_base = TCG_AREG0, table_base = TCG_AREG0;
+ int fast_off = TLB_MASK_TABLE_OFS(mem_index);
+ int mask_off = fast_off + offsetof(CPUTLBDescFast, mask);
+ int table_off = fast_off + offsetof(CPUTLBDescFast, table);
unsigned s_bits = opc & MO_SIZE;
unsigned a_bits = get_alignment_bits(opc);
- if (table_off > 0xfff) {
- int mask_hi = mask_off & ~0xfff;
- int table_hi = table_off & ~0xfff;
- int rot;
-
- table_base = TCG_REG_R2;
- if (mask_hi == table_hi) {
- mask_base = table_base;
- } else if (mask_hi) {
- mask_base = TCG_REG_TMP;
- rot = encode_imm(mask_hi);
- assert(rot >= 0);
- tcg_out_dat_imm(s, COND_AL, ARITH_ADD, mask_base, TCG_AREG0,
- rotl(mask_hi, rot) | (rot << 7));
- }
- rot = encode_imm(table_hi);
- assert(rot >= 0);
- tcg_out_dat_imm(s, COND_AL, ARITH_ADD, table_base, TCG_AREG0,
- rotl(table_hi, rot) | (rot << 7));
-
- mask_off -= mask_hi;
- table_off -= table_hi;
- }
-
/* Load tlb_mask[mmu_idx] and tlb_table[mmu_idx]. */
- tcg_out_ld(s, TCG_TYPE_I32, TCG_REG_TMP, mask_base, mask_off);
- tcg_out_ld(s, TCG_TYPE_I32, TCG_REG_R2, table_base, table_off);
+ tcg_out_ld(s, TCG_TYPE_I32, TCG_REG_TMP, TCG_AREG0, mask_off);
+ tcg_out_ld(s, TCG_TYPE_I32, TCG_REG_R2, TCG_AREG0, table_off);
/* Extract the tlb index from the address into TMP. */
tcg_out_dat_reg(s, COND_AL, ARITH_AND, TCG_REG_TMP, TCG_REG_TMP, addrlo,
TARGET_PAGE_BITS - CPU_TLB_ENTRY_BITS);
tcg_out_modrm_offset(s, OPC_AND_GvEv + trexw, r0, TCG_AREG0,
- offsetof(CPUArchState, tlb_.f[mem_index].mask));
+ TLB_MASK_TABLE_OFS(mem_index) +
+ offsetof(CPUTLBDescFast, mask));
tcg_out_modrm_offset(s, OPC_ADD_GvEv + hrexw, r0, TCG_AREG0,
- offsetof(CPUArchState, tlb_.f[mem_index].table));
+ TLB_MASK_TABLE_OFS(mem_index) +
+ offsetof(CPUTLBDescFast, table));
/* If the required alignment is at least as large as the access, simply
copy the address and mask. For lesser alignments, check that we don't
return i;
}
+/* We expect to use a 16-bit negative offset from ENV. */
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) > 0);
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) < -32768);
+
/*
* Perform the tlb comparison operation.
* The complete host address is placed in BASE.
unsigned s_bits = opc & MO_SIZE;
unsigned a_bits = get_alignment_bits(opc);
int mem_index = get_mmuidx(oi);
- int mask_off = offsetof(CPUArchState, tlb_.f[mem_index].mask);
- int table_off = offsetof(CPUArchState, tlb_.f[mem_index].mask);
+ int fast_off = TLB_MASK_TABLE_OFS(mem_index);
+ int mask_off = fast_off + offsetof(CPUTLBDescFast, mask);
+ int table_off = fast_off + offsetof(CPUTLBDescFast, table);
int add_off = offsetof(CPUTLBEntry, addend);
int cmp_off = (is_load ? offsetof(CPUTLBEntry, addr_read)
: offsetof(CPUTLBEntry, addr_write));
- TCGReg mask_base = TCG_AREG0, table_base = TCG_AREG0;
target_ulong mask;
- if (table_off > 0x7fff) {
- int mask_hi = mask_off - (int16_t)mask_off;
- int table_hi = table_off - (int16_t)table_off;
-
- table_base = TCG_TMP1;
- if (likely(mask_hi == table_hi)) {
- mask_base = table_base;
- tcg_out_opc_imm(s, OPC_LUI, mask_base, TCG_REG_ZERO, mask_hi >> 16);
- tcg_out_opc_reg(s, ALIAS_PADD, mask_base, mask_base, TCG_AREG0);
- mask_off -= mask_hi;
- table_off -= mask_hi;
- } else {
- if (mask_hi != 0) {
- mask_base = TCG_TMP0;
- tcg_out_opc_imm(s, OPC_LUI,
- mask_base, TCG_REG_ZERO, mask_hi >> 16);
- tcg_out_opc_reg(s, ALIAS_PADD,
- mask_base, mask_base, TCG_AREG0);
- }
- table_off -= mask_off;
- mask_off -= mask_hi;
- tcg_out_opc_imm(s, ALIAS_PADDI, table_base, mask_base, mask_off);
- }
- }
-
/* Load tlb_mask[mmu_idx] and tlb_table[mmu_idx]. */
- tcg_out_ld(s, TCG_TYPE_PTR, TCG_TMP0, mask_base, mask_off);
- tcg_out_ld(s, TCG_TYPE_PTR, TCG_TMP1, table_base, table_off);
+ tcg_out_ld(s, TCG_TYPE_PTR, TCG_TMP0, TCG_AREG0, mask_off);
+ tcg_out_ld(s, TCG_TYPE_PTR, TCG_TMP1, TCG_AREG0, table_off);
/* Extract the TLB index from the address into TMP3. */
tcg_out_opc_sa(s, ALIAS_TSRL, TCG_TMP3, addrl,
[MO_BEQ] = helper_be_stq_mmu,
};
+/* We expect to use a 16-bit negative offset from ENV. */
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) > 0);
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) < -32768);
+
/* Perform the TLB load and compare. Places the result of the comparison
in CR7, loads the addend of the TLB into R3, and returns the register
containing the guest address (zero-extended into R4). Clobbers R0 and R2. */
= (is_read
? offsetof(CPUTLBEntry, addr_read)
: offsetof(CPUTLBEntry, addr_write));
- int mask_off = offsetof(CPUArchState, tlb_.f[mem_index].mask);
- int table_off = offsetof(CPUArchState, tlb_.f[mem_index].table);
- TCGReg mask_base = TCG_AREG0, table_base = TCG_AREG0;
+ int fast_off = TLB_MASK_TABLE_OFS(mem_index);
+ int mask_off = fast_off + offsetof(CPUTLBDescFast, mask);
+ int table_off = fast_off + offsetof(CPUTLBDescFast, table);
unsigned s_bits = opc & MO_SIZE;
unsigned a_bits = get_alignment_bits(opc);
- if (table_off > 0x7fff) {
- int mask_hi = mask_off - (int16_t)mask_off;
- int table_hi = table_off - (int16_t)table_off;
-
- table_base = TCG_REG_R4;
- if (mask_hi == table_hi) {
- mask_base = table_base;
- } else if (mask_hi) {
- mask_base = TCG_REG_R3;
- tcg_out32(s, ADDIS | TAI(mask_base, TCG_AREG0, mask_hi >> 16));
- }
- tcg_out32(s, ADDIS | TAI(table_base, TCG_AREG0, table_hi >> 16));
- mask_off -= mask_hi;
- table_off -= table_hi;
- }
-
/* Load tlb_mask[mmu_idx] and tlb_table[mmu_idx]. */
- tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_R3, mask_base, mask_off);
- tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_R4, table_base, table_off);
+ tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_R3, TCG_AREG0, mask_off);
+ tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_R4, TCG_AREG0, table_off);
/* Extract the page index, shifted into place for tlb index. */
if (TCG_TARGET_REG_BITS == 32) {
/* We don't support oversize guests */
QEMU_BUILD_BUG_ON(TCG_TARGET_REG_BITS < TARGET_LONG_BITS);
+/* We expect to use a 12-bit negative offset from ENV. */
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) > 0);
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) < -(1 << 11));
+
static void tcg_out_tlb_load(TCGContext *s, TCGReg addrl,
TCGReg addrh, TCGMemOpIdx oi,
tcg_insn_unit **label_ptr, bool is_load)
unsigned a_bits = get_alignment_bits(opc);
tcg_target_long compare_mask;
int mem_index = get_mmuidx(oi);
- int mask_off, table_off;
+ int fast_ofs = TLB_MASK_TABLE_OFS(mem_index);
+ int mask_ofs = fast_ofs + offsetof(CPUTLBDescFast, mask);
+ int table_ofs = fast_ofs + offsetof(CPUTLBDescFast, table);
TCGReg mask_base = TCG_AREG0, table_base = TCG_AREG0;
- mask_off = offsetof(CPUArchState, tlb_.f[mem_index].mask);
- table_off = offsetof(CPUArchState, tlb_.f[mem_index].table);
- if (table_off > 0x7ff) {
- int mask_hi = mask_off - sextreg(mask_off, 0, 12);
- int table_hi = table_off - sextreg(table_off, 0, 12);
-
- if (likely(mask_hi == table_hi)) {
- mask_base = table_base = TCG_REG_TMP1;
- tcg_out_opc_upper(s, OPC_LUI, mask_base, mask_hi);
- tcg_out_opc_reg(s, OPC_ADD, mask_base, mask_base, TCG_AREG0);
- mask_off -= mask_hi;
- table_off -= mask_hi;
- } else {
- mask_base = TCG_REG_TMP0;
- table_base = TCG_REG_TMP1;
- tcg_out_opc_upper(s, OPC_LUI, mask_base, mask_hi);
- tcg_out_opc_reg(s, OPC_ADD, mask_base, mask_base, TCG_AREG0);
- table_off -= mask_off;
- mask_off -= mask_hi;
- tcg_out_opc_imm(s, OPC_ADDI, table_base, mask_base, mask_off);
- }
- }
-
tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP0, mask_base, mask_off);
tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP1, table_base, table_off);
#if defined(CONFIG_SOFTMMU)
#include "tcg-ldst.inc.c"
-/* We're expecting to use a 20-bit signed offset on the tlb memory ops. */
-QEMU_BUILD_BUG_ON(offsetof(CPUArchState, tlb_.f[NB_MMU_MODES - 1].table)
- > 0x7ffff);
+/* We're expecting to use a 20-bit negative offset on the tlb memory ops. */
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) > 0);
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) < -(1 << 19));
/* Load and compare a TLB entry, leaving the flags set. Loads the TLB
addend into R2. Returns a register with the santitized guest address. */
unsigned a_bits = get_alignment_bits(opc);
unsigned s_mask = (1 << s_bits) - 1;
unsigned a_mask = (1 << a_bits) - 1;
- int mask_off = offsetof(CPUArchState, tlb_.f[mem_index].mask);
- int table_off = offsetof(CPUArchState, tlb_.f[mem_index].table);
+ int fast_off = TLB_MASK_TABLE_OFS(mem_index);
+ int mask_off = fast_off + offsetof(CPUTLBDescFast, mask);
+ int table_off = fast_off + offsetof(CPUTLBDescFast, table);
int ofs, a_off;
uint64_t tlb_mask;
}
#if defined(CONFIG_SOFTMMU)
+
+/* We expect to use a 13-bit negative offset from ENV. */
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) > 0);
+QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) < -(1 << 12));
+
/* Perform the TLB load and compare.
Inputs:
static TCGReg tcg_out_tlb_load(TCGContext *s, TCGReg addr, int mem_index,
TCGMemOp opc, int which)
{
- int mask_off = offsetof(CPUArchState, tlb_.f[mem_index].mask);
- int table_off = offsetof(CPUArchState, tlb_.f[mem_index].table);
- TCGReg base = TCG_AREG0;
+ int fast_off = TLB_MASK_TABLE_OFS(mem_index);
+ int mask_off = fast_off + offsetof(CPUTLBDescFast, mask);
+ int table_off = fast_off + offsetof(CPUTLBDescFast, table);
const TCGReg r0 = TCG_REG_O0;
const TCGReg r1 = TCG_REG_O1;
const TCGReg r2 = TCG_REG_O2;
unsigned a_bits = get_alignment_bits(opc);
tcg_target_long compare_mask;
- if (!check_fit_i32(table_off, 13)) {
- int table_hi;
-
- base = r1;
- if (table_off <= 2 * 0xfff) {
- table_hi = 0xfff;
- tcg_out_arithi(s, base, TCG_AREG0, table_hi, ARITH_ADD);
- } else {
- table_hi = table_off & ~0x3ff;
- tcg_out_sethi(s, base, table_hi);
- tcg_out_arith(s, base, TCG_AREG0, base, ARITH_ADD);
- }
- mask_off -= table_hi;
- table_off -= table_hi;
- tcg_debug_assert(check_fit_i32(mask_off, 13));
- }
-
/* Load tlb_mask[mmu_idx] and tlb_table[mmu_idx]. */
- tcg_out_ld(s, TCG_TYPE_PTR, r0, base, mask_off);
- tcg_out_ld(s, TCG_TYPE_PTR, r1, base, table_off);
+ tcg_out_ld(s, TCG_TYPE_PTR, r0, TCG_AREG0, mask_off);
+ tcg_out_ld(s, TCG_TYPE_PTR, r1, TCG_AREG0, table_off);
/* Extract the page index, shifted into place for tlb index. */
tcg_out_arithi(s, r2, addr, TARGET_PAGE_BITS - CPU_TLB_ENTRY_BITS,