阶段 C 步骤 8:vm 读侧型号匹配 + SHA-256 校验。

- vm/Image.cpp:执行器侧 SHA-256 实现(与 compiler 各一份)、fill_model_id、
  model_id()/model_matches()/sha_ok();data_len 不含 SHA 尾
- Machine::create:sha256 不匹配 → 'image sha256 mismatch';型号不匹配 →
  'image model mismatch'(内建 kModelName=STATOR/kModelVersion=1),均拒绝
- vm_test:手拼映像填型号(修越界读 bug)+ compiler::sha256 填尾(95 断言);
  新增篡改拒绝(sha256 mismatch)与型号不匹配(重算 SHA 后 model mismatch)用例
- 验证:line1 正常执行 Q=1;篡改一字节 → 执行器 'image sha256 mismatch' 拒绝;ctest 12/12
This commit is contained in:
2026-08-21 22:29:33 +08:00
parent 5fccbadaa8
commit b13585711a
4 changed files with 246 additions and 4 deletions
+70
View File
@@ -13,6 +13,7 @@
#include "compiler/Codegen.h"
#include "compiler/Linker.h"
#include "compiler/Project.h"
#include "compiler/Stb.h"
#include "compiler/Typecheck.h"
#include "isa/Encode.h"
#include "isa/Instr.h"
@@ -155,6 +156,16 @@ static std::vector<uint8_t> hand_image(const std::vector<vm::ConstEntry>& consts
for (size_t i = 0; i < data.size(); ++i) {
b[off_data + i] = data[i];
}
// 型号标识(STATOR1+ 真实 SHA-256 尾(用 compiler 实现填)
const char* mid = "STATOR1";
for (size_t i = 0; i < 32; ++i) {
b[72 + i] = i < 7 ? static_cast<uint8_t>(mid[i]) : 0;
}
uint8_t digest[32];
compiler::sha256(b.data(), b.size() - 32, digest);
for (int i = 0; i < 32; ++i) {
b[b.size() - 32 + i] = digest[i];
}
return b;
}
@@ -428,6 +439,64 @@ static bool test_bad_const() {
return true;
}
// ---- 10. 12.13 校验:SHA 篡改拒绝 + 型号不匹配拒绝 ----
static bool test_verify() {
std::string err;
vm::Machine m;
// 正常映像可跑
CHECK(make_machine("tests/cases/01_empty_main", &m, &err));
CHECK(m.run_cycle() == vm::Fault::None);
// SHA 篡改:改一字节 → create 拒绝
{
const std::string toml = std::string(REPO_ROOT) + "/tests/cases/01_empty_main/project.toml";
compiler::Project p;
CHECK(compiler::parse_project(toml, &p, &err));
std::vector<compiler::SourceUnit> units;
for (const std::string& f : compiler::compile_files(p)) {
compiler::SourceUnit u;
CHECK(compiler::load_unit(p.base_dir + "/" + f, &u, &err));
units.push_back(std::move(u));
}
compiler::LinkResult link;
CHECK(compiler::link_project(p, units, &link, &err));
compiler::MachineConfig cfg;
CHECK(cfg.load(std::string(REPO_ROOT) + "/compiler/machine.toml", &err));
std::vector<uint8_t> img;
CHECK(compiler::codegen_project(p, units, link, cfg, &img, &err));
img[100] ^= 0x01; // 篡改代码段一字节
vm::Machine bad;
CHECK(!vm::Machine::create(img, &bad, &err));
CHECK(err.find("sha256 mismatch") != std::string::npos);
}
// 型号不匹配:手拼映像改型号(重算 SHA,仅型号不一致)→ create 拒绝
{
const std::vector<vm::ConstEntry> consts;
const std::vector<HFunc> funcs = {{
{
isa::enc_ret(),
},
8,
}};
std::vector<uint8_t> img = hand_image(consts, funcs);
for (int i = 0; i < 5; ++i) {
img[72 + i] = "OTHER"[i];
}
uint8_t digest[32];
compiler::sha256(img.data(), img.size() - 32, digest);
for (int i = 0; i < 32; ++i) {
img[img.size() - 32 + i] = digest[i];
}
vm::Machine bad;
CHECK(!vm::Machine::create(img, &bad, &err));
CHECK(err.find("model mismatch") != std::string::npos);
}
return true;
}
int main() {
if (!test_hand_scalar()) return 1;
if (!test_hand_call()) return 1;
@@ -438,6 +507,7 @@ int main() {
if (!test_determinism()) return 1;
if (!test_step()) return 1;
if (!test_bad_const()) return 1;
if (!test_verify()) return 1;
std::printf("vm_test: %d checks passed\n", g_checks);
return 0;
}
+10 -1
View File
@@ -17,6 +17,10 @@
namespace vm {
// 执行器内建支持型号(与 machine.toml [meta] 对齐;.stb 型号不匹配直接拒绝)
static const char* const kModelName = "STATOR";
static const uint32_t kModelVersion = 1;
// 映像头(字段与 Doc/isa/指令与映像.md 一致)
struct ImageHeader {
uint32_t cycle_limit = 0;
@@ -62,7 +66,12 @@ namespace vm {
const uint8_t* code_bytes() const; // 字节码段起点
const uint8_t* data_bytes() const; // 数据段起点
size_t data_len() const; // 字节数
size_t data_len() const; // 字节数(不含 SHA 尾)
// 12.13:型号标识与 SHA-256 校验
std::string model_id() const;
bool model_matches(const std::string& name, uint32_t version) const;
bool sha_ok() const; // 文件尾 32 字节 SHA-256 校验
public:
// 默认构造为无效态(ok() == false
+152 -3
View File
@@ -7,6 +7,9 @@
#include "vm/Image.h"
#include <cstring>
#include <string>
namespace vm {
namespace {
@@ -16,7 +19,8 @@ namespace {
const size_t kHeaderSize = 104;
const size_t kConstEntrySize = 12;
const size_t kFuncRowSize = 12;
const size_t kSha256Size = 32; // 文件尾(步骤 8 校验)
const size_t kSha256Size = 32;
const size_t kModelIdSize = 32;
uint32_t get_le32(const uint8_t* p) {
return static_cast<uint32_t>(p[0])
@@ -33,6 +37,124 @@ namespace {
return v;
}
// ---- SHA-256(执行器侧实现,与 compiler 各一份)----
const uint32_t kShaK[64] = {
0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5, 0x3956c25b, 0x59f111f1,
0x923f82a4, 0xab1c5ed5, 0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3,
0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174, 0xe49b69c1, 0xefbe4786,
0x0fc19dc6, 0x240ca1cc, 0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da,
0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7, 0xc6e00bf3, 0xd5a79147,
0x06ca6351, 0x14292967, 0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13,
0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85, 0xa2bfe8a1, 0xa81a664b,
0xc24b8b70, 0xc76c51a3, 0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070,
0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5, 0x391c0cb3, 0x4ed8aa4a,
0x5b9cca4f, 0x682e6ff3, 0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208,
0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2,
};
inline uint32_t rotr(uint32_t x, uint32_t n) { return (x >> n) | (x << (32 - n)); }
struct Sha256 {
uint32_t h[8] = {0x6a09e667, 0xbb67ae85, 0x3c6ef372, 0xa54ff53a,
0x510e527f, 0x9b05688c, 0x1f83d9ab, 0x5be0cd19};
uint64_t total = 0;
uint8_t block[64];
size_t block_len = 0;
void update(const uint8_t* data, size_t len) {
total += len;
while (len > 0) {
const size_t n = (block_len < 64) ? (64 - block_len) : 0;
const size_t take = len < n ? len : n;
if (take == 0) {
break;
}
for (size_t i = 0; i < take; ++i) {
block[block_len + i] = data[i];
}
block_len += take;
data += take;
len -= take;
if (block_len == 64) {
process();
block_len = 0;
}
}
}
void process() {
uint32_t w[64];
for (int i = 0; i < 16; ++i) {
w[i] = get_be32(block + i * 4);
}
for (int i = 16; i < 64; ++i) {
const uint32_t s0 = rotr(w[i - 15], 7) ^ rotr(w[i - 15], 18) ^ (w[i - 15] >> 3);
const uint32_t s1 = rotr(w[i - 2], 17) ^ rotr(w[i - 2], 19) ^ (w[i - 2] >> 10);
w[i] = w[i - 16] + s0 + w[i - 7] + s1;
}
uint32_t a = h[0], b = h[1], c = h[2], d = h[3];
uint32_t e = h[4], f = h[5], g = h[6], hh = h[7];
for (int i = 0; i < 64; ++i) {
const uint32_t s1 = rotr(e, 6) ^ rotr(e, 11) ^ rotr(e, 25);
const uint32_t ch = (e & f) ^ (~e & g);
const uint32_t t1 = hh + s1 + ch + kShaK[i] + w[i];
const uint32_t s0 = rotr(a, 2) ^ rotr(a, 13) ^ rotr(a, 22);
const uint32_t maj = (a & b) ^ (a & c) ^ (b & c);
const uint32_t t2 = s0 + maj;
hh = g;
g = f;
f = e;
e = d + t1;
d = c;
c = b;
b = a;
a = t1 + t2;
}
h[0] += a; h[1] += b; h[2] += c; h[3] += d;
h[4] += e; h[5] += f; h[6] += g; h[7] += hh;
}
void final(uint8_t out[32]) {
const uint64_t bitlen = total * 8;
const uint8_t pad = 0x80;
update(&pad, 1);
const uint8_t zeros[64] = {0};
while (block_len != 56) {
const size_t n = (block_len < 56) ? (56 - block_len) : (64 - block_len);
update(zeros, n);
}
for (int i = 0; i < 8; ++i) {
const uint8_t b2[1] = {static_cast<uint8_t>((bitlen >> (56 - 8 * i)) & 0xFF)};
update(b2, 1);
}
for (int i = 0; i < 8; ++i) {
out[i * 4 + 0] = static_cast<uint8_t>((h[i] >> 24) & 0xFF);
out[i * 4 + 1] = static_cast<uint8_t>((h[i] >> 16) & 0xFF);
out[i * 4 + 2] = static_cast<uint8_t>((h[i] >> 8) & 0xFF);
out[i * 4 + 3] = static_cast<uint8_t>(h[i] & 0xFF);
}
}
static uint32_t get_be32(const uint8_t* p) {
return (static_cast<uint32_t>(p[0]) << 24) | (static_cast<uint32_t>(p[1]) << 16) |
(static_cast<uint32_t>(p[2]) << 8) | static_cast<uint32_t>(p[3]);
}
};
void sha256(const uint8_t* data, size_t len, uint8_t out[32]) {
Sha256 s;
s.update(data, len);
s.final(out);
}
void fill_model_id(const std::string& name, uint32_t version, char out[32]) {
const std::string id = name + std::to_string(version);
for (size_t i = 0; i < 32; ++i) {
out[i] = i < id.size() ? id[i] : '\0';
}
}
} // namespace
Image Image::from(const uint8_t* buf, size_t len) {
@@ -144,8 +266,35 @@ const uint8_t* Image::data_bytes() const {
}
size_t Image::data_len() const {
// 步骤 7:含 SHA 尾(步骤 8 收紧并校验)
return ok_ ? len_ - hdr_.offset_data : 0;
return ok_ ? (len_ - kSha256Size) - hdr_.offset_data : 0;
}
std::string Image::model_id() const {
if (!ok_) {
return "";
}
std::string s(reinterpret_cast<const char*>(buf_ + 72), kModelIdSize);
const size_t z = s.find('\0');
if (z != std::string::npos) {
s.resize(z);
}
return s;
}
bool Image::model_matches(const std::string& name, uint32_t version) const {
char want[kModelIdSize];
fill_model_id(name, version, want);
return std::memcmp(buf_ + 72, want, kModelIdSize) == 0;
}
bool Image::sha_ok() const {
if (!ok_) {
return false;
}
const size_t content_len = len_ - kSha256Size;
uint8_t digest[32];
sha256(buf_, content_len, digest);
return std::memcmp(buf_ + content_len, digest, 32) == 0;
}
} // namespace vm
+14
View File
@@ -32,6 +32,20 @@ bool Machine::create(const std::vector<uint8_t>& image, Machine* out,
}
return false;
}
// 12.13:型号匹配与 SHA-256 校验(不匹配/篡改直接拒绝)
if (!out->image_.sha_ok()) {
if (err) {
*err = "image sha256 mismatch";
}
return false;
}
if (!out->image_.model_matches(kModelName, kModelVersion)) {
if (err) {
*err = "image model mismatch: '" + out->image_.model_id() +
"' (expect " + kModelName + std::to_string(kModelVersion) + ")";
}
return false;
}
// 数据区工作副本(8 字节定宽槽)
out->data_.assign(out->image_.data_bytes(),
out->image_.data_bytes() + out->image_.data_len());