ruby--ruby/ujit_compile.c

#include <assert.h>
#include "insns.inc"
#include "internal.h"
#include "vm_core.h"
#include "vm_sync.h"
#include "vm_callinfo.h"
#include "builtin.h"
#include "internal/compile.h"
#include "internal/class.h"
#include "insns_info.inc"
#include "ujit_compile.h"
#include "ujit_asm.h"
#include "ujit_utils.h"

// TODO: give ujit_examples.inc some more meaningful file name
// eg ujit_hook.h
#include "ujit_examples.inc"

#ifdef _WIN32
#define PLATFORM_SUPPORTED_P 0
#else
#define PLATFORM_SUPPORTED_P 1
#endif

#ifndef UJIT_CHECK_MODE
#define UJIT_CHECK_MODE 0
#endif

// >= 1: print when output code invalidation happens
// >= 2: dump list of instructions when regions compile
#ifndef UJIT_DUMP_MODE
#define UJIT_DUMP_MODE 0
#endif

bool rb_ujit_enabled;

// Hash table of encoded instructions
extern st_table *rb_encoded_insn_data;

// Code generation context
typedef struct ctx_struct
{
    // Current PC
    VALUE *pc;

    // Difference between the current stack pointer and actual stack top
    int32_t stack_diff;

    // The iseq that owns the region that is compiling
    const rb_iseq_t *iseq;

    // Index in the iseq of the opcode we are replacing
    size_t start_idx;

    // The start of the generated code
    uint8_t *code_ptr;

} ctx_t;

// MicroJIT code generation function signature
typedef bool (*codegen_fn)(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx);

// Map from YARV opcodes to code generation functions
static st_table *gen_fns;

// Code block into which we write machine code
static codeblock_t block;
static codeblock_t* cb = NULL;

// Code block into which we write out-of-line machine code
static codeblock_t outline_block;
static codeblock_t* ocb = NULL;

// Register MicroJIT receives the CFP and EC into
#define REG_CFP RDI
#define REG_EC RSI

// Register MicroJIT loads the SP into
#define REG_SP RDX

// Scratch registers used by MicroJIT
#define REG0 RAX
#define REG1 RCX
#define REG0_32 EAX
#define REG1_32 ECX

// Keep track of mapping from instructions to generated code
// See comment for rb_encoded_insn_data in iseq.c
static void
addr2insn_bookkeeping(void *code_ptr, int insn)
{
    const void * const *table = rb_vm_get_insns_address_table();
    const void * const translated_address = table[insn];
    st_data_t encoded_insn_data;
    if (st_lookup(rb_encoded_insn_data, (st_data_t)translated_address, &encoded_insn_data)) {
        st_insert(rb_encoded_insn_data, (st_data_t)code_ptr, encoded_insn_data);
    }
    else {
        rb_bug("ujit: failed to find info for original instruction while dealing with addr2insn");
    }
}

// GC root for interacting with the GC
struct ujit_root_struct {};

// Map cme_or_cc => [[iseq, offset]]. An entry in the map means compiled code at iseq[offset]
// is only valid when cme_or_cc is valid
static st_table *method_lookup_dependency;

struct compiled_region_array {
    int32_t size;
    int32_t capa;
    struct compiled_region {
        const rb_iseq_t *iseq;
        size_t start_idx;
        uint8_t *code;
    } data[];
};

// Add an element to a region array, or allocate a new region array.
static struct compiled_region_array *
add_compiled_region(struct compiled_region_array *array, const rb_iseq_t *iseq, size_t start_idx, uint8_t *code)
{
    if (!array) {
        // Allocate a brand new array with space for one
        array = malloc(sizeof(*array) + sizeof(struct compiled_region));
        if (!array) {
            return NULL;
        }
        array->size = 0;
        array->capa = 1;
    }
    if (array->size == INT32_MAX) {
        return NULL;
    }
    // Check if the region is already present
    for (int32_t i = 0; i < array->size; i++) {
        if (array->data[i].iseq == iseq && array->data[i].start_idx == start_idx) {
            return array;
        }
    }
    if (array->size + 1 > array->capa) {
        // Double the array's capacity.
        int64_t double_capa = ((int64_t)array->capa) * 2;
        int32_t new_capa = (int32_t)double_capa;
        if (new_capa != double_capa) {
            return NULL;
        }
        array = realloc(array, sizeof(*array) + new_capa * sizeof(struct compiled_region));
        if (array == NULL) {
            return NULL;
        }
        array->capa = new_capa;
    }

    int32_t size = array->size;
    array->data[size].iseq = iseq;
    array->data[size].start_idx = start_idx;
    array->data[size].code = code;
    array->size++;
    return array;
}

static int
add_lookup_dependency_i(st_data_t *key, st_data_t *value, st_data_t data, int existing)
{
    ctx_t *ctx = (ctx_t *)data;
    struct compiled_region_array *regions = NULL;
    if (existing) {
        regions = (struct compiled_region_array *)*value;
    }
    regions = add_compiled_region(regions, ctx->iseq, ctx->start_idx, ctx->code_ptr);
    if (!regions) {
        rb_bug("ujit: failed to add method lookup dependency"); // TODO: we could bail out of compiling instead
    }
    *value = (st_data_t)regions;
    return ST_CONTINUE;
}

// Store info to remember that the currently compiling region is only valid while cme and cc and valid.
static void
ujit_assume_method_lookup_stable(const struct rb_callcache *cc, const rb_callable_method_entry_t *cme,  ctx_t *ctx)
{
    st_update(method_lookup_dependency, (st_data_t)cme, add_lookup_dependency_i, (st_data_t)ctx);
    st_update(method_lookup_dependency, (st_data_t)cc, add_lookup_dependency_i, (st_data_t)ctx);
    // FIXME: This is a leak! When either the cme or the cc become invalid, the other also needs to go
}

static int
ujit_root_mark_i(st_data_t k, st_data_t v, st_data_t ignore)
{
    // FIXME: This leaks everything that end up in the dependency table!
    // One way to deal with this is with weak references...
    rb_gc_mark((VALUE)k);
    struct compiled_region_array *regions = (void *)v;
    for (int32_t i = 0; i < regions->size; i++) {
        rb_gc_mark((VALUE)regions->data[i].iseq);
    }

    return ST_CONTINUE;
}

// GC callback during mark phase
static void
ujit_root_mark(void *ptr)
{
    if (method_lookup_dependency) {
        st_foreach(method_lookup_dependency, ujit_root_mark_i, 0);
    }
}

static void
ujit_root_free(void *ptr)
{
    // Do nothing. The root lives as long as the process.
}

static size_t
ujit_root_memsize(const void *ptr)
{
    // Count off-gc-heap allocation size of the dependency table
    return st_memsize(method_lookup_dependency); // TODO: more accurate accounting
}

// Custom type for interacting with the GC
// TODO: compaction support
// TODO: make this write barrier protected
static const rb_data_type_t ujit_root_type = {
    "ujit_root",
    {ujit_root_mark, ujit_root_free, ujit_root_memsize, },
    0, 0, RUBY_TYPED_FREE_IMMEDIATELY
};

static int
opcode_at_pc(const rb_iseq_t *iseq, const VALUE *pc)
{
    const VALUE at_pc = *pc;
    if (FL_TEST_RAW((VALUE)iseq, ISEQ_TRANSLATED)) {
        return rb_vm_insn_addr2opcode((const void *)at_pc);
    }
    else {
        return (int)at_pc;
    }
}

// Get the current instruction opcode from the context object
static int
ctx_get_opcode(ctx_t *ctx)
{
    return opcode_at_pc(ctx->iseq, ctx->pc);
}

// Get an instruction argument from the context object
static VALUE
ctx_get_arg(ctx_t* ctx, size_t arg_idx)
{
    assert (arg_idx + 1 < insn_len(ctx_get_opcode(ctx)));
    return *(ctx->pc + arg_idx + 1);
}

/*
Get an operand for the adjusted stack pointer address
*/
static x86opnd_t
ctx_sp_opnd(ctx_t* ctx, int32_t offset_bytes)
{
    int32_t offset = (ctx->stack_diff) * 8 + offset_bytes;
    return mem_opnd(64, REG_SP, offset);
}

/*
Make space on the stack for N values
Return a pointer to the new stack top
*/
static x86opnd_t
ctx_stack_push(ctx_t* ctx, size_t n)
{
    ctx->stack_diff += n;

    // SP points just above the topmost value
    int32_t offset = (ctx->stack_diff - 1) * 8;
    return mem_opnd(64, REG_SP, offset);
}

/*
Pop N values off the stack
Return a pointer to the stack top before the pop operation
*/
static x86opnd_t
ctx_stack_pop(ctx_t* ctx, size_t n)
{
    // SP points just above the topmost value
    int32_t offset = (ctx->stack_diff - 1) * 8;
    x86opnd_t top = mem_opnd(64, REG_SP, offset);

    ctx->stack_diff -= n;

    return top;
}

static x86opnd_t
ctx_stack_opnd(ctx_t* ctx, int32_t idx)
{
    // SP points just above the topmost value
    int32_t offset = (ctx->stack_diff - 1 - idx) * 8;
    x86opnd_t opnd = mem_opnd(64, REG_SP, offset);

    return opnd;
}

// Ruby instruction entry
static void
ujit_gen_entry(codeblock_t* cb)
{
    for (size_t i = 0; i < sizeof(ujit_with_ec_pre_call_bytes); ++i)
        cb_write_byte(cb, ujit_with_ec_pre_call_bytes[i]);
}

/**
Generate an inline exit to return to the interpreter
*/
static void
ujit_gen_exit(codeblock_t* cb, ctx_t* ctx, VALUE* exit_pc)
{
    // Write the adjusted SP back into the CFP
    if (ctx->stack_diff != 0)
    {
        x86opnd_t stack_pointer = ctx_sp_opnd(ctx, 0);
        lea(cb, REG_SP, stack_pointer);
        mov(cb, member_opnd(REG_CFP, rb_control_frame_t, sp), REG_SP);
    }

    // Directly return the next PC, which is a constant
    mov(cb, RAX, const_ptr_opnd(exit_pc));
    mov(cb, member_opnd(REG_CFP, rb_control_frame_t, pc), RAX);

    // Write the post call bytes
    for (size_t i = 0; i < sizeof(ujit_with_ec_post_call_bytes); ++i)
        cb_write_byte(cb, ujit_with_ec_post_call_bytes[i]);
}

/**
Generate an out-of-line exit to return to the interpreter
*/
static uint8_t *
ujit_side_exit(codeblock_t* cb, ctx_t* ctx, VALUE* exit_pc)
{
    uint8_t* code_ptr = cb_get_ptr(cb, cb->write_pos);

    // Table mapping opcodes to interpreter handlers
    const void * const *table = rb_vm_get_insns_address_table();

    // Write back the old instruction at the entry PC
    // To deotimize the code block this instruction belongs to
    VALUE* entry_pc = &ctx->iseq->body->iseq_encoded[ctx->start_idx];
    int entry_opcode = opcode_at_pc(ctx->iseq, entry_pc);
    void* entry_instr = (void*)table[entry_opcode];
    mov(cb, RAX, const_ptr_opnd(entry_pc));
    mov(cb, RCX, const_ptr_opnd(entry_instr));
    mov(cb, mem_opnd(64, RAX, 0), RCX);

    // Write back the old instruction at the exit PC
    // Otherwise the interpreter may jump right back to the
    // JITted code we're trying to exit
    int exit_opcode = opcode_at_pc(ctx->iseq, exit_pc);
    void* exit_instr = (void*)table[exit_opcode];
    mov(cb, RAX, const_ptr_opnd(exit_pc));
    mov(cb, RCX, const_ptr_opnd(exit_instr));
    mov(cb, mem_opnd(64, RAX, 0), RCX);

    // Generate the code to exit to the interpreters
    ujit_gen_exit(cb, ctx, exit_pc);

    return code_ptr;
}

/*
Compile a sequence of bytecode instructions starting at `insn_idx`.
Return the index to the first instruction not compiled in the sequence
through `next_ujit_idx`. Return `NULL` in case compilation fails.
*/
static uint8_t *
ujit_compile_insn(const rb_iseq_t *iseq, unsigned int insn_idx, unsigned int *next_ujit_idx)
{
    assert (cb != NULL);
    unsigned first_insn_idx = insn_idx;
    VALUE *encoded = iseq->body->iseq_encoded;

    // NOTE: if we are ever deployed in production, we
    // should probably just log an error and return NULL here,
    // so we can fail more gracefully
    if (cb->write_pos + 1024 >= cb->mem_size) {
        rb_bug("out of executable memory");
    }
    if (ocb->write_pos + 1024 >= ocb->mem_size) {
        rb_bug("out of executable memory (outlined block)");
    }

    // Align the current write positon to cache line boundaries
    cb_align_pos(cb, 64);

    // Get a pointer to the current write position in the code block
    uint8_t *code_ptr = &cb->mem_block[cb->write_pos];
    //printf("write pos: %ld\n", cb->write_pos);

    // Get the first opcode in the sequence
    int first_opcode = opcode_at_pc(iseq, &encoded[insn_idx]);

    // Create codegen context
    ctx_t ctx;
    ctx.pc = NULL;
    ctx.stack_diff = 0;
    ctx.iseq = iseq;
    ctx.code_ptr = code_ptr;
    ctx.start_idx = insn_idx;

    // For each instruction to compile
    unsigned num_instrs = 0;
    for (;;) {
        // Set the current PC
        ctx.pc = &encoded[insn_idx];

        // Get the current opcode
        int opcode = ctx_get_opcode(&ctx);

        // Lookup the codegen function for this instruction
        st_data_t st_gen_fn;
        if (!rb_st_lookup(gen_fns, opcode, &st_gen_fn)) {
            //print_int(cb, imm_opnd(num_instrs));
            //print_str(cb, insn_name(opcode));
            break;
        }

        // Write the pre call bytes before the first instruction
        if (num_instrs == 0) {
            ujit_gen_entry(cb);

            // Load the current SP from the CFP into REG_SP
            mov(cb, REG_SP, member_opnd(REG_CFP, rb_control_frame_t, sp));
        }

        // Call the code generation function
        codegen_fn gen_fn = (codegen_fn)st_gen_fn;
        if (!gen_fn(cb, ocb, &ctx)) {
            break;
        }

    	// Move to the next instruction
        insn_idx += insn_len(opcode);
        num_instrs++;

        // Ensure we only have one send per region. Our code invalidation mechanism can't
        // invalidate running code and one send could invalidate the other if we had
        // multiple in the same region.
        if (opcode == BIN(opt_send_without_block)) {
            break;
        }
    }

    // Let the caller know how many instructions ujit compiled
    *next_ujit_idx = insn_idx;

    // If no instructions were compiled
    if (num_instrs == 0) {
        return NULL;
    }

    // Generate code to exit to the interpreter
    ujit_gen_exit(cb, &ctx, &encoded[*next_ujit_idx]);

    addr2insn_bookkeeping(code_ptr, first_opcode);

    if (UJIT_DUMP_MODE >= 2) {
        // Dump list of compiled instrutions
        fprintf(stderr, "Compiled the following for iseq=%p:\n", (void *)iseq);
        VALUE *pc = &encoded[first_insn_idx];
        VALUE *end_pc = &encoded[*next_ujit_idx];
        while (pc < end_pc) {
            int opcode = opcode_at_pc(iseq, pc);
            fprintf(stderr, "  %04td %s\n", pc - encoded, insn_name(opcode));
            pc += insn_len(opcode);
        }
    }

    return code_ptr;
}

static bool
gen_dup(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    x86opnd_t dup_val = ctx_stack_pop(ctx, 1);
    x86opnd_t loc0 = ctx_stack_push(ctx, 1);
    x86opnd_t loc1 = ctx_stack_push(ctx, 1);
    mov(cb, RAX, dup_val);
    mov(cb, loc0, RAX);
    mov(cb, loc1, RAX);
    return true;
}

static bool
gen_nop(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    // Do nothing
    return true;
}

static bool
gen_pop(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    // Decrement SP
    ctx_stack_pop(ctx, 1);
    return true;
}

static bool
gen_putnil(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    // Write constant at SP
    x86opnd_t stack_top = ctx_stack_push(ctx, 1);
    mov(cb, stack_top, imm_opnd(Qnil));
    return true;
}

static bool
gen_putobject(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    // Load the argument from the bytecode sequence.
    // We need to do this as the argument can chanage due to GC compaction.
    x86opnd_t pc_imm = const_ptr_opnd((void*)ctx->pc);
    mov(cb, RAX, pc_imm);
    mov(cb, RAX, mem_opnd(64, RAX, 8)); // One after the opcode

    // Write argument at SP
    x86opnd_t stack_top = ctx_stack_push(ctx, 1);
    mov(cb, stack_top, RAX);

    return true;
}

static bool
gen_putobject_int2fix(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    int opcode = ctx_get_opcode(ctx);
    int cst_val = (opcode == BIN(putobject_INT2FIX_0_))? 0:1;

    // Write constant at SP
    x86opnd_t stack_top = ctx_stack_push(ctx, 1);
    mov(cb, stack_top, imm_opnd(INT2FIX(cst_val)));

    return true;
}

static bool
gen_putself(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    // Load self from CFP
    mov(cb, RAX, member_opnd(REG_CFP, rb_control_frame_t, self));

    // Write it on the stack
    x86opnd_t stack_top = ctx_stack_push(ctx, 1);
    mov(cb, stack_top, RAX);

    return true;
}

static bool
gen_getlocal_wc0(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    // Load environment pointer EP from CFP
    mov(cb, REG0, member_opnd(REG_CFP, rb_control_frame_t, ep));

    // Compute the offset from BP to the local
    int32_t local_idx = (int32_t)ctx_get_arg(ctx, 0);
    const int32_t offs = -8 * local_idx;

    // Load the local from the block
    mov(cb, REG0, mem_opnd(64, REG0, offs));

    // Write the local at SP
    x86opnd_t stack_top = ctx_stack_push(ctx, 1);
    mov(cb, stack_top, REG0);

    return true;
}

static bool
gen_setlocal_wc0(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    /*
    vm_env_write(const VALUE *ep, int index, VALUE v)
    {
        VALUE flags = ep[VM_ENV_DATA_INDEX_FLAGS];
        if (LIKELY((flags & VM_ENV_FLAG_WB_REQUIRED) == 0)) {
    	VM_STACK_ENV_WRITE(ep, index, v);
        }
        else {
    	vm_env_write_slowpath(ep, index, v);
        }
    }
    */

    // Load environment pointer EP from CFP
    mov(cb, REG0, member_opnd(REG_CFP, rb_control_frame_t, ep));

    // flags & VM_ENV_FLAG_WB_REQUIRED
    x86opnd_t flags_opnd = mem_opnd(64, REG0, 8 * VM_ENV_DATA_INDEX_FLAGS);
    test(cb, flags_opnd, imm_opnd(VM_ENV_FLAG_WB_REQUIRED));

    // Create a size-exit to fall back to the interpreter
    uint8_t* side_exit = ujit_side_exit(ocb, ctx, ctx->pc);

    // if (flags & VM_ENV_FLAG_WB_REQUIRED) != 0
    jnz_ptr(cb, side_exit);

    // Pop the value to write from the stack
    x86opnd_t stack_top = ctx_stack_pop(ctx, 1);
    mov(cb, REG1, stack_top);

    // Write the value at the environment pointer
    int32_t local_idx = (int32_t)ctx_get_arg(ctx, 0);
    const int32_t offs = -8 * local_idx;
    mov(cb, mem_opnd(64, REG0, offs), REG1);

    return true;
}

static bool
gen_getinstancevariable(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    IVC ic = (IVC)ctx_get_arg(ctx, 1);

    // Check that the inline cache has been set, slot index is known
    if (!ic->entry)
    {
        return false;
    }

    // If the class uses the default allocator, instances should all be T_OBJECT
    // NOTE: This assumes nobody changes the allocator of the class after allocation.
    //       Eventually, we can encode whether an object is T_OBJECT or not
    //       inside object shapes.
    if (rb_get_alloc_func(ic->entry->class_value) != rb_class_allocate_instance)
    {
        return false;
    }

    uint32_t ivar_index = ic->entry->index;

    // Create a size-exit to fall back to the interpreter
    uint8_t* side_exit = ujit_side_exit(ocb, ctx, ctx->pc);

    // Load self from CFP
    mov(cb, REG0, member_opnd(REG_CFP, rb_control_frame_t, self));

    // Check that the receiver is a heap object
    test(cb, REG0, imm_opnd(RUBY_IMMEDIATE_MASK));
    jnz_ptr(cb, side_exit);
    cmp(cb, REG0, imm_opnd(Qfalse));
    je_ptr(cb, side_exit);
    cmp(cb, REG0, imm_opnd(Qnil));
    je_ptr(cb, side_exit);

    // Bail if receiver class is different from compiled time call cache class
    x86opnd_t klass_opnd = mem_opnd(64, REG0, offsetof(struct RBasic, klass));
    mov(cb, REG1, klass_opnd);
    x86opnd_t serial_opnd = mem_opnd(64, REG1, offsetof(struct RClass, class_serial));
    cmp(cb, serial_opnd, imm_opnd(ic->entry->class_serial));
    jne_ptr(cb, side_exit);

    // Bail if the ivars are not on the extended table
    // See ROBJECT_IVPTR() from include/ruby/internal/core/robject.h
    x86opnd_t flags_opnd = member_opnd(REG0, struct RBasic, flags);
    test(cb, flags_opnd, imm_opnd(ROBJECT_EMBED));
    jnz_ptr(cb, side_exit);

    // Get a pointer to the extended table
    x86opnd_t tbl_opnd = mem_opnd(64, REG0, offsetof(struct RObject, as.heap.ivptr));
    mov(cb, REG0, tbl_opnd);

    // Read the ivar from the extended table
    x86opnd_t ivar_opnd = mem_opnd(64, REG0, sizeof(VALUE) * ivar_index);
    mov(cb, REG0, ivar_opnd);

    // Check that the ivar is not Qundef
    cmp(cb, REG0, imm_opnd(Qundef));
    je_ptr(cb, side_exit);

    // Push the ivar on the stack
    x86opnd_t out_opnd = ctx_stack_push(ctx, 1);
    mov(cb, out_opnd, REG0);

    return true;
}

static bool
gen_setinstancevariable(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    IVC ic = (IVC)ctx_get_arg(ctx, 1);

    // Check that the inline cache has been set, slot index is known
    if (!ic->entry)
    {
        return false;
    }

    // If the class uses the default allocator, instances should all be T_OBJECT
    // NOTE: This assumes nobody changes the allocator of the class after allocation.
    //       Eventually, we can encode whether an object is T_OBJECT or not
    //       inside object shapes.
    if (rb_get_alloc_func(ic->entry->class_value) != rb_class_allocate_instance)
    {
        return false;
    }

    uint32_t ivar_index = ic->entry->index;

    // Create a size-exit to fall back to the interpreter
    uint8_t* side_exit = ujit_side_exit(ocb, ctx, ctx->pc);

    // Load self from CFP
    mov(cb, REG0, member_opnd(REG_CFP, rb_control_frame_t, self));

    // Check that the receiver is a heap object
    test(cb, REG0, imm_opnd(RUBY_IMMEDIATE_MASK));
    jnz_ptr(cb, side_exit);
    cmp(cb, REG0, imm_opnd(Qfalse));
    je_ptr(cb, side_exit);
    cmp(cb, REG0, imm_opnd(Qnil));
    je_ptr(cb, side_exit);

    // Bail if receiver class is different from compiled time call cache class
    x86opnd_t klass_opnd = mem_opnd(64, REG0, offsetof(struct RBasic, klass));
    mov(cb, REG1, klass_opnd);
    x86opnd_t serial_opnd = mem_opnd(64, REG1, offsetof(struct RClass, class_serial));
    cmp(cb, serial_opnd, imm_opnd(ic->entry->class_serial));
    jne_ptr(cb, side_exit);

    // Bail if the ivars are not on the extended table
    // See ROBJECT_IVPTR() from include/ruby/internal/core/robject.h
    x86opnd_t flags_opnd = member_opnd(REG0, struct RBasic, flags);
    test(cb, flags_opnd, imm_opnd(ROBJECT_EMBED));
    jnz_ptr(cb, side_exit);

    // If we can't guarantee that the extended table is big enoughg
    if (ivar_index >= ROBJECT_EMBED_LEN_MAX + 1)
    {
        // Check that the slot is inside the extended table (num_slots > index)
        x86opnd_t num_slots = mem_opnd(32, REG0, offsetof(struct RObject, as.heap.numiv));
        cmp(cb, num_slots, imm_opnd(ivar_index));
        jle_ptr(cb, side_exit);
    }

    // Get a pointer to the extended table
    x86opnd_t tbl_opnd = mem_opnd(64, REG0, offsetof(struct RObject, as.heap.ivptr));
    mov(cb, REG0, tbl_opnd);

    // Pop the value to write from the stack
    x86opnd_t stack_top = ctx_stack_pop(ctx, 1);
    mov(cb, REG1, stack_top);

    // Bail if this is a heap object, because this needs a write barrier
    test(cb, REG1, imm_opnd(RUBY_IMMEDIATE_MASK));
    jz_ptr(cb, side_exit);

    // Write the ivar to the extended table
    x86opnd_t ivar_opnd = mem_opnd(64, REG0, sizeof(VALUE) * ivar_index);
    mov(cb, ivar_opnd, REG1);

    return true;
}

static bool
gen_opt_minus(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    // Create a size-exit to fall back to the interpreter
    // Note: we generate the side-exit before popping operands from the stack
    uint8_t* side_exit = ujit_side_exit(ocb, ctx, ctx->pc);

    // TODO: make a helper function for this
    // Make sure that minus isn't redefined for integers
    mov(cb, RAX, const_ptr_opnd(ruby_current_vm_ptr));
    test(
        cb,
        member_opnd_idx(RAX, rb_vm_t, redefined_flag, BOP_MINUS),
        imm_opnd(INTEGER_REDEFINED_OP_FLAG)
    );
    jnz_ptr(cb, side_exit);

    // Get the operands and destination from the stack
    x86opnd_t arg1 = ctx_stack_pop(ctx, 1);
    x86opnd_t arg0 = ctx_stack_pop(ctx, 1);

    // If not fixnums, fall back
    test(cb, arg0, imm_opnd(RUBY_FIXNUM_FLAG));
    jz_ptr(cb, side_exit);
    test(cb, arg1, imm_opnd(RUBY_FIXNUM_FLAG));
    jz_ptr(cb, side_exit);

    // Subtract arg0 - arg1 and test for overflow
    mov(cb, RAX, arg0);
    sub(cb, RAX, arg1);
    jo_ptr(cb, side_exit);
    add(cb, RAX, imm_opnd(1));

    // Push the output on the stack
    x86opnd_t dst = ctx_stack_push(ctx, 1);
    mov(cb, dst, RAX);

    return true;
}

// Verify that calling with cd on receiver goes to callee
static void
check_cfunc_dispatch(VALUE receiver, struct rb_call_data *cd, void *callee, rb_callable_method_entry_t *compile_time_cme)
{
    if (METHOD_ENTRY_INVALIDATED(compile_time_cme)) {
        rb_bug("ujit: output code uses invalidated cme %p", (void *)compile_time_cme);
    }

    bool callee_correct = false;
    const rb_callable_method_entry_t *cme = rb_callable_method_entry(CLASS_OF(receiver), vm_ci_mid(cd->ci));
    if (cme->def->type == VM_METHOD_TYPE_CFUNC) {
        const rb_method_cfunc_t *cfunc = UNALIGNED_MEMBER_PTR(cme->def, body.cfunc);
        if ((void *)cfunc->func == callee) {
            callee_correct = true;
        }
    }
    if (!callee_correct) {
        rb_bug("ujit: output code calls wrong method cd->cc->klass: %p", (void *)cd->cc->klass);
    }
}

MJIT_FUNC_EXPORTED VALUE rb_hash_has_key(VALUE hash, VALUE key);

static bool
cfunc_needs_frame(const rb_method_cfunc_t *cfunc)
{
    void* fptr = (void*)cfunc->func;

    // Leaf C functions do not need a stack frame
    // or a stack overflow check
    return !(
        // Hash#key?
        fptr == (void*)rb_hash_has_key
    );
}

static bool
gen_opt_send_without_block(codeblock_t* cb, codeblock_t* ocb, ctx_t* ctx)
{
    // Relevant definitions:
    // rb_execution_context_t       : vm_core.h
    // invoker, cfunc logic         : method.h, vm_method.c
    // rb_callable_method_entry_t   : method.h
    // vm_call_cfunc_with_frame     : vm_insnhelper.c
    // rb_callcache                 : vm_callinfo.h

    struct rb_call_data * cd = (struct rb_call_data *)ctx_get_arg(ctx, 0);
    int32_t argc = (int32_t)vm_ci_argc(cd->ci);

    // Don't JIT calls with keyword splat
    if (vm_ci_flag(cd->ci) & VM_CALL_KW_SPLAT)
    {
        return false;
    }

    // Don't JIT calls that aren't simple
    if (!(vm_ci_flag(cd->ci) & VM_CALL_ARGS_SIMPLE))
    {
        return false;
    }

    // Don't JIT if the inline cache is not set
    if (!cd->cc || !cd->cc->klass) {
        return false;
    }

    const rb_callable_method_entry_t *cme = vm_cc_cme(cd->cc);

    // Don't JIT if the method entry is out of date
    if (METHOD_ENTRY_INVALIDATED(cme)) {
        return false;
    }

    // Don't JIT if this is not a C call
    if (cme->def->type != VM_METHOD_TYPE_CFUNC)
    {
        return false;
    }

    const rb_method_cfunc_t *cfunc = UNALIGNED_MEMBER_PTR(cme->def, body.cfunc);

    // Don't JIT if the argument count doesn't match
    if (cfunc->argc < 0 || cfunc->argc != argc)
    {
        return false;
    }

    // Don't JIT functions that need C stack arguments for now
    if (argc + 1 > NUM_C_ARG_REGS)
    {
        return false;
    }

    // Create a size-exit to fall back to the interpreter
    uint8_t* side_exit = ujit_side_exit(ocb, ctx, ctx->pc);

    // Check for interrupts
    // RUBY_VM_CHECK_INTS(ec)
    mov(cb, REG0_32, member_opnd(REG_EC, rb_execution_context_t, interrupt_mask));
    not(cb, REG0_32);
    test(cb, member_opnd(REG_EC, rb_execution_context_t, interrupt_flag), REG0_32);
    jnz_ptr(cb, side_exit);

    // Points to the receiver operand on the stack
    x86opnd_t recv = ctx_stack_opnd(ctx, argc);
    mov(cb, REG0, recv);

    // Callee method ID
    //ID mid = vm_ci_mid(cd->ci);
    //printf("JITting call to C function \"%s\", argc: %lu\n", rb_id2name(mid), argc);
    //print_str(cb, "");
    //print_str(cb, "calling CFUNC:");
    //print_str(cb, rb_id2name(mid));
    //print_str(cb, "recv");
    //print_ptr(cb, recv);

    // Check that the receiver is a heap object
    test(cb, REG0, imm_opnd(RUBY_IMMEDIATE_MASK));
    jnz_ptr(cb, side_exit);
    cmp(cb, REG0, imm_opnd(Qfalse));
    je_ptr(cb, side_exit);
    cmp(cb, REG0, imm_opnd(Qnil));
    je_ptr(cb, side_exit);

    // Pointer to the klass field of the receiver &(recv->klass)
    x86opnd_t klass_opnd = mem_opnd(64, REG0, offsetof(struct RBasic, klass));

    // Bail if receiver class is different from compiled time call cache class
    mov(cb, REG1, imm_opnd(cd->cc->klass));
    cmp(cb, klass_opnd, REG1);
    jne_ptr(cb, side_exit);

    // Store incremented PC into current control frame in case callee raises.
    mov(cb, REG0, const_ptr_opnd(ctx->pc + insn_len(BIN(opt_send_without_block))));
    mov(cb, mem_opnd(64, REG_CFP, offsetof(rb_control_frame_t, pc)), REG0);

    // If this function needs a Ruby stack frame
    if (cfunc_needs_frame(cfunc))
    {
        // Stack overflow check
        // #define CHECK_VM_STACK_OVERFLOW0(cfp, sp, margin)
        // REG_CFP <= REG_SP + 4 * sizeof(VALUE) + sizeof(rb_control_frame_t)
        lea(cb, REG0, ctx_sp_opnd(ctx, sizeof(VALUE) * 4 + sizeof(rb_control_frame_t)));
        cmp(cb, REG_CFP, REG0);
        jle_ptr(cb, side_exit);

        // Increment the stack pointer by 3 (in the callee)
        // sp += 3
        lea(cb, REG0, ctx_sp_opnd(ctx, sizeof(VALUE) * 3));

        // Put compile time cme into REG1. It's assumed to be valid because we are notified when
        // any cme we depend on become outdated. See rb_ujit_method_lookup_change().
        mov(cb, REG1, const_ptr_opnd(cme));
        // Write method entry at sp[-3]
        // sp[-3] = me;
        mov(cb, mem_opnd(64, REG0, 8 * -3), REG1);

        // Write block handler at sp[-2]
        // sp[-2] = block_handler;
        mov(cb, mem_opnd(64, REG0, 8 * -2), imm_opnd(VM_BLOCK_HANDLER_NONE));

        // Write env flags at sp[-1]
        // sp[-1] = frame_type;
        uint64_t frame_type = VM_FRAME_MAGIC_CFUNC | VM_FRAME_FLAG_CFRAME | VM_ENV_FLAG_LOCAL;
        mov(cb, mem_opnd(64, REG0, 8 * -1), imm_opnd(frame_type));

        // Allocate a new CFP (ec->cfp--)
        sub(
            cb,
            member_opnd(REG_EC, rb_execution_context_t, cfp),
            imm_opnd(sizeof(rb_control_frame_t))
        );

        // Setup the new frame
        // *cfp = (const struct rb_control_frame_struct) {
        //    .pc         = 0,
        //    .sp         = sp,
        //    .iseq       = 0,
        //    .self       = recv,
        //    .ep         = sp - 1,
        //    .block_code = 0,
        //    .__bp__     = sp,
        // };
        mov(cb, REG1, member_opnd(REG_EC, rb_execution_context_t, cfp));
        mov(cb, member_opnd(REG1, rb_control_frame_t, pc), imm_opnd(0));
        mov(cb, member_opnd(REG1, rb_control_frame_t, sp), REG0);
        mov(cb, member_opnd(REG1, rb_control_frame_t, iseq), imm_opnd(0));
        mov(cb, member_opnd(REG1, rb_control_frame_t, block_code), imm_opnd(0));
        mov(cb, member_opnd(REG1, rb_control_frame_t, __bp__), REG0);
        sub(cb, REG0, imm_opnd(sizeof(VALUE)));
        mov(cb, member_opnd(REG1, rb_control_frame_t, ep), REG0);
        mov(cb, REG0, recv);
        mov(cb, member_opnd(REG1, rb_control_frame_t, self), REG0);
    }

    if (UJIT_CHECK_MODE > 0) {
        // Verify that we are calling the right function
        // Save MicroJIT registers
        push(cb, REG_CFP);
        push(cb, REG_EC);
        push(cb, REG_SP);
        // Maintain 16-byte RSP alignment
        sub(cb, RSP, imm_opnd(8));

        // Call check_cfunc_dispatch
        mov(cb, RDI, recv);
        mov(cb, RSI, const_ptr_opnd(cd));
        mov(cb, RDX, const_ptr_opnd((void *)cfunc->func));
        mov(cb, RCX, const_ptr_opnd(cme));
        call_ptr(cb, REG0, (void *)&check_cfunc_dispatch);

        // Restore registers
        add(cb, RSP, imm_opnd(8));
        pop(cb, REG_SP);
        pop(cb, REG_EC);
        pop(cb, REG_CFP);
    }

    // Save the MicroJIT registers
    push(cb, REG_CFP);
    push(cb, REG_EC);
    push(cb, REG_SP);

    // Maintain 16-byte RSP alignment
    sub(cb, RSP, imm_opnd(8));

    // Copy SP into RAX because REG_SP will get overwritten
    lea(cb, RAX, ctx_sp_opnd(ctx, 0));

    // Copy the arguments from the stack to the C argument registers
    // self is the 0th argument and is at index argc from the stack top
    for (int32_t i = 0; i < argc + 1; ++i)
    {
        x86opnd_t stack_opnd = mem_opnd(64, RAX, -(argc + 1 - i) * 8);
        //print_ptr(cb, stack_opnd);
        x86opnd_t c_arg_reg = C_ARG_REGS[i];
        mov(cb, c_arg_reg, stack_opnd);
    }

    // Pop the C function arguments from the stack (in the caller)
    ctx_stack_pop(ctx, argc + 1);

    //print_str(cb, "before C call");

    ujit_assume_method_lookup_stable(cd->cc, cme, ctx);
    // Call the C function
    // VALUE ret = (cfunc->func)(recv, argv[0], argv[1]);
    // cfunc comes from compile-time cme->def, which we assume to be stable.
    // Invalidation logic is in rb_ujit_method_lookup_change()
    call_ptr(cb, REG0, (void*)cfunc->func);

    //print_str(cb, "after C call");

    // Maintain 16-byte RSP alignment
    add(cb, RSP, imm_opnd(8));

    // Restore MicroJIT registers
    pop(cb, REG_SP);
    pop(cb, REG_EC);
    pop(cb, REG_CFP);

    // Push the return value on the Ruby stack
    x86opnd_t stack_ret = ctx_stack_push(ctx, 1);
    mov(cb, stack_ret, RAX);

    // If this function needs a Ruby stack frame
    if (cfunc_needs_frame(cfunc))
    {
        // Pop the stack frame (ec->cfp++)
        add(
            cb,
            member_opnd(REG_EC, rb_execution_context_t, cfp),
            imm_opnd(sizeof(rb_control_frame_t))
        );
    }

    return true;
}

void
rb_ujit_compile_iseq(const rb_iseq_t *iseq)
{
#if OPT_DIRECT_THREADED_CODE || OPT_CALL_THREADED_CODE
    RB_VM_LOCK_ENTER();
    VALUE *encoded = (VALUE *)iseq->body->iseq_encoded;

    unsigned int insn_idx;
    unsigned int next_ujit_idx = 0;

    for (insn_idx = 0; insn_idx < iseq->body->iseq_size; /* */) {
        int insn = opcode_at_pc(iseq, &encoded[insn_idx]);
        int len = insn_len(insn);

        uint8_t *native_code_ptr = NULL;

        // If ujit hasn't already compiled this instruction
        if (insn_idx >= next_ujit_idx) {
            native_code_ptr = ujit_compile_insn(iseq, insn_idx, &next_ujit_idx);
        }

        if (native_code_ptr) {
            encoded[insn_idx] = (VALUE)native_code_ptr;
        }
        insn_idx += len;
    }
    RB_VM_LOCK_LEAVE();
#endif
}

// Callback when cme or cc become invalid
void
rb_ujit_method_lookup_change(VALUE cme_or_cc)
{
    if (!method_lookup_dependency) return;

    RUBY_ASSERT(IMEMO_TYPE_P(cme_or_cc, imemo_ment) || IMEMO_TYPE_P(cme_or_cc, imemo_callcache));

    st_data_t image;
    if (st_lookup(method_lookup_dependency, (st_data_t)cme_or_cc, &image)) {
        struct compiled_region_array *array = (void *)image;
        // Invalidate all regions that depend on the cme or cc
        for (int32_t i = 0; i < array->size; i++) {
            struct compiled_region *region = &array->data[i];
            const struct rb_iseq_constant_body *body = region->iseq->body;
            RUBY_ASSERT((unsigned int)region->start_idx < body->iseq_size);
            // Restore region address to interpreter address in bytecode sequence
            if (body->iseq_encoded[region->start_idx] == (VALUE)region->code) {
                const void *const *code_threading_table = rb_vm_get_insns_address_table();
                int opcode = rb_vm_insn_addr2insn(region->code);
                body->iseq_encoded[region->start_idx] = (VALUE)code_threading_table[opcode];
                if (UJIT_DUMP_MODE > 0) {
                    fprintf(stderr, "cc_or_cme=%p now out of date. Restored idx=%u in iseq=%p\n", (void *)cme_or_cc, (unsigned)region->start_idx, (void *)region->iseq);
                }
            }
        }

        array->size = 0;
    }
}

void
rb_ujit_init(void)
{
    if (!ujit_scrape_successful || !PLATFORM_SUPPORTED_P)
    {
        return;
    }

    rb_ujit_enabled = true;
    // Initialize the code blocks
    size_t mem_size = 128 * 1024 * 1024;
    uint8_t* mem_block = alloc_exec_mem(mem_size);
    cb = &block;
    cb_init(cb, mem_block, mem_size/2);
    ocb = &outline_block;
    cb_init(ocb, mem_block + mem_size/2, mem_size/2);

    // Initialize the codegen function table
    gen_fns = rb_st_init_numtable();

    // Map YARV opcodes to the corresponding codegen functions
    st_insert(gen_fns, (st_data_t)BIN(dup), (st_data_t)&gen_dup);
    st_insert(gen_fns, (st_data_t)BIN(nop), (st_data_t)&gen_nop);
    st_insert(gen_fns, (st_data_t)BIN(pop), (st_data_t)&gen_pop);
    st_insert(gen_fns, (st_data_t)BIN(putnil), (st_data_t)&gen_putnil);
    st_insert(gen_fns, (st_data_t)BIN(putobject), (st_data_t)&gen_putobject);
    st_insert(gen_fns, (st_data_t)BIN(putobject_INT2FIX_0_), (st_data_t)&gen_putobject_int2fix);
    st_insert(gen_fns, (st_data_t)BIN(putobject_INT2FIX_1_), (st_data_t)&gen_putobject_int2fix);
    st_insert(gen_fns, (st_data_t)BIN(putself), (st_data_t)&gen_putself);
    st_insert(gen_fns, (st_data_t)BIN(getlocal_WC_0), (st_data_t)&gen_getlocal_wc0);
    st_insert(gen_fns, (st_data_t)BIN(setlocal_WC_0), (st_data_t)&gen_setlocal_wc0);
    st_insert(gen_fns, (st_data_t)BIN(getinstancevariable), (st_data_t)&gen_getinstancevariable);
    st_insert(gen_fns, (st_data_t)BIN(setinstancevariable), (st_data_t)&gen_setinstancevariable);
    st_insert(gen_fns, (st_data_t)BIN(opt_minus), (st_data_t)&gen_opt_minus);
    st_insert(gen_fns, (st_data_t)BIN(opt_send_without_block), (st_data_t)&gen_opt_send_without_block);

    method_lookup_dependency = st_init_numtable();
    struct ujit_root_struct *root;
    VALUE ujit_root = TypedData_Make_Struct(0, struct ujit_root_struct, &ujit_root_type, root);
    rb_gc_register_mark_object(ujit_root);
}