Add x86_64-linux-gnu emulator

I wanted a tiny scriptable meltdown proof way to run userspace programs
and visualize how program execution impacts memory. It helps to explain
how things like Actually Portable Executable works. It can show you how
the GCC generated code is going about manipulating matrices and more. I
didn't feel fully comfortable with Qemu and Bochs because I'm not smart
enough to understand them. I wanted something like gVisor but with much
stronger levels of assurances. I wanted a single binary that'll run, on
all major operating systems with an embedded GPL barrier ZIP filesystem
that is tiny enough to transpile to JavaScript and run in browsers too.

https://justine.storage.googleapis.com/emulator625.mp4

libc/alg/alg.h

@@ -18,9 +18,8 @@ void qsort(void *, size_t, size_t, int (*)(const void *, const void *))
 void qsort_r(void *, size_t, size_t,
              int cmp(const void *, const void *, void *), void *arg)
     paramsnonnull((1, 4));
-int tarjan(uint32_t, const uint32_t (*)[2], uint32_t, uint32_t[], uint32_t[],
-           uint32_t *) paramsnonnull((2, 4)) nocallback nothrow;
-void heapsortcar(int32_t (*)[2], unsigned) paramsnonnull() nocallback nothrow;
+int tarjan(int, const int (*)[2], int, int[], int[], int *)
+    paramsnonnull((2, 4)) nocallback nothrow;
 
 void *memmem(const void *, size_t, const void *, size_t)
     paramsnonnull() nothrow nocallback nosideeffect;

libc/alg/arraylist2.h

@@ -19,6 +19,11 @@ COSMOPOLITAN_C_START_
     size_t SizE = sizeof(*Item);                                      \
     size_t Count = (COUNT);                                           \
     ssize_t Entry = -1;                                               \
+    /* NOTE: We use `<` to guarantee one additional slot */           \
+    /*       grow() will memset(0) extended memory, thus */           \
+    /*       you get a nul-terminator for free sometimes */           \
+    /*       the exception is if you list.i=0 and re-use */           \
+    /*       so you need concat(...); list.p[list.i++]=0 */           \
     if (*ListI + Count < *ListN || grow(ListP, ListN, SizE, Count)) { \
       memcpy(&(*ListP)[*ListI], Item, (SizE) * (Count));              \
       Entry = *ListI;                                                 \

libc/alg/qsort.c

@@ -71,25 +71,25 @@ static void cycle(size_t width, unsigned char *ar[], size_t n) {
 
 forceinline void shl(unsigned p[2], size_t n) {
   assert(n > 0);
-  if (n >= 8 * sizeof(unsigned)) {
-    n -= 8 * sizeof(unsigned);
+  if (n >= CHAR_BIT * sizeof(unsigned)) {
+    n -= CHAR_BIT * sizeof(unsigned);
     p[1] = p[0];
     p[0] = 0;
   }
   p[1] <<= n;
-  p[1] |= p[0] >> (sizeof(unsigned) * 8 - n);
+  p[1] |= p[0] >> (sizeof(unsigned) * CHAR_BIT - n);
   p[0] <<= n;
 }
 
 forceinline void shr(unsigned p[2], size_t n) {
   assert(n > 0);
-  if (n >= 8 * sizeof(unsigned)) {
-    n -= 8 * sizeof(unsigned);
+  if (n >= CHAR_BIT * sizeof(unsigned)) {
+    n -= CHAR_BIT * sizeof(unsigned);
     p[0] = p[1];
     p[1] = 0;
   }
   p[0] >>= n;
-  p[0] |= p[1] << (sizeof(unsigned) * 8 - n);
+  p[0] |= p[1] << (sizeof(unsigned) * CHAR_BIT - n);
   p[1] >>= n;
 }
 

libc/alg/replacestr16.c

@@ -23,7 +23,9 @@
 #include "libc/str/str.h"
 #include "libc/sysv/errfuns.h"
 
+#undef strlen
 #undef replacestr
 #define replacestr replacestr16
 #define char char16_t
+#define strlen strlen16
 #include "libc/alg/replacestr.c"

libc/alg/tarjan.c

@@ -19,8 +19,8 @@
 ╚─────────────────────────────────────────────────────────────────────────────*/
 #include "libc/alg/alg.h"
 #include "libc/assert.h"
-#include "libc/bits/safemacros.h"
 #include "libc/limits.h"
+#include "libc/macros.h"
 #include "libc/mem/mem.h"
 
 /**
@@ -33,79 +33,77 @@
  *     topological sorting as a byproduct.” ──D.E. Knuth
  */
 
-struct Vertex {
-  uint32_t Vi;
-  uint32_t Ei;
-  uint32_t index;
-  uint32_t lowlink;
-  bool onstack;
-  bool selfreferential;
-};
-
-struct TarjanStack {
-  size_t i;
-  size_t n;
-  uint32_t *p;
-};
-
 struct Tarjan {
-  uint32_t Vn;
-  uint32_t En;
-  struct Vertex *V;
-  const uint32_t (*E)[2];
-  uint32_t *R;
-  uint32_t *C;
-  uint32_t Ci;
-  uint32_t Ri;
-  uint32_t index;
-  struct TarjanStack S;
+  int Vn, En, Ci, Ri, *R, *C, index;
+  const int (*E)[2];
+  struct Vertex {
+    int Vi;
+    int Ei;
+    int index;
+    int lowlink;
+    bool onstack;
+    bool selfreferential;
+  } * V;
+  struct TarjanStack {
+    int i;
+    int n;
+    int *p;
+  } S;
 };
 
-static uint32_t TarjanPush(struct TarjanStack *st, uint32_t Vi) {
-  if (st->i < st->n || grow(&st->p, &st->n, sizeof(uint32_t), 0)) {
-    return (st->p[st->i++] = Vi);
-  } else {
-    return -1;
+static bool TarjanPush(struct Tarjan *t, int v) {
+  int *q;
+  assert(t->S.i >= 0);
+  assert(t->S.n >= 0);
+  assert(0 <= v && v < t->Vn);
+  if (t->S.i == t->S.n) {
+    if ((q = realloc(t->S.p, (t->S.n + (t->S.n >> 1) + 8) * sizeof(*t->S.p)))) {
+      t->S.p = q;
+    } else {
+      return false;
+    }
   }
+  t->S.p[t->S.i++] = v;
+  return true;
 }
 
-static uint32_t TarjanPop(struct TarjanStack *st) {
-  assert(st->i != 0);
-  return st->p[--st->i];
+static int TarjanPop(struct Tarjan *t) {
+  assert(t->S.i > 0);
+  return t->S.p[--t->S.i];
 }
 
-static int TarjanConnect(struct Tarjan **tj, uint32_t Vi) {
-  struct Vertex *v = &(*tj)->V[Vi];
-  v->index = (*tj)->index;
-  v->lowlink = (*tj)->index;
-  v->onstack = true;
-  (*tj)->index++;
-  if (TarjanPush(&(*tj)->S, Vi) == -1) return -1;
-  uint32_t fs = (*tj)->V[Vi].Ei;
+static bool TarjanConnect(struct Tarjan *t, int v) {
+  int fs, w, e;
+  assert(0 <= v && v < t->Vn);
+  t->V[v].index = t->index;
+  t->V[v].lowlink = t->index;
+  t->V[v].onstack = true;
+  t->index++;
+  if (!TarjanPush(t, v)) return false;
+  fs = t->V[v].Ei;
   if (fs != -1) {
-    for (uint32_t Ei = fs; Ei < (*tj)->En && Vi == (*tj)->E[Ei][0]; ++Ei) {
-      struct Vertex *w = &(*tj)->V[(*tj)->E[Ei][1]];
-      if (!w->index) {
-        if (TarjanConnect(tj, w->Vi) == -1) return -1;
-        v->lowlink = min(v->lowlink, w->lowlink);
-      } else if (w->onstack) {
-        v->lowlink = min(v->lowlink, w->index);
+    for (e = fs; e < t->En && v == t->E[e][0]; ++e) {
+      w = t->E[e][1];
+      if (!t->V[w].index) {
+        if (!TarjanConnect(t, t->V[w].Vi)) return false;
+        t->V[v].lowlink = MIN(t->V[v].lowlink, t->V[w].lowlink);
+      } else if (t->V[w].onstack) {
+        t->V[v].lowlink = MIN(t->V[v].lowlink, t->V[w].index);
       }
       if (w == v) {
-        w->selfreferential = true;
+        t->V[w].selfreferential = true;
       }
     }
   }
-  if (v->lowlink == v->index) {
-    struct Vertex *w;
+  if (t->V[v].lowlink == t->V[v].index) {
     do {
-      w = &(*tj)->V[TarjanPop(&(*tj)->S)];
-      w->onstack = false;
-      (*tj)->R[(*tj)->Ri++] = w->Vi;
+      w = TarjanPop(t);
+      t->V[w].onstack = false;
+      t->R[t->Ri++] = t->V[w].Vi;
     } while (w != v);
-    if ((*tj)->C) (*tj)->C[(*tj)->Ci++] = (*tj)->Ri;
+    if (t->C) t->C[t->Ci++] = t->Ri;
   }
-  return 0;
+  return true;
 }
 
 /**
@@ -133,51 +131,53 @@ static int TarjanConnect(struct Tarjan **tj, uint32_t Vi) {
  * @error ENOMEM
  * @note Tarjan's Algorithm is O(|V|+|E|)
  */
-int tarjan(uint32_t vertex_count, const uint32_t (*edges)[2],
-           uint32_t edge_count, uint32_t out_sorted[],
-           uint32_t out_opt_components[], uint32_t *out_opt_componentcount) {
-  assert(edge_count <= INT_MAX);
-  assert(vertex_count <= INT_MAX);
-  for (unsigned i = 0; i < edge_count; ++i) {
+int tarjan(int vertex_count, const int (*edges)[2], int edge_count,
+           int out_sorted[], int out_opt_components[],
+           int *out_opt_componentcount) {
+  int i, rc, v, e;
+  struct Tarjan *t;
+  assert(0 <= edge_count && edge_count <= INT_MAX);
+  assert(0 <= vertex_count && vertex_count <= INT_MAX);
+  for (i = 0; i < edge_count; ++i) {
     if (i) assert(edges[i - 1][0] <= edges[i][0]);
     assert(edges[i][0] < vertex_count);
     assert(edges[i][1] < vertex_count);
   }
-  int rc;
-  struct Tarjan *tj;
-  if ((tj = calloc(1, (sizeof(struct Tarjan) +
+  if (!(t = calloc(1, (sizeof(struct Tarjan) +
                        sizeof(struct Vertex) * vertex_count)))) {
-    tj->V = (struct Vertex *)((char *)tj + sizeof(struct Tarjan));
-    tj->Vn = vertex_count;
-    tj->E = edges;
-    tj->En = edge_count;
-    tj->R = out_sorted;
-    tj->C = out_opt_components;
-    tj->index = 1;
-    uint32_t Vi, Ei;
-    for (Vi = 0; Vi < tj->Vn; ++Vi) {
-      tj->V[Vi].Vi = Vi;
-      tj->V[Vi].Ei = -1u;
-    }
-    for (Ei = 0, Vi = -1u; Ei < tj->En; ++Ei) {
-      if (tj->E[Ei][0] == Vi) continue;
-      Vi = tj->E[Ei][0];
-      tj->V[Vi].Ei = Ei;
-    }
-    rc = 0;
-    for (Vi = 0; Vi < tj->Vn; ++Vi) {
-      if (!tj->V[Vi].index) {
-        if ((rc = TarjanConnect(&tj, Vi)) == -1) {
-          break;
-        }
+    return -1;
+  }
+  t->V = (struct Vertex *)((char *)t + sizeof(struct Tarjan));
+  t->Vn = vertex_count;
+  t->E = edges;
+  t->En = edge_count;
+  t->R = out_sorted;
+  t->C = out_opt_components;
+  t->index = 1;
+  for (v = 0; v < t->Vn; ++v) {
+    t->V[v].Vi = v;
+    t->V[v].Ei = -1;
+  }
+  for (e = 0, v = -1; e < t->En; ++e) {
+    if (t->E[e][0] == v) continue;
+    v = t->E[e][0];
+    t->V[v].Ei = e;
+  }
+  rc = 0;
+  for (v = 0; v < t->Vn; ++v) {
+    if (!t->V[v].index) {
+      if (!TarjanConnect(t, v)) {
+        free(t->S.p);
+        free(t);
+        return -1;
       }
     }
-    free(tj->S.p);
-    assert(tj->Ri == vertex_count);
-    if (out_opt_components) *out_opt_componentcount = tj->Ci;
-  } else {
-    rc = -1;
   }
-  free(tj);
+  if (out_opt_components) {
+    *out_opt_componentcount = t->Ci;
+  }
+  assert(t->Ri == vertex_count);
+  free(t->S.p);
+  free(t);
   return rc;
 }