YJIT: Bounds check every byte in the assembler

Previously, YJIT assumed that basic blocks never consume more than
1 KiB of memory. This assumption does not hold for long Ruby methods
such as the one in the following:

```ruby
eval(<<RUBY)
def set_local_a_lot
  #{'_=0;'*0x40000}
end
RUBY

set_local_a_lot
```

For low `--yjit-exec-mem-size` values, one basic block could exhaust the
entire buffer.

Introduce a new field `codeblock_t::dropped_bytes` that the assembler
sets whenever it runs out of space. Check this field in
gen_single_block() to respond to out of memory situations and other
error conditions. This design avoids making the control flow graph of
existing code generation functions more complex.

Use POSIX shell in misc/test_yjit_asm.sh since bash is expanding
`0%/*/*` differently.

Co-authored-by: Aaron Patterson <tenderlove@ruby-lang.org>

misc/test_yjit_asm.sh

@@ -1,9 +1,9 @@
-#!/bin/bash
+#!/bin/sh
 
 set -e
 set -x
 
-clang -std=gnu99 -Wall -Werror -Wno-error=unused-function -Wshorten-64-to-32 -I ${0%/*/*} ${0%/*}/yjit_asm_tests.c -o asm_test
+clang -std=gnu99 -Wall -Werror -Wno-error=unused-function -Wshorten-64-to-32 -I "${0%/*/*}" "${0%/*}/yjit_asm_tests.c" -o asm_test
 
 ./asm_test
 

misc/yjit_asm_tests.c

@@ -26,7 +26,7 @@ void print_bytes(codeblock_t* cb)
 {
     for (uint32_t i = 0; i < cb->write_pos; ++i)
     {
-        printf("%02X", (int)cb->mem_block[i]);
+        printf("%02X", (int)*cb_get_ptr(cb, i));
     }
 
     printf("\n");
@@ -59,7 +59,7 @@ void check_bytes(codeblock_t* cb, const char* bytes)
         char* endptr;
         long int byte = strtol(byte_str, &endptr, 16);
 
-        uint8_t cb_byte = cb->mem_block[i];
+        uint8_t cb_byte = *cb_get_ptr(cb, i);
 
         if (cb_byte != byte)
         {

yjit_asm.c

@@ -211,11 +211,10 @@ static uint8_t *alloc_exec_mem(uint32_t mem_size)
     }
 
     codeblock_t block;
-    block.current_aligned_write_pos = ALIGNED_WRITE_POSITION_NONE;
-    block.mem_block = mem_block;
-    block.mem_size = mem_size;
+    codeblock_t *cb = █
+
+    cb_init(cb, mem_block, mem_size);
 
-    codeblock_t * cb = █
     // Fill the executable memory with INT3 (0xCC) so that
     // executing uninitialized memory will fault
     cb_mark_all_writeable(cb);
@@ -233,61 +232,74 @@ static uint8_t *alloc_exec_mem(uint32_t mem_size)
 void cb_init(codeblock_t *cb, uint8_t *mem_block, uint32_t mem_size)
 {
     assert (mem_block);
-    cb->mem_block = mem_block;
+    cb->mem_block_ = mem_block;
     cb->mem_size = mem_size;
     cb->write_pos = 0;
     cb->num_labels = 0;
     cb->num_refs = 0;
     cb->current_aligned_write_pos = ALIGNED_WRITE_POSITION_NONE;
 }
 
+// Set the current write position
+void cb_set_pos(codeblock_t *cb, uint32_t pos)
+{
+    // Assert here since while assembler functions do bounds checking, there is
+    // nothing stopping users from taking out an out-of-bounds pointer and
+    // doing bad accesses with it.
+    assert (pos < cb->mem_size);
+    cb->write_pos = pos;
+}
+
 // Align the current write position to a multiple of bytes
 void cb_align_pos(codeblock_t *cb, uint32_t multiple)
 {
     // Compute the pointer modulo the given alignment boundary
-    uint8_t *ptr = &cb->mem_block[cb->write_pos];
+    uint8_t *ptr = cb_get_write_ptr(cb);
     uint8_t *aligned_ptr = align_ptr(ptr, multiple);
+    const uint32_t write_pos = cb->write_pos;
 
     // Pad the pointer by the necessary amount to align it
     ptrdiff_t pad = aligned_ptr - ptr;
-    cb->write_pos += (int32_t)pad;
-}
-
-// Set the current write position
-void cb_set_pos(codeblock_t *cb, uint32_t pos)
-{
-    assert (pos < cb->mem_size);
-    cb->write_pos = pos;
+    cb_set_pos(cb, write_pos + (int32_t)pad);
 }
 
 // Set the current write position from a pointer
 void cb_set_write_ptr(codeblock_t *cb, uint8_t *code_ptr)
 {
-    intptr_t pos = code_ptr - cb->mem_block;
+    intptr_t pos = code_ptr - cb->mem_block_;
     assert (pos < cb->mem_size);
-    cb->write_pos = (uint32_t)pos;
+    cb_set_pos(cb, (uint32_t)pos);
 }
 
 // Get a direct pointer into the executable memory block
-uint8_t *cb_get_ptr(codeblock_t *cb, uint32_t index)
+uint8_t *cb_get_ptr(const codeblock_t *cb, uint32_t index)
 {
-    assert (index < cb->mem_size);
-    return &cb->mem_block[index];
+    if (index < cb->mem_size) {
+        return &cb->mem_block_[index];
+    }
+    else {
+        return NULL;
+    }
 }
 
 // Get a direct pointer to the current write position
-uint8_t *cb_get_write_ptr(codeblock_t *cb)
+uint8_t *cb_get_write_ptr(const codeblock_t *cb)
 {
     return cb_get_ptr(cb, cb->write_pos);
 }
 
 // Write a byte at the current position
 void cb_write_byte(codeblock_t *cb, uint8_t byte)
 {
-    assert (cb->mem_block);
-    assert (cb->write_pos + 1 <= cb->mem_size);
-    cb_mark_position_writeable(cb, cb->write_pos);
-    cb->mem_block[cb->write_pos++] = byte;
+    assert (cb->mem_block_);
+    if (cb->write_pos < cb->mem_size) {
+        cb_mark_position_writeable(cb, cb->write_pos);
+        cb->mem_block_[cb->write_pos] = byte;
+        cb->write_pos++;
+    }
+    else {
+        cb->dropped_bytes = true;
+    }
 }
 
 // Write multiple bytes starting from the current position
@@ -890,14 +902,18 @@ static void cb_write_jcc_ptr(codeblock_t *cb, const char *mnem, uint8_t op0, uin
     cb_write_byte(cb, op1);
 
     // Pointer to the end of this jump instruction
-    uint8_t *end_ptr = &cb->mem_block[cb->write_pos] + 4;
+    uint8_t *end_ptr = cb_get_ptr(cb, cb->write_pos + 4);
 
     // Compute the jump offset
     int64_t rel64 = (int64_t)(dst_ptr - end_ptr);
-    assert (rel64 >= INT32_MIN && rel64 <= INT32_MAX);
-
-    // Write the relative 32-bit jump offset
-    cb_write_int(cb, (int32_t)rel64, 32);
+    if (rel64 >= INT32_MIN && rel64 <= INT32_MAX) {
+        // Write the relative 32-bit jump offset
+        cb_write_int(cb, (int32_t)rel64, 32);
+    }
+    else {
+        // Offset doesn't fit in 4 bytes. Report error.
+        cb->dropped_bytes = true;
+    }
 }
 
 // Encode a conditional move instruction
@@ -971,14 +987,13 @@ void call_ptr(codeblock_t *cb, x86opnd_t scratch_reg, uint8_t *dst_ptr)
     assert (scratch_reg.type == OPND_REG);
 
     // Pointer to the end of this call instruction
-    uint8_t *end_ptr = &cb->mem_block[cb->write_pos] + 5;
+    uint8_t *end_ptr = cb_get_ptr(cb, cb->write_pos + 5);
 
     // Compute the jump offset
     int64_t rel64 = (int64_t)(dst_ptr - end_ptr);
 
     // If the offset fits in 32-bit
-    if (rel64 >= INT32_MIN && rel64 <= INT32_MAX)
-    {
+    if (rel64 >= INT32_MIN && rel64 <= INT32_MAX) {
         call_rel32(cb, (int32_t)rel64);
         return;
     }
@@ -1784,8 +1799,8 @@ void cb_write_lock_prefix(codeblock_t *cb)
 
 void cb_mark_all_writeable(codeblock_t * cb)
 {
-    if (mprotect(cb->mem_block, cb->mem_size, PROT_READ | PROT_WRITE)) {
-        fprintf(stderr, "Couldn't make JIT page (%p) writeable, errno: %s", (void *)cb->mem_block, strerror(errno));
+    if (mprotect(cb->mem_block_, cb->mem_size, PROT_READ | PROT_WRITE)) {
+        fprintf(stderr, "Couldn't make JIT page (%p) writeable, errno: %s", (void *)cb->mem_block_, strerror(errno));
         abort();
     }
 }
@@ -1797,8 +1812,9 @@ void cb_mark_position_writeable(codeblock_t * cb, uint32_t write_pos)
 
     if (cb->current_aligned_write_pos != aligned_position) {
         cb->current_aligned_write_pos = aligned_position;
-        if (mprotect(cb->mem_block + aligned_position, pagesize, PROT_READ | PROT_WRITE)) {
-            fprintf(stderr, "Couldn't make JIT page (%p) writeable, errno: %s", (void *)(cb->mem_block + aligned_position), strerror(errno));
+        void *const page_addr = cb_get_ptr(cb, aligned_position);
+        if (mprotect(page_addr, pagesize, PROT_READ | PROT_WRITE)) {
+            fprintf(stderr, "Couldn't make JIT page (%p) writeable, errno: %s", page_addr, strerror(errno));
             abort();
         }
     }
@@ -1807,8 +1823,8 @@ void cb_mark_position_writeable(codeblock_t * cb, uint32_t write_pos)
 void cb_mark_all_executable(codeblock_t * cb)
 {
     cb->current_aligned_write_pos = ALIGNED_WRITE_POSITION_NONE;
-    if (mprotect(cb->mem_block, cb->mem_size, PROT_READ | PROT_EXEC)) {
-        fprintf(stderr, "Couldn't make JIT page (%p) executable, errno: %s", (void *)cb->mem_block, strerror(errno));
+    if (mprotect(cb->mem_block_, cb->mem_size, PROT_READ | PROT_EXEC)) {
+        fprintf(stderr, "Couldn't make JIT page (%p) executable, errno: %s", (void *)cb->mem_block_, strerror(errno));
         abort();
     }
 }

yjit_asm.h

@@ -26,12 +26,13 @@ typedef struct LabelRef
 typedef struct CodeBlock
 {
     // Memory block
-    uint8_t *mem_block;
+    // Users are advised to not use this directly.
+    uint8_t *mem_block_;
 
     // Memory block size
     uint32_t mem_size;
 
-    /// Current writing position
+    // Current writing position
     uint32_t write_pos;
 
     // Table of registered label addresses
@@ -50,14 +51,20 @@ typedef struct CodeBlock
     // Number of references to labels
     uint32_t num_refs;
 
-    // TODO: system for disassembly/comment strings, indexed by position
-
-    // Flag to enable or disable comments
-    bool has_asm;
 
     // Keep track of the current aligned write position.
     // Used for changing protection when writing to the JIT buffer
     uint32_t current_aligned_write_pos;
+
+    // Set if the assembler is unable to output some instructions,
+    // for example, when there is not enough space or when a jump
+    // target is too far away.
+    bool dropped_bytes;
+
+    // Flag to enable or disable comments
+    bool has_asm;
+
+
 } codeblock_t;
 
 // 1 is not aligned so this won't match any pages
@@ -258,8 +265,8 @@ static inline void cb_init(codeblock_t *cb, uint8_t *mem_block, uint32_t mem_siz
 static inline void cb_align_pos(codeblock_t *cb, uint32_t multiple);
 static inline void cb_set_pos(codeblock_t *cb, uint32_t pos);
 static inline void cb_set_write_ptr(codeblock_t *cb, uint8_t *code_ptr);
-static inline uint8_t *cb_get_ptr(codeblock_t *cb, uint32_t index);
-static inline uint8_t *cb_get_write_ptr(codeblock_t *cb);
+static inline uint8_t *cb_get_ptr(const codeblock_t *cb, uint32_t index);
+static inline uint8_t *cb_get_write_ptr(const codeblock_t *cb);
 static inline void cb_write_byte(codeblock_t *cb, uint8_t byte);
 static inline void cb_write_bytes(codeblock_t *cb, uint32_t num_bytes, ...);
 static inline void cb_write_int(codeblock_t *cb, uint64_t val, uint32_t num_bits);

yjit_codegen.c

@@ -554,7 +554,7 @@ yjit_entry_prologue(codeblock_t *cb, const rb_iseq_t *iseq)
 
     const uint32_t old_write_pos = cb->write_pos;
 
-    // Align the current write positon to cache line boundaries
+    // Align the current write position to cache line boundaries
     cb_align_pos(cb, 64);
 
     uint8_t *code_ptr = cb_get_ptr(cb, cb->write_pos);
@@ -640,16 +640,6 @@ gen_single_block(blockid_t blockid, const ctx_t *start_ctx, rb_execution_context
 {
     RUBY_ASSERT(cb != NULL);
 
-    // Check if there is enough executable memory.
-    // FIXME: This bound isn't enforced and long blocks can potentially use more.
-    enum { MAX_CODE_PER_BLOCK = 1024 };
-    if (cb->write_pos + MAX_CODE_PER_BLOCK >= cb->mem_size) {
-        return NULL;
-    }
-    if (ocb->write_pos + MAX_CODE_PER_BLOCK >= ocb->mem_size) {
-        return NULL;
-    }
-
     // Allocate the new block
     block_t *block = calloc(1, sizeof(block_t));
     if (!block) {
@@ -778,6 +768,12 @@ gen_single_block(blockid_t blockid, const ctx_t *start_ctx, rb_execution_context
     // doesn't go to the next instruction.
     RUBY_ASSERT(!jit.record_boundary_patch_point);
 
+    // If code for the block doesn't fit, free the block and fail.
+    if (cb->dropped_bytes || ocb->dropped_bytes) {
+        yjit_free_block(block);
+        return NULL;
+    }
+
     if (YJIT_DUMP_MODE >= 2) {
         // Dump list of compiled instrutions
         fprintf(stderr, "Compiled the following for iseq=%p:\n", (void *)iseq);

yjit_iface.c

@@ -1150,7 +1150,7 @@ yjit_get_code_page(uint32_t cb_bytes_needed, uint32_t ocb_bytes_needed)
     code_page_t *new_code_page = rb_yjit_code_page_unwrap(yjit_cur_code_page);
 
     // Jump to the new code page
-    jmp_ptr(&code_page->cb, new_code_page->cb.mem_block);
+    jmp_ptr(&code_page->cb, cb_get_ptr(&new_code_page->cb, 0));
 
     return yjit_cur_code_page;
 }