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
423 lines
15 KiB
C++
423 lines
15 KiB
C++
/**
|
||
* @file Image.cpp
|
||
* @brief vm 自带的 .stb 只读视图实现（执行器侧；compiler 各有实现）
|
||
* @author
|
||
* @date 2026-08-21
|
||
*/
|
||
|
||
#include "vm/Image.h"
|
||
|
||
#include <cstring>
|
||
#include <string>
|
||
|
||
namespace vm {
|
||
|
||
namespace {
|
||
|
||
/// 映像魔数："STSC" 小端。
|
||
const uint32_t kMagic = 0x43545353u;
|
||
/// 映像格式版本。
|
||
const uint32_t kVersion = 1;
|
||
/// 映像头字节数（72 原字段 + 型号标识[32] @72，12.13 修订）。
|
||
const size_t kHeaderSize = 104;
|
||
/// 常量表一行字节数（tag:4 + value:8）。
|
||
const size_t kConstEntrySize = 12;
|
||
/// 函数表一行字节数（nregs/code_offset/code_len 各 4）。
|
||
const size_t kFuncRowSize = 12;
|
||
/// SHA-256 摘要长度（文件尾）。
|
||
const size_t kSha256Size = 32;
|
||
/// 型号标识长度（头内 @72，不足补 '\0'）。
|
||
const size_t kModelIdSize = 32;
|
||
|
||
// ---- 头字段偏移（与 Doc/isa/指令与映像.md 字段表一致；compiler 侧另有同值一份）----
|
||
constexpr size_t kOffMagic = 0; ///< 魔数 "STSC"
|
||
constexpr size_t kOffVersion = 4; ///< 格式版本
|
||
constexpr size_t kOffCycleLimit = 8; ///< 每周期指令数上限
|
||
constexpr size_t kOffDtMs = 12; ///< 周期时长（毫秒）
|
||
constexpr size_t kOffProjectHash = 16; ///< 工程哈希（8 字节）
|
||
constexpr size_t kOffEntryFnId = 24; ///< 入口函数 fn_id
|
||
constexpr size_t kOffNGlobals = 28; ///< 全局槽数
|
||
constexpr size_t kOffNI = 32; ///< 输入（I）槽数
|
||
constexpr size_t kOffNQ = 36; ///< 输出（Q）槽数
|
||
constexpr size_t kOffNM = 40; ///< 中间（M）槽数
|
||
constexpr size_t kOffNConsts = 44; ///< 常量表条目数
|
||
constexpr size_t kOffNFuncs = 48; ///< 函数表行数
|
||
constexpr size_t kOffConst = 52; ///< 常量表段偏移
|
||
constexpr size_t kOffFuncs = 56; ///< 函数表段偏移
|
||
constexpr size_t kOffCode = 60; ///< 字节码段偏移
|
||
constexpr size_t kOffFb = 64; ///< FB 表段偏移
|
||
constexpr size_t kOffData = 68; ///< 数据段偏移
|
||
constexpr size_t kOffModelId = 72; ///< 型号标识[32]（12.13 修订）
|
||
|
||
// ---- 段内行偏移 ----
|
||
constexpr size_t kConstTagOff = 0; ///< 常量表行内：tag
|
||
constexpr size_t kConstValueOff = 4; ///< 常量表行内：value（8 字节）
|
||
constexpr size_t kFuncNregsOff = 0; ///< 函数表行内：nregs
|
||
constexpr size_t kFuncCodeOff = 4; ///< 函数表行内：code_offset（字节）
|
||
constexpr size_t kFuncLenOff = 8; ///< 函数表行内：code_len（指令条数）
|
||
|
||
// 防漂移：头布局自洽性（与 Doc/isa/指令与映像.md 冻结表核对）
|
||
static_assert(kHeaderSize == kOffModelId + kModelIdSize, "header size = 72 + model id[32]");
|
||
static_assert(kOffData > kOffFb && kOffFb > kOffCode && kOffCode > kOffFuncs &&
|
||
kOffFuncs > kOffConst,
|
||
"header field offsets monotonic");
|
||
static_assert(kOffModelId == kOffData + 4, "model id follows data offset field");
|
||
static_assert(kConstEntrySize == kConstValueOff + 8, "const row: tag + value:8");
|
||
static_assert(kFuncRowSize == kFuncLenOff + 4, "func row: nregs/code_offset/code_len");
|
||
|
||
/// 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（执行器侧实现，与 compiler 各一份）----
|
||
|
||
/// 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 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;
|
||
}
|
||
}
|
||
}
|
||
|
||
/// 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};
|
||
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);
|
||
}
|
||
}
|
||
|
||
/// 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]);
|
||
}
|
||
};
|
||
|
||
/**
|
||
* @brief 一次性 SHA-256。
|
||
* @param data 输入字节
|
||
* @param len 字节数
|
||
* @param out 32 字节摘要输出
|
||
*/
|
||
void sha256(const uint8_t* data, size_t len, uint8_t out[32]) {
|
||
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"），余下补 '\0'。
|
||
* @param name 型号名
|
||
* @param version 版本号
|
||
* @param out 32 字节输出缓冲
|
||
*/
|
||
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
|
||
|
||
/**
|
||
* @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
|
||
*/
|
||
Image Image::from(const uint8_t* buf, size_t len) {
|
||
Image 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;
|
||
}
|
||
ImageHeader& h = v.hdr_;
|
||
h.cycle_limit = get_le32(buf + kOffCycleLimit);
|
||
h.dt_ms = get_le32(buf + kOffDtMs);
|
||
h.project_hash = get_le64(buf + kOffProjectHash);
|
||
h.entry_fn_id = get_le32(buf + kOffEntryFnId);
|
||
h.n_globals = get_le32(buf + kOffNGlobals);
|
||
h.n_i = get_le32(buf + kOffNI);
|
||
h.n_q = get_le32(buf + kOffNQ);
|
||
h.n_m = get_le32(buf + kOffNM);
|
||
h.n_consts = get_le32(buf + kOffNConsts);
|
||
h.n_funcs = get_le32(buf + kOffNFuncs);
|
||
h.offset_const = get_le32(buf + kOffConst);
|
||
h.offset_funcs = get_le32(buf + kOffFuncs);
|
||
h.offset_code = get_le32(buf + kOffCode);
|
||
h.offset_fb = get_le32(buf + kOffFb);
|
||
h.offset_data = get_le32(buf + kOffData);
|
||
|
||
// 段校验（数据段之后是 SHA-256 文件尾；这里仅保证有尾，sha_ok() 做完整性校验）
|
||
if (len < static_cast<size_t>(h.offset_data) + kSha256Size) {
|
||
v.err_ = "missing sha256 tail";
|
||
return v;
|
||
}
|
||
const uint64_t offs[5] = {h.offset_const, h.offset_funcs, h.offset_code, h.offset_fb,
|
||
h.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>(h.n_consts) * kConstEntrySize) {
|
||
v.err_ = "const table size mismatch";
|
||
return v;
|
||
}
|
||
if (offs[2] - offs[1] != static_cast<uint64_t>(h.n_funcs) * kFuncRowSize) {
|
||
v.err_ = "function table size mismatch";
|
||
return v;
|
||
}
|
||
if ((h.offset_fb - h.offset_code) % 4 != 0) {
|
||
v.err_ = "code segment not 4-byte aligned";
|
||
return v;
|
||
}
|
||
if (h.entry_fn_id >= h.n_funcs && h.n_funcs != 0) {
|
||
v.err_ = "entry fn_id out of range";
|
||
return v;
|
||
}
|
||
v.ok_ = true;
|
||
v.err_.clear();
|
||
return v;
|
||
}
|
||
|
||
/**
|
||
* @brief 从 vector 构造只读视图。
|
||
* @param buf 映像缓冲（引用 data()，调用方保证生命周期）
|
||
* @return 同 from(const uint8_t*, size_t)
|
||
*/
|
||
Image Image::from(const std::vector<uint8_t>& buf) {
|
||
return from(buf.data(), buf.size());
|
||
}
|
||
|
||
/**
|
||
* @brief 读函数表一行。
|
||
* @param i 函数下标
|
||
* @return 该行字段；越界（或校验失败）时全 0
|
||
*/
|
||
FuncRow Image::func_row(size_t i) const {
|
||
FuncRow r;
|
||
if (ok_ && i < hdr_.n_funcs) {
|
||
const uint8_t* p = buf_ + hdr_.offset_funcs + i * kFuncRowSize;
|
||
r.nregs = get_le32(p + kFuncNregsOff);
|
||
r.code_offset = get_le32(p + kFuncCodeOff);
|
||
r.code_len = get_le32(p + kFuncLenOff);
|
||
}
|
||
return r;
|
||
}
|
||
|
||
/**
|
||
* @brief 读常量表一项。
|
||
* @param i 常量下标
|
||
* @return 该条目；越界（或校验失败）时全 0
|
||
*/
|
||
ConstEntry Image::const_entry(size_t i) const {
|
||
ConstEntry e;
|
||
if (ok_ && i < hdr_.n_consts) {
|
||
const uint8_t* p = buf_ + hdr_.offset_const + i * kConstEntrySize;
|
||
e.tag = get_le32(p + kConstTagOff);
|
||
e.value = get_le64(p + kConstValueOff);
|
||
}
|
||
return e;
|
||
}
|
||
|
||
/**
|
||
* @brief 字节码段起点。
|
||
* @return 段指针；校验失败返回 nullptr
|
||
*/
|
||
const uint8_t* Image::code_bytes() const {
|
||
return ok_ ? buf_ + hdr_.offset_code : nullptr;
|
||
}
|
||
|
||
/**
|
||
* @brief 数据段起点。
|
||
* @return 段指针；校验失败返回 nullptr
|
||
*/
|
||
const uint8_t* Image::data_bytes() const {
|
||
return ok_ ? buf_ + hdr_.offset_data : nullptr;
|
||
}
|
||
|
||
/**
|
||
* @brief 数据段字节数。
|
||
* @return 段长度（不含 SHA-256 尾）；校验失败返回 0
|
||
*/
|
||
size_t Image::data_len() const {
|
||
return ok_ ? (len_ - kSha256Size) - hdr_.offset_data : 0;
|
||
}
|
||
|
||
/**
|
||
* @brief 型号标识字符串。
|
||
* @return 头 @72 起 32 字节，去 '\0' 截断；校验失败返回空串
|
||
*/
|
||
std::string Image::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;
|
||
}
|
||
|
||
/**
|
||
* @brief 型号标识是否匹配。
|
||
* @param name 期望型号名（如 "STATOR"）
|
||
* @param version 期望版本（如 1）
|
||
* @return 头内 32 字节与 "name + version" 补零后逐字节相等
|
||
*/
|
||
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_ + kOffModelId, want, kModelIdSize) == 0;
|
||
}
|
||
|
||
/**
|
||
* @brief 文件尾 SHA-256 完整性校验。
|
||
* @return 对 len-32 字节内容算摘要，与文件尾 32 字节相等
|
||
*/
|
||
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
|