Remove all remaining explicit bytecodes from insns.dat

Get rid of the last vestiges of the explicit byte codes in insns.dat.
The only files that now depend on actual byte code numbers are
insns.pl, assemble.c and disasm.c.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>

diff --git a/assemble.c b/assemble.c
index 5a7bd1e..eba1b32 100644 (file)
--- a/assemble.c
+++ b/assemble.c
@@ -127,8 +127,8 @@
  * \333          - REP prefix (0xF3 byte) used as opcode extension.
  * \334          - LOCK prefix used as REX.R (used in non-64-bit mode)
  * \335          - disassemble a rep (0xF3 byte) prefix as repe not rep.
- * \336          - force a REP(E) prefix (0xF2) even if not specified.
- * \337          - force a REPNE prefix (0xF3) even if not specified.
+ * \336          - force a REP(E) prefix (0xF3) even if not specified.
+ * \337          - force a REPNE prefix (0xF2) even if not specified.
  *                 \336-\337 are still listed as prefixes in the disassembler.
  * \340          - reserve <operand 0> bytes of uninitialized storage.
  *                 Operand 0 had better be a segmentless constant.
@@ -145,7 +145,7 @@
  * \365          - address-size prefix (0x67) not permitted
  * \366          - operand-size prefix (0x66) used as opcode extension
  * \367          - address-size prefix (0x67) used as opcode extension
- * \370,\371,\372 - match only if operand 0 meets byte jump criteria.
+ * \370,\371     - match only if operand 0 meets byte jump criteria.
  *                 370 is used for Jcc, 371 is used for JMP.
  * \373          - assemble 0x03 if bits==16, 0x05 if bits==32;
  *                 used for conditional jump over longer jump
@@ -326,7 +326,7 @@ static bool jmp_match(int32_t segment, int64_t offset, int bits,
     const uint8_t *code = temp->code;
     uint8_t c = code[0];
 
-    if ((c != 0370 && c != 0371) || (ins->oprs[0].type & STRICT))
+    if (((c & ~1) != 0370) || (ins->oprs[0].type & STRICT))
         return false;
     if (!optimizing)
         return false;
@@ -1831,7 +1831,6 @@ static void gencode(int32_t segment, int64_t offset, int bits,
 
         case 0370:
         case 0371:
-        case 0372:
             break;
 
         case 0373:
@@ -2250,7 +2249,7 @@ static enum match_result matches(const struct itemplate *itemp,
     /*
      * Check if special handling needed for Jumps
      */
-    if ((itemp->code[0] & 0374) == 0370)
+    if ((itemp->code[0] & ~1) == 0370)
         return MOK_JUMP;
 
     return MOK_GOOD;

diff --git a/insns.dat b/insns.dat
index 9df17d0..8001681 100644 (file)
--- a/insns.dat
+++ b/insns.dat
@@ -55,7 +55,7 @@ DQ            ignore                          ignore                                          ignore
 DT             ignore                          ignore                                          ignore
 DO             ignore                          ignore                                          ignore
 DY             ignore                          ignore                                          ignore
-RESB           imm                             \340                                            8086
+RESB           imm                             [       resb]                                   8086
 RESW           ignore                          ignore                                          ignore
 RESD           ignore                          ignore                                          ignore
 RESQ           ignore                          ignore                                          ignore
@@ -338,8 +338,8 @@ DIV         rm64                            [m:     o64 f7 /6]                              X64
 DMINT          void                            [       0f 39]                                  P6,CYRIX
 EMMS           void                            [       0f 77]                                  PENT,MMX
 ENTER          imm,imm                         [ij:    c8 iw ib,u]                             186
-EQU            imm                             \0                                              8086
-EQU            imm:imm                         \0                                              8086
+EQU            imm                             ignore                                          8086
+EQU            imm:imm                         ignore                                          8086
 F2XM1          void                            [       d9 f0]                                  8086,FPU
 FABS           void                            [       d9 e1]                                  8086,FPU
 FADD           mem32                           [m:     d8 /0]                                  8086,FPU
@@ -652,7 +652,7 @@ JCXZ                imm                             [i:     a16 e3 rel8]                            8086,NOLONG
 JECXZ          imm                             [i:     a32 e3 rel8]                            386
 JRCXZ          imm                             [i:     a64 e3 rel8]                            X64
 JMP            imm|short                       [i:     eb rel8]                                8086
-JMP            imm                             [i:     \371 eb rel8]                           8086,ND
+JMP            imm                             [i:     jmp8 eb rel8]                           8086,ND
 JMP            imm                             [i:     odf e9 rel]                             8086
 JMP            imm|near                        [i:     odf e9 rel]                             8086,ND
 JMP            imm|far                         [i:     odf ea iwd seg]                         8086,ND,NOLONG
@@ -730,23 +730,23 @@ LODSB             void                            [       ac]                                     8086
 LODSD          void                            [       o32 ad]                                 386
 LODSQ          void                            [       o64 ad]                                 X64
 LODSW          void                            [       o16 ad]                                 8086
-LOOP           imm                             [i:     \312 e2 rel8]                           8086
+LOOP           imm                             [i:     adf e2 rel8]                            8086
 LOOP           imm,reg_cx                      [i-:    a16 e2 rel8]                            8086,NOLONG
 LOOP           imm,reg_ecx                     [i-:    a32 e2 rel8]                            386
 LOOP           imm,reg_rcx                     [i-:    a64 e2 rel8]                            X64
-LOOPE          imm                             [i:     \312 e1 rel8]                           8086
+LOOPE          imm                             [i:     adf e1 rel8]                            8086
 LOOPE          imm,reg_cx                      [i-:    a16 e1 rel8]                            8086,NOLONG
 LOOPE          imm,reg_ecx                     [i-:    a32 e1 rel8]                            386
 LOOPE          imm,reg_rcx                     [i-:    a64 e1 rel8]                            X64
-LOOPNE         imm                             [i:     \312 e0 rel8]                           8086
+LOOPNE         imm                             [i:     adf e0 rel8]                            8086
 LOOPNE         imm,reg_cx                      [i-:    a16 e0 rel8]                            8086,NOLONG
 LOOPNE         imm,reg_ecx                     [i-:    a32 e0 rel8]                            386
 LOOPNE         imm,reg_rcx                     [i-:    a64 e0 rel8]                            X64
-LOOPNZ         imm                             [i:     \312 e0 rel8]                           8086
+LOOPNZ         imm                             [i:     adf e0 rel8]                            8086
 LOOPNZ         imm,reg_cx                      [i-:    a16 e0 rel8]                            8086,NOLONG
 LOOPNZ         imm,reg_ecx                     [i-:    a32 e0 rel8]                            386
 LOOPNZ         imm,reg_rcx                     [i-:    a64 e0 rel8]                            X64
-LOOPZ          imm                             [i:     \312 e1 rel8]                           8086
+LOOPZ          imm                             [i:     adf e1 rel8]                            8086
 LOOPZ          imm,reg_cx                      [i-:    a16 e1 rel8]                            8086,NOLONG
 LOOPZ          imm,reg_ecx                     [i-:    a32 e1 rel8]                            386
 LOOPZ          imm,reg_rcx                     [i-:    a64 e1 rel8]                            X64
@@ -868,7 +868,7 @@ NEG         rm8                             [m:     f6 /3]                                  8086
 NEG            rm16                            [m:     o16 f7 /3]                              8086
 NEG            rm32                            [m:     o32 f7 /3]                              386
 NEG            rm64                            [m:     o64 f7 /3]                              X64
-NOP            void                            [       \314 90]                                8086
+NOP            void                            [       norexb 90]                              8086
 NOP            rm16                            [m:     o16 0f 1f /0]                           P6
 NOP            rm32                            [m:     o32 0f 1f /0]                           P6
 NOP            rm64                            [m:     o64 0f 1f /0]                           X64
@@ -931,7 +931,7 @@ PADDUSW             mmxreg,mmxrm                    [rm:    np o64nw 0f dd /r]                      PENT,MMX,SQ
 PADDW          mmxreg,mmxrm                    [rm:    np o64nw 0f fd /r]                      PENT,MMX,SQ
 PAND           mmxreg,mmxrm                    [rm:    np o64nw 0f db /r]                      PENT,MMX,SQ
 PANDN          mmxreg,mmxrm                    [rm:    np o64nw 0f df /r]                      PENT,MMX,SQ
-PAUSE          void                            [       \314 \333 90]                           8086
+PAUSE          void                            [       norexb f3i 90]                          8086
 PAVEB          mmxreg,mmxrm                    [rm:    o64nw 0f 50 /r]                         PENT,MMX,SQ,CYRIX
 PAVGUSB                mmxreg,mmxrm                    [rm:    o64nw 0f 0f /r bf]                      PENT,3DNOW,SQ
 PCMPEQB                mmxreg,mmxrm                    [rm:    np o64nw 0f 74 /r]                      PENT,MMX,SQ
@@ -977,8 +977,8 @@ POP         rm16                            [m:     o16 8f /0]                              8086
 POP            rm32                            [m:     o32 8f /0]                              386,NOLONG
 POP            rm64                            [m:     o64nw 8f /0]                            X64
 POP            reg_cs                          [-:     0f]                                     8086,UNDOC,ND
-POP            reg_dess                        [-:     \345]                                   8086,NOLONG
-POP            reg_fsgs                        [-:     0f \347]                                386
+POP            reg_dess                        [-:     popseg]                                 8086,NOLONG
+POP            reg_fsgs                        [-:     0f popseg2]                             386
 POPA           void                            [       odf 61]                                 186,NOLONG
 POPAD          void                            [       o32 61]                                 386,NOLONG
 POPAW          void                            [       o16 61]                                 186,NOLONG
@@ -1025,9 +1025,9 @@ PUSH              reg64                           [r:     o64nw 50+r]                             X64
 PUSH           rm16                            [m:     o16 ff /6]                              8086
 PUSH           rm32                            [m:     o32 ff /6]                              386,NOLONG
 PUSH           rm64                            [m:     o64nw ff /6]                            X64
-PUSH           reg_cs                          [-:     \344]                                   8086,NOLONG
-PUSH           reg_dess                        [-:     \344]                                   8086,NOLONG
-PUSH           reg_fsgs                        [-:     0f \346]                                386
+PUSH           reg_cs                          [-:     pushseg]                                8086,NOLONG
+PUSH           reg_dess                        [-:     pushseg]                                8086,NOLONG
+PUSH           reg_fsgs                        [-:     0f pushseg2]                            386
 PUSH           imm8                            [i:     6a ibx]                                 186
 PUSH           imm16                           [i:     o16 68+s ibw]                           186,AR0,SZ
 PUSH           imm32                           [i:     o32 68+s ibd]                           386,NOLONG,AR0,SZ
@@ -1421,9 +1421,9 @@ Jcc               imm|near                        [i:     odf 0f 80+c rel]                        386
 Jcc            imm16|near                      [i:     o16 0f 80+c rel]                        386
 Jcc            imm32|near                      [i:     o32 0f 80+c rel]                        386
 Jcc            imm|short                       [i:     70+c rel8]                              8086,ND
-Jcc            imm                             [i:     \370 70+c rel8]                         8086,ND
+Jcc            imm                             [i:     jcc8 70+c rel8]                         8086,ND
 Jcc            imm                             [i:     0f 80+c rel]                            386,ND
-Jcc            imm                             [i:     71+c \373 e9 rel]                       8086,ND
+Jcc            imm                             [i:     71+c jlen e9 rel]                       8086,ND
 Jcc            imm                             [i:     70+c rel8]                              8086
 SETcc          mem                             [m:     0f 90+c /0]                             386,SB
 SETcc          reg8                            [m:     0f 90+c /0]                             386
@@ -1575,10 +1575,10 @@ LFENCE          void                            [       0f ae e8]                               WILLAMETTE,SSE2
 MFENCE         void                            [       0f ae f0]                               WILLAMETTE,SSE2
 
 ;# Willamette MMX instructions (SSE2 SIMD Integer Instructions)
-MOVD           mem,xmmreg                      [mr:    66 \317 0f 7e /r]                       WILLAMETTE,SSE2,SD
-MOVD           xmmreg,mem                      [rm:    66 \317 0f 6e /r]                       WILLAMETTE,SSE2,SD
-MOVD           xmmreg,rm32                     [rm:    66 \317 0f 6e /r]                       WILLAMETTE,SSE2
-MOVD           rm32,xmmreg                     [mr:    66 \317 0f 7e /r]                       WILLAMETTE,SSE2
+MOVD           mem,xmmreg                      [mr:    66 norexw 0f 7e /r]                     WILLAMETTE,SSE2,SD
+MOVD           xmmreg,mem                      [rm:    66 norexw 0f 6e /r]                     WILLAMETTE,SSE2,SD
+MOVD           xmmreg,rm32                     [rm:    66 norexw 0f 6e /r]                     WILLAMETTE,SSE2
+MOVD           rm32,xmmreg                     [mr:    66 norexw 0f 7e /r]                     WILLAMETTE,SSE2
 MOVDQA         xmmreg,xmmreg                   [rm:    66 0f 6f /r]                            WILLAMETTE,SSE2
 MOVDQA         mem,xmmreg                      [mr:    66 0f 7f /r]                            WILLAMETTE,SSE2,SO
 MOVDQA         xmmreg,mem                      [rm:    66 0f 6f /r]                            WILLAMETTE,SSE2,SO
@@ -1845,9 +1845,9 @@ MOVNTSD           mem,xmmreg                      [mr:    f2 0f 2b /r]                            SSE4A,AMD,SQ
 MOVNTSS                mem,xmmreg                      [mr:    f3 0f 2b /r]                            SSE4A,AMD,SD
 
 ;# New instructions in Barcelona
-LZCNT          reg16,rm16                      [rm:    o16 \333 0f bd /r]                      P6,AMD
-LZCNT          reg32,rm32                      [rm:    o32 \333 0f bd /r]                      P6,AMD
-LZCNT          reg64,rm64                      [rm:    o64 \333 0f bd /r]                      X64,AMD
+LZCNT          reg16,rm16                      [rm:    o16 f3i 0f bd /r]                       P6,AMD
+LZCNT          reg32,rm32                      [rm:    o32 f3i 0f bd /r]                       P6,AMD
+LZCNT          reg64,rm64                      [rm:    o64 f3i 0f bd /r]                       X64,AMD
 
 ;# Penryn New Instructions (SSE4.1)
 BLENDPD                xmmreg,xmmrm,imm                [rmi:   66 0f 3a 0d /r ib,u]                    SSE41
@@ -1910,19 +1910,19 @@ ROUNDSD         xmmreg,xmmrm,imm                [rmi:   66 0f 3a 0b /r ib,u]                    SSE41
 ROUNDSS                xmmreg,xmmrm,imm                [rmi:   66 0f 3a 0a /r ib,u]                    SSE41
 
 ;# Nehalem New Instructions (SSE4.2)
-CRC32          reg32,rm8                       [rm:    \332 0f 38 f0 /r]                       SSE42
-CRC32          reg32,rm16                      [rm:    o16 \332 0f 38 f1 /r]                   SSE42
-CRC32          reg32,rm32                      [rm:    o32 \332 0f 38 f1 /r]                   SSE42
-CRC32          reg64,rm8                       [rm:    o64 \332 0f 38 f0 /r]                   SSE42,X64
-CRC32          reg64,rm64                      [rm:    o64 \332 0f 38 f1 /r]                   SSE42,X64
+CRC32          reg32,rm8                       [rm:    f2i 0f 38 f0 /r]                        SSE42
+CRC32          reg32,rm16                      [rm:    o16 f2i 0f 38 f1 /r]                    SSE42
+CRC32          reg32,rm32                      [rm:    o32 f2i 0f 38 f1 /r]                    SSE42
+CRC32          reg64,rm8                       [rm:    o64 f2i 0f 38 f0 /r]                    SSE42,X64
+CRC32          reg64,rm64                      [rm:    o64 f2i 0f 38 f1 /r]                    SSE42,X64
 PCMPESTRI      xmmreg,xmmrm,imm                [rmi:   66 0f 3a 61 /r ib,u]                    SSE42
 PCMPESTRM      xmmreg,xmmrm,imm                [rmi:   66 0f 3a 60 /r ib,u]                    SSE42
 PCMPISTRI      xmmreg,xmmrm,imm                [rmi:   66 0f 3a 63 /r ib,u]                    SSE42
 PCMPISTRM      xmmreg,xmmrm,imm                [rmi:   66 0f 3a 62 /r ib,u]                    SSE42
 PCMPGTQ                xmmreg,xmmrm                    [rm:    66 0f 38 37 /r]                         SSE42
-POPCNT         reg16,rm16                      [rm:    o16 \333 0f b8 /r]                      NEHALEM,SW
-POPCNT         reg32,rm32                      [rm:    o32 \333 0f b8 /r]                      NEHALEM,SD
-POPCNT         reg64,rm64                      [rm:    o64 \333 0f b8 /r]                      NEHALEM,SQ,X64
+POPCNT         reg16,rm16                      [rm:    o16 f3i 0f b8 /r]                       NEHALEM,SW
+POPCNT         reg32,rm32                      [rm:    o32 f3i 0f b8 /r]                       NEHALEM,SD
+POPCNT         reg64,rm64                      [rm:    o64 f3i 0f b8 /r]                       NEHALEM,SQ,X64
 
 ;# Intel SMX
 GETSEC         void                            [       0f 37]                                  KATMAI
@@ -2877,14 +2877,14 @@ VCVTPS2PH       xmmrm64,xmmreg,imm8             [mri:   vex.128.66.0f3a.w0 1d /r ib]            AVX,FUTURE
 
 ;# VIA (Centaur) security instructions
 XSTORE         void                            [       0f a7 c0]                               PENT,CYRIX
-XCRYPTECB      void                            [       \336 0f a7 c8]                          PENT,CYRIX
-XCRYPTCBC      void                            [       \336 0f a7 d0]                          PENT,CYRIX
-XCRYPTCTR      void                            [       \336 0f a7 d8]                          PENT,CYRIX
-XCRYPTCFB      void                            [       \336 0f a7 e0]                          PENT,CYRIX
-XCRYPTOFB      void                            [       \336 0f a7 e8]                          PENT,CYRIX
-MONTMUL                void                            [       \336 0f a6 c0]                          PENT,CYRIX
-XSHA1          void                            [       \336 0f a6 c8]                          PENT,CYRIX
-XSHA256                void                            [       \336 0f a6 d0]                          PENT,CYRIX
+XCRYPTECB      void                            [       mustrep 0f a7 c8]                       PENT,CYRIX
+XCRYPTCBC      void                            [       mustrep 0f a7 d0]                       PENT,CYRIX
+XCRYPTCTR      void                            [       mustrep 0f a7 d8]                       PENT,CYRIX
+XCRYPTCFB      void                            [       mustrep 0f a7 e0]                       PENT,CYRIX
+XCRYPTOFB      void                            [       mustrep 0f a7 e8]                       PENT,CYRIX
+MONTMUL                void                            [       mustrep 0f a6 c0]                       PENT,CYRIX
+XSHA1          void                            [       mustrep 0f a6 c8]                       PENT,CYRIX
+XSHA256                void                            [       mustrep 0f a6 d0]                       PENT,CYRIX
 
 ;# AMD Lightweight Profiling (LWP) instructions
 ;

diff --git a/insns.pl b/insns.pl
index 5ae4594..2ec891c 100755 (executable)
--- a/insns.pl
+++ b/insns.pl
@@ -527,18 +527,20 @@ sub decodify($$) {
     my $c = $codestr;
     my @codes = ();
 
-    while ($c ne '') {
-        if ($c =~ /^\\x([0-9a-f]+)(.*)$/i) {
-            push(@codes, hex $1);
-            $c = $2;
-            next;
-        } elsif ($c =~ /^\\([0-7]{1,3})(.*)$/) {
-            push(@codes, oct $1);
-            $c = $2;
-            next;
-        } else {
-            die "$fname: unknown code format in \"$codestr\"\n";
-        }
+    unless ($codestr eq 'ignore') {
+       while ($c ne '') {
+           if ($c =~ /^\\x([0-9a-f]+)(.*)$/i) {
+               push(@codes, hex $1);
+               $c = $2;
+               next;
+           } elsif ($c =~ /^\\([0-7]{1,3})(.*)$/) {
+               push(@codes, oct $1);
+               $c = $2;
+               next;
+           } else {
+               die "$fname: unknown code format in \"$codestr\"\n";
+           }
+       }
     }
 
     return @codes;
@@ -700,13 +702,30 @@ sub byte_code_compile($$) {
        'o64nw' => 0323,        # Implied 64-bit operand size (no REX.W)
        'a16' => 0310,
        'a32' => 0311,
+       'adf' => 0312,          # Address size is default
        'a64' => 0313,
        '!osp' => 0364,
        '!asp' => 0365,
+       'f2i' => 0332,          # F2 prefix, but 66 for operand size is OK
+       'f3i' => 0333,          # F3 prefix, but 66 for operand size is OK
+       'pushseg' => 0344,
+       'popseg' => 0345,
+       'pushseg2' => 0346,
+       'popseg2' => 0347,
+       'mustrep' => 0336,
+       'mustrepne' => 0337,
        'rex.l' => 0334,
+       'norexb' => 0314,
+       'norexx' => 0315,
+       'norexr' => 0316,
+       'norexw' => 0317,
        'repe' => 0335,
        'nohi' => 0325,         # Use spl/bpl/sil/dil even without REX
        'wait' => 0341,         # Needs a wait prefix
+       'resb' => 0340,
+       'jcc8' => 0370,         # Match only if Jcc possible with single byte
+       'jmp8' => 0371,         # Match only if JMP possible with single byte
+       'jlen' => 0373,         # Length of jump
        'nohle' => 0264,
        'hlexr' => 0265,
        'hlenl' => 0266,