Admin
7aa2bbefcf
- vm/Image.cpp：匿名空间 constexpr 偏移常量（kOffMagic..kOffModelId + 段内行偏移）， from/model_id/model_matches/func_row/const_entry 全部替换 - compiler/Stb.h：同值 inline constexpr 表（Codegen 写侧 + Stb 读侧共用） - compiler/Codegen.cpp assemble_image：put_* 头字段/行内偏移替换 - compiler/Stb.cpp StbView：读侧偏移替换 - 两侧 static_assert：头大小 = 72 + 型号标识、字段偏移单调、行内布局自洽
482 lines
17 KiB
C++
482 lines
17 KiB
C++
/**
|
||
* @file Stb.cpp
|
||
* @brief 编译器自带的 .stb 映像规范（写侧）+ 只读视图 + FNV-1a + sidecar
|
||
* @author
|
||
* @date 2026-08-21
|
||
*
|
||
* @details 与 vm 侧各一份实现；格式契约见 Doc/isa/指令与映像.md。
|
||
* 12.13 修订已落地：头 104 = 原 72 + 型号标识[32] @72，文件尾 SHA-256[32]；
|
||
* 工程哈希为 FNV-1a 64。
|
||
*/
|
||
|
||
#include "compiler/Stb.h"
|
||
|
||
#include <cstdio>
|
||
#include <cstring>
|
||
#include <fstream>
|
||
|
||
#include "compiler/Project.h"
|
||
|
||
namespace compiler {
|
||
|
||
/**
|
||
* @brief FNV-1a 64 增量哈希更新（工程哈希，非密码学）。
|
||
* @param h 当前哈希（首轮传 kFnvBasis）
|
||
* @param data 输入字节
|
||
* @param len 字节数
|
||
* @return 更新后的哈希值
|
||
* @details 每字节：h ^= byte; h *= kFnvPrime。
|
||
*/
|
||
uint64_t fnv1a64_update(uint64_t h, const uint8_t* data, size_t len) {
|
||
for (size_t i = 0; i < len; ++i) {
|
||
h ^= data[i];
|
||
h *= kFnvPrime;
|
||
}
|
||
return h;
|
||
}
|
||
|
||
/**
|
||
* @brief FNV-1a 64 一次性哈希（工程哈希）。
|
||
* @param data 输入字节
|
||
* @param len 字节数
|
||
* @return 哈希值（从 kFnvBasis 起）
|
||
*/
|
||
uint64_t fnv1a64(const uint8_t* data, size_t len) {
|
||
return fnv1a64_update(kFnvBasis, data, len);
|
||
}
|
||
|
||
namespace {
|
||
|
||
/// 小端读 32 位。
|
||
uint32_t get_le32(const uint8_t* p) {
|
||
return static_cast<uint32_t>(p[0])
|
||
| (static_cast<uint32_t>(p[1]) << 8)
|
||
| (static_cast<uint32_t>(p[2]) << 16)
|
||
| (static_cast<uint32_t>(p[3]) << 24);
|
||
}
|
||
|
||
/// 小端读 64 位。
|
||
uint64_t get_le64(const uint8_t* p) {
|
||
uint64_t v = 0;
|
||
for (int i = 0; i < 8; ++i) {
|
||
v |= static_cast<uint64_t>(p[i]) << (8 * i);
|
||
}
|
||
return v;
|
||
}
|
||
|
||
// ---- SHA-256（FIPS 180-4）----
|
||
|
||
/// SHA-256 轮常量 K[0..63]。
|
||
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,
|
||
};
|
||
|
||
/// 循环右移。
|
||
inline uint32_t rotr(uint32_t x, uint32_t n) { return (x >> n) | (x << (32 - n)); }
|
||
|
||
/**
|
||
* @brief SHA-256 增量状态机。
|
||
* @details 标准 FIPS 180-4 实现：update() 吸收任意长度字节流，final() 输出
|
||
* 32 字节大端摘要。按 64 字节块 process()。
|
||
*/
|
||
struct Sha256 {
|
||
uint32_t h[8] = {0x6a09e667, 0xbb67ae85, 0x3c6ef372, 0xa54ff53a,
|
||
0x510e527f, 0x9b05688c, 0x1f83d9ab, 0x5be0cd19}; ///< 初始哈希值
|
||
uint64_t total = 0; ///< 已吸收字节数（final 时编码进长度域）
|
||
uint8_t block[64]; ///< 当前块缓冲
|
||
size_t block_len = 0; ///< 块缓冲已用字节数
|
||
|
||
/**
|
||
* @brief 吸收数据。
|
||
* @param data 输入字节
|
||
* @param len 字节数
|
||
*/
|
||
void update(const uint8_t* data, size_t len) {
|
||
total += len;
|
||
while (len > 0) {
|
||
const size_t take = (block_len < 64) ? (64 - block_len) : 0;
|
||
const size_t n = len < take ? len : take;
|
||
if (n > 0) {
|
||
for (size_t i = 0; i < n; ++i) {
|
||
block[block_len + i] = data[i];
|
||
}
|
||
block_len += n;
|
||
data += n;
|
||
len -= n;
|
||
if (block_len == 64) {
|
||
process();
|
||
block_len = 0;
|
||
}
|
||
} else {
|
||
break;
|
||
}
|
||
}
|
||
}
|
||
|
||
/// 压缩一个满块（64 字节：w[0..63] 展开 + 64 轮）。
|
||
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;
|
||
}
|
||
|
||
/**
|
||
* @brief 结束并输出摘要。
|
||
* @param out 32 字节输出缓冲（大端）
|
||
*/
|
||
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};
|
||
// 补零到 block_len == 56（跨块时先补满当前块再补）
|
||
while (block_len != 56) {
|
||
const size_t n = (block_len < 56) ? (56 - block_len) : (64 - block_len);
|
||
update(zeros, n);
|
||
}
|
||
// 64 位大端比特长度
|
||
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);
|
||
}
|
||
}
|
||
|
||
/// 大端读 32 位。
|
||
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]);
|
||
}
|
||
};
|
||
|
||
} // namespace
|
||
|
||
/**
|
||
* @brief 一次性 SHA-256（文件完整性校验）。
|
||
* @param data 输入字节
|
||
* @param len 字节数
|
||
* @param out 32 字节摘要输出（大端）
|
||
*/
|
||
void sha256(const uint8_t* data, size_t len, uint8_t out[kSha256Size]) {
|
||
Sha256 s;
|
||
s.update(data, len);
|
||
s.final(out);
|
||
}
|
||
|
||
/**
|
||
* @brief 按型号标识约定填充 32 字节：name + version（如 "STATOR1"）。
|
||
* @param name 型号名
|
||
* @param version 版本号
|
||
* @param out 32 字节输出缓冲
|
||
* @details 不足补 '\0'，超长截断。
|
||
*/
|
||
void fill_model_id(const std::string& name, uint32_t version, char out[kModelIdSize]) {
|
||
const std::string id = name + std::to_string(version);
|
||
for (size_t i = 0; i < kModelIdSize; ++i) {
|
||
out[i] = i < id.size() ? id[i] : '\0';
|
||
}
|
||
}
|
||
|
||
/**
|
||
* @brief 从原始字节构造只读视图（不拷贝，调用方保证生命周期）。
|
||
* @param buf 映像缓冲；可为 nullptr
|
||
* @param len 缓冲字节数
|
||
* @return 校验结果；ok() 判成功，error() 取失败原因
|
||
* @details 校验顺序：长度 ≥ 头 → 魔数 → 版本 → 段偏移范围/单调 →
|
||
* 常量表/函数表大小 → 代码段 4 字节对齐 → 入口 fn_id 范围；
|
||
* 错误消息："null buffer" / "image too short" / "bad magic" /
|
||
* "bad version" / "missing sha256 tail" / "segment offset out of range" /
|
||
* "segment offsets not monotonic" / "const table size mismatch" /
|
||
* "function table size mismatch" / "code segment not 4-byte aligned" /
|
||
* "entry fn_id out of range"。
|
||
*/
|
||
StbView StbView::from(const uint8_t* buf, size_t len) {
|
||
StbView v;
|
||
v.buf_ = buf;
|
||
v.len_ = len;
|
||
if (buf == nullptr) {
|
||
v.err_ = "null buffer";
|
||
return v;
|
||
}
|
||
if (len < kHeaderSize) {
|
||
v.err_ = "image too short";
|
||
return v;
|
||
}
|
||
if (get_le32(buf + kOffMagic) != kMagic) {
|
||
v.err_ = "bad magic";
|
||
return v;
|
||
}
|
||
if (get_le32(buf + kOffVersion) != kVersion) {
|
||
v.err_ = "bad version";
|
||
return v;
|
||
}
|
||
v.cycle_limit_ = get_le32(buf + kOffCycleLimit);
|
||
v.dt_ms_ = get_le32(buf + kOffDtMs);
|
||
v.project_hash_ = get_le64(buf + kOffProjectHash);
|
||
v.entry_fn_id_ = get_le32(buf + kOffEntryFnId);
|
||
v.n_globals_ = get_le32(buf + kOffNGlobals);
|
||
v.n_consts_ = get_le32(buf + kOffNConsts);
|
||
v.n_funcs_ = get_le32(buf + kOffNFuncs);
|
||
v.offset_code_ = get_le32(buf + kOffCode);
|
||
v.offset_fb_ = get_le32(buf + kOffFb);
|
||
v.offset_data_ = get_le32(buf + kOffData);
|
||
|
||
// 段校验（12.13：文件尾 SHA-256[32] 在数据段之后）
|
||
if (len < static_cast<size_t>(v.offset_data_) + kSha256Size) {
|
||
v.err_ = "missing sha256 tail";
|
||
return v;
|
||
}
|
||
const uint64_t offs[5] = {get_le32(buf + kOffConst), get_le32(buf + kOffFuncs), v.offset_code_,
|
||
v.offset_fb_, v.offset_data_};
|
||
for (int i = 0; i < 5; ++i) {
|
||
if (offs[i] < kHeaderSize || offs[i] > len - kSha256Size) {
|
||
v.err_ = "segment offset out of range";
|
||
return v;
|
||
}
|
||
if (i > 0 && offs[i] < offs[i - 1]) {
|
||
v.err_ = "segment offsets not monotonic";
|
||
return v;
|
||
}
|
||
}
|
||
if (offs[1] - offs[0] != static_cast<uint64_t>(v.n_consts_) * kConstEntrySize) {
|
||
v.err_ = "const table size mismatch";
|
||
return v;
|
||
}
|
||
if (offs[2] - offs[1] != static_cast<uint64_t>(v.n_funcs_) * kFuncRowSize) {
|
||
v.err_ = "function table size mismatch";
|
||
return v;
|
||
}
|
||
if ((v.offset_fb_ - v.offset_code_) % 4 != 0) {
|
||
v.err_ = "code segment not 4-byte aligned";
|
||
return v;
|
||
}
|
||
if (v.entry_fn_id_ >= v.n_funcs_ && v.n_funcs_ != 0) {
|
||
v.err_ = "entry fn_id out of range";
|
||
return v;
|
||
}
|
||
v.ok_ = true;
|
||
v.err_.clear();
|
||
return v;
|
||
}
|
||
|
||
/// 从 vector 构造（转发 from(buf.data(), buf.size())）。
|
||
StbView StbView::from(const std::vector<uint8_t>& buf) {
|
||
return from(buf.data(), buf.size());
|
||
}
|
||
|
||
/// 读常量表一行；越界或未 ok() 时返回全 0。
|
||
ConstEntry StbView::const_entry(size_t i) const {
|
||
ConstEntry e;
|
||
if (ok_ && i < n_consts_) {
|
||
const uint8_t* p = buf_ + offs_of_const() + i * kConstEntrySize;
|
||
e.tag = get_le32(p);
|
||
e.value = get_le64(p + kConstValueOff);
|
||
}
|
||
return e;
|
||
}
|
||
|
||
/// 常量表段起点（头 @52；由 from() 已校验的段起点）。
|
||
uint32_t StbView::offs_of_const() const {
|
||
// offset_const = offs[0]，由 from() 已校验的段起点
|
||
return static_cast<uint32_t>(get_le32(buf_ + 52));
|
||
}
|
||
|
||
/// 读函数表一行；越界或未 ok() 时返回全 0。
|
||
StbView::FuncRow StbView::func_row(size_t i) const {
|
||
FuncRow r;
|
||
if (ok_ && i < n_funcs_) {
|
||
const uint8_t* p = buf_ + get_le32(buf_ + 56) + i * kFuncRowSize;
|
||
r.nregs = get_le32(p + kFuncNregsOff);
|
||
r.code_offset = get_le32(p + kFuncCodeOff);
|
||
r.code_len = get_le32(p + kFuncLenOff);
|
||
}
|
||
return r;
|
||
}
|
||
|
||
/// 字节码段起点（未 ok() 时返回 nullptr）。
|
||
const uint8_t* StbView::code_bytes() const {
|
||
return ok_ ? buf_ + offset_code_ : nullptr;
|
||
}
|
||
|
||
/// 字节码段字节数（未 ok() 时返回 0）。
|
||
size_t StbView::code_len() const {
|
||
return ok_ ? offset_fb_ - offset_code_ : 0;
|
||
}
|
||
|
||
/// 数据段起点（未 ok() 时返回 nullptr）。
|
||
const uint8_t* StbView::data_bytes() const {
|
||
return ok_ ? buf_ + offset_data_ : nullptr;
|
||
}
|
||
|
||
/// 数据段字节数（不含文件尾 SHA-256；未 ok() 时返回 0）。
|
||
size_t StbView::data_len() const {
|
||
return ok_ ? (len_ - kSha256Size) - offset_data_ : 0;
|
||
}
|
||
|
||
/// 型号标识字符串（头 72..103，截断到首个 '\0'；未 ok() 时返回空串）。
|
||
std::string StbView::model_id() const {
|
||
if (!ok_) {
|
||
return "";
|
||
}
|
||
std::string s(reinterpret_cast<const char*>(buf_ + kOffModelId), kModelIdSize);
|
||
const size_t z = s.find('\0');
|
||
if (z != std::string::npos) {
|
||
s.resize(z);
|
||
}
|
||
return s;
|
||
}
|
||
|
||
/// 型号标识是否匹配 name + version（未 ok() 时返回 false）。
|
||
bool StbView::model_matches(const std::string& name, uint32_t version) const {
|
||
char want[kModelIdSize];
|
||
fill_model_id(name, version, want);
|
||
return std::memcmp(buf_ + kOffModelId, want, kModelIdSize) == 0;
|
||
}
|
||
|
||
/// 文件尾 32 字节 SHA-256 校验（对文件尾之前全部内容重算；未 ok() 时返回 false）。
|
||
bool StbView::sha_ok() const {
|
||
if (!ok_) {
|
||
return false;
|
||
}
|
||
const size_t content_len = len_ - kSha256Size;
|
||
uint8_t digest[kSha256Size];
|
||
sha256(buf_, content_len, digest);
|
||
return std::memcmp(buf_ + content_len, digest, kSha256Size) == 0;
|
||
}
|
||
|
||
/**
|
||
* @brief 读整个文件为字节（纯字节；校验交给 StbView）。
|
||
* @param path 文件路径
|
||
* @param out 输出字节
|
||
* @param err 错误输出；可为 nullptr（静默）
|
||
* @return true 成功；false（err "cannot open for read: <path>"）
|
||
*/
|
||
bool read_stb_file(const char* path, std::vector<uint8_t>* out, std::string* err) {
|
||
std::ifstream in(path, std::ios::binary);
|
||
if (!in) {
|
||
if (err) {
|
||
*err = "cannot open for read: " + std::string(path);
|
||
}
|
||
return false;
|
||
}
|
||
out->assign(std::istreambuf_iterator<char>(in), std::istreambuf_iterator<char>());
|
||
return true;
|
||
}
|
||
|
||
/**
|
||
* @brief 写字节为文件（纯字节）。
|
||
* @param path 文件路径
|
||
* @param img 映像字节
|
||
* @param err 错误输出；可为 nullptr（静默）
|
||
* @return true 成功；false（err "cannot open for write: <path>" / "write failed: <path>"）
|
||
*/
|
||
bool write_stb_file(const char* path, const std::vector<uint8_t>& img, std::string* err) {
|
||
std::FILE* f = std::fopen(path, "wb");
|
||
if (f == nullptr) {
|
||
if (err) {
|
||
*err = "cannot open for write: " + std::string(path);
|
||
}
|
||
return false;
|
||
}
|
||
const bool ok = img.empty() || std::fwrite(&img[0], 1, img.size(), f) == img.size();
|
||
std::fclose(f);
|
||
if (!ok && err) {
|
||
*err = "write failed: " + std::string(path);
|
||
}
|
||
return ok;
|
||
}
|
||
|
||
/**
|
||
* @brief 生成 sidecar 文本（TOML）。
|
||
* @param bindings I/O 绑定列表
|
||
* @return TOML 文本（每个绑定一节 [[io.input]] / [[io.output]]，含
|
||
* var / slot / channel / bit 四字段）
|
||
* @details I/O 绑定 var → 槽号 → channel/bit（执行器采样用）。
|
||
*/
|
||
std::string make_sidecar(const std::vector<IoBinding>& bindings) {
|
||
std::string out;
|
||
for (const IoBinding& b : bindings) {
|
||
char buf[64];
|
||
out += b.is_input ? "[[io.input]]\n" : "[[io.output]]\n";
|
||
out += "var = \"";
|
||
out += b.var;
|
||
out += "\"\n";
|
||
std::snprintf(buf, sizeof buf, "slot = %u\n", static_cast<unsigned>(b.slot));
|
||
out += buf;
|
||
std::snprintf(buf, sizeof buf, "channel = %u\n", static_cast<unsigned>(b.channel));
|
||
out += buf;
|
||
std::snprintf(buf, sizeof buf, "bit = %u\n", static_cast<unsigned>(b.bit));
|
||
out += buf;
|
||
out += "\n";
|
||
}
|
||
return out;
|
||
}
|
||
|
||
/**
|
||
* @brief 生成并写 sidecar 文件。
|
||
* @param path 文件路径
|
||
* @param bindings I/O 绑定列表
|
||
* @param err 错误输出；可为 nullptr（静默）
|
||
* @return true 成功；false（err "cannot open for write: <path>" / "write failed: <path>"）
|
||
*/
|
||
bool write_sidecar_file(const char* path, const std::vector<IoBinding>& bindings,
|
||
std::string* err) {
|
||
std::FILE* f = std::fopen(path, "wb");
|
||
if (f == nullptr) {
|
||
if (err) {
|
||
*err = "cannot open for write: " + std::string(path);
|
||
}
|
||
return false;
|
||
}
|
||
const std::string s = make_sidecar(bindings);
|
||
const bool ok = s.empty() || std::fwrite(s.data(), 1, s.size(), f) == s.size();
|
||
std::fclose(f);
|
||
if (!ok && err) {
|
||
*err = "write failed: " + std::string(path);
|
||
}
|
||
return ok;
|
||
}
|
||
|
||
} // namespace compiler
|