1 files changed, 52 insertions, 10 deletions

diff --git a/dynasm/dasm_x86.h b/dynasm/dasm_x86.h
index c3089d91..f0327302 100644
--- a/dynasm/dasm_x86.h
+++ b/dynasm/dasm_x86.h
@@ -170,7 +170,7 @@ void dasm_put(Dst_DECL, int start, ...)
   dasm_State *D = Dst_REF;
   dasm_ActList p = D->actionlist + start;
   dasm_Section *sec = D->section;
-  int pos = sec->pos, ofs = sec->ofs, mrm = 4;
+  int pos = sec->pos, ofs = sec->ofs, mrm = -1;
   int *b;
 
   if (pos >= sec->epos) {
@@ -193,7 +193,7 @@ void dasm_put(Dst_DECL, int start, ...)
       b[pos++] = n;
       switch (action) {
       case DASM_DISP:
-        if (n == 0) { if ((mrm&7) == 4) mrm = p[-2]; if ((mrm&7) != 5) break; }
+        if (n == 0) { if (mrm < 0) mrm = p[-2]; if ((mrm&7) != 5) break; }
         /* fallthrough */
       case DASM_IMM_DB: if (((n+128)&-256) == 0) goto ob; /* fallthrough */
       case DASM_REL_A: /* Assumes ptrdiff_t is int. !x64 */
@@ -204,11 +204,17 @@ void dasm_put(Dst_DECL, int start, ...)
       case DASM_IMM_W: CK((n&-65536) == 0, RANGE_I); ofs += 2; break;
       case DASM_SPACE: p++; ofs += n; break;
       case DASM_SETLABEL: b[pos-2] = -0x40000000; break;  /* Neg. label ofs. */
-      case DASM_VREG: CK((n&-8) == 0 && (n != 4 || (*p&1) == 0), RANGE_VREG);
-        if (*p++ == 1 && *p == DASM_DISP) mrm = n;
+      case DASM_VREG: CK((n&-16) == 0 && (n != 4 || (*p>>5) != 2), RANGE_VREG);
+        if (*p < 0x40 && p[1] == DASM_DISP) mrm = n;
+        if (*p < 0x20 && (n&7) == 4) ofs++;
+        switch ((*p++ >> 3) & 3) {
+        case 3: n |= b[pos-3]; /* fallthrough */
+        case 2: n |= b[pos-2]; /* fallthrough */
+        case 1: if (n <= 7) { b[pos-1] |= 0x10; ofs--; }
+        }
         continue;
       }
-      mrm = 4;
+      mrm = -1;
     } else {
       int *pl, n;
       switch (action) {
@@ -233,8 +239,11 @@ void dasm_put(Dst_DECL, int start, ...)
         }
         pos++;
         ofs += 4;  /* Maximum offset needed. */
-        if (action == DASM_REL_LG || action == DASM_REL_PC)
+        if (action == DASM_REL_LG || action == DASM_REL_PC) {
           b[pos++] = ofs;  /* Store pass1 offset estimate. */
+        } else if (sizeof(ptrdiff_t) == 8) {
+          ofs += 4;
+        }
         break;
       case DASM_LABEL_LG: pl = D->lglabels + *p++; CKPL(lg, LG); goto putlabel;
       case DASM_LABEL_PC: pl = D->pclabels + va_arg(ap, int); CKPL(pc, PC);
@@ -359,10 +368,22 @@ int dasm_link(Dst_DECL, size_t *szp)
   do { *((unsigned short *)cp) = (unsigned short)(x); cp+=2; } while (0)
 #define dasmd(x) \
   do { *((unsigned int *)cp) = (unsigned int)(x); cp+=4; } while (0)
+#define dasmq(x) \
+  do { *((unsigned long long *)cp) = (unsigned long long)(x); cp+=8; } while (0)
 #else
 #define dasmw(x)        do { dasmb(x); dasmb((x)>>8); } while (0)
 #define dasmd(x)        do { dasmw(x); dasmw((x)>>16); } while (0)
+#define dasmq(x)        do { dasmd(x); dasmd((x)>>32); } while (0)
 #endif
+static unsigned char *dasma_(unsigned char *cp, ptrdiff_t x)
+{
+  if (sizeof(ptrdiff_t) == 8)
+    dasmq((unsigned long long)x);
+  else
+    dasmd((unsigned int)x);
+  return cp;
+}
+#define dasma(x)        (cp = dasma_(cp, (x)))
 
 /* Pass 3: Encode sections. */
 int dasm_encode(Dst_DECL, void *buffer)
@@ -402,7 +423,27 @@ int dasm_encode(Dst_DECL, void *buffer)
         case DASM_IMM_WB: if (((n+128)&-256) == 0) goto db; else mark = NULL;
           /* fallthrough */
         case DASM_IMM_W: dasmw(n); break;
-        case DASM_VREG: { int t = *p++; if (t >= 2) n<<=3; cp[-1] |= n; break; }
+        case DASM_VREG: {
+          int t = *p++;
+          unsigned char *ex = cp - (t&7);
+          if ((n & 8) && t < 0xa0) {
+            if (*ex & 0x80) ex[1] ^= 0x20 << (t>>6); else *ex ^= 1 << (t>>6);
+            n &= 7;
+          } else if (n & 0x10) {
+            if (*ex & 0x80) {
+              *ex = 0xc5; ex[1] = (ex[1] & 0x80) | ex[2]; ex += 2;
+            }
+            while (++ex < cp) ex[-1] = *ex;
+            if (mark) mark--;
+            cp--;
+            n &= 7;
+          }
+          if (t >= 0xc0) n <<= 4;
+          else if (t >= 0x40) n <<= 3;
+          else if (n == 4 && t < 0x20) { cp[-1] ^= n; *cp++ = 0x20; }
+          cp[-1] ^= n;
+          break;
+        }
         case DASM_REL_LG: p++; if (n >= 0) goto rel_pc;
           b++; n = (int)(ptrdiff_t)D->globals[-n];
           /* fallthrough */
@@ -417,12 +458,13 @@ int dasm_encode(Dst_DECL, void *buffer)
           goto wb;
         }
         case DASM_IMM_LG:
-          p++; if (n < 0) { n = (int)(ptrdiff_t)D->globals[-n]; goto wd; }
+          p++;
+          if (n < 0) { dasma((ptrdiff_t)D->globals[-n]); break; }
           /* fallthrough */
         case DASM_IMM_PC: {
           int *pb = DASM_POS2PTR(D, n);
-          n = *pb < 0 ? pb[1] : (*pb + (int)(ptrdiff_t)base);
-          goto wd;
+          dasma(*pb < 0 ? (ptrdiff_t)pb[1] : (*pb + (ptrdiff_t)base));
+          break;
         }
         case DASM_LABEL_LG: {
           int idx = *p++;