From ed37aa8070cabf80de269508cbf022bc2b464d26 Mon Sep 17 00:00:00 2001
From: "H. Peter Anvin" <hpa@zytor.com>
Date: Wed, 18 Mar 2009 23:10:19 -0700
Subject: [PATCH] disasm: when no instruction is found, consider a naked prefix

If we can't find a matching instruction, rather than printing it as a
"db" literal, consider first if we can disassemble it as a naked prefix.
---
 disasm.c  | 82 +++++++++++++++++++++++++++++++++++++++++++++++++++++--
 disasm.h  |  2 +-
 ndisasm.c |  9 +++---
 3 files changed, 86 insertions(+), 7 deletions(-)

diff --git a/disasm.c b/disasm.c
index 63fd37be..f4104186 100644
--- a/disasm.c
+++ b/disasm.c
@@ -1420,8 +1420,86 @@ int32_t disasm(uint8_t *data, char *output, int outbufsize, int segsize,
     return length;
 }
 
-int32_t eatbyte(uint8_t *data, char *output, int outbufsize)
+/*
+ * This is called when we don't have a complete instruction.  If it
+ * is a standalone *single-byte* prefix show it as such, otherwise
+ * print it as a literal.
+ */
+int32_t eatbyte(uint8_t *data, char *output, int outbufsize, int segsize)
 {
-    snprintf(output, outbufsize, "db 0x%02X", *data);
+    uint8_t byte = *data;
+    const char *str = NULL;
+    
+    switch (byte) {
+    case 0xF2:
+	str = "rep";
+	break;
+    case 0xF3:
+	str = "repne";
+	break;
+    case 0x9B:
+	str = "wait";
+	break;
+    case 0xF0:
+	str = "lock";
+	break;
+    case 0x2E:
+	str = "cs";
+	break;
+    case 0x36:
+	str = "ss";
+	break;
+    case 0x3E:
+	str = "ss";
+	break;
+    case 0x26:
+	str = "es";
+	break;
+    case 0x64:
+	str = "fs";
+	break;
+    case 0x65:
+	str = "gs";
+	break;
+    case 0x66:
+	str = (segsize == 16) ? "o32" : "o16";
+	break;
+    case 0x67:
+	str = (segsize == 32) ? "a16" : "a32";
+	break;
+    case REX_P + 0x0:
+    case REX_P + 0x1:
+    case REX_P + 0x2:
+    case REX_P + 0x3:
+    case REX_P + 0x4:
+    case REX_P + 0x5:
+    case REX_P + 0x6:
+    case REX_P + 0x7:
+    case REX_P + 0x8:
+    case REX_P + 0x9:
+    case REX_P + 0xA:
+    case REX_P + 0xB:
+    case REX_P + 0xC:
+    case REX_P + 0xD:
+    case REX_P + 0xE:
+    case REX_P + 0xF:
+	if (segsize == 64) {
+	    snprintf(output, outbufsize, "rex%s%s%s%s%s",
+		     (byte == REX_P) ? "" : ".",
+		     (byte & REX_W) ? "w" : "",
+		     (byte & REX_R) ? "r" : "",
+		     (byte & REX_X) ? "x" : "",
+		     (byte & REX_B) ? "b" : "");
+	    break;
+	}
+	/* else fall through */
+    default:
+	snprintf(output, outbufsize, "db 0x%02x", byte);
+	break;
+    }
+
+    if (str)
+	strcpy(output, str);
+
     return 1;
 }
diff --git a/disasm.h b/disasm.h
index 56785ce6..6467730a 100644
--- a/disasm.h
+++ b/disasm.h
@@ -13,6 +13,6 @@
 
 int32_t disasm(uint8_t *data, char *output, int outbufsize, int segsize,
             int32_t offset, int autosync, uint32_t prefer);
-int32_t eatbyte(uint8_t *data, char *output, int outbufsize);
+int32_t eatbyte(uint8_t *data, char *output, int outbufsize, int segsize);
 
 #endif
diff --git a/ndisasm.c b/ndisasm.c
index 766119b9..75fc3256 100644
--- a/ndisasm.c
+++ b/ndisasm.c
@@ -273,7 +273,8 @@ int main(int argc, char **argv)
         if ((nextsync || synclen) &&
 	    (uint32_t)offset == nextsync) {
             if (synclen) {
-                fprintf(stdout, "%08"PRIX32"  skipping 0x%"PRIX32" bytes\n", offset, synclen);
+                fprintf(stdout, "%08"PRIX32"  skipping 0x%"PRIX32" bytes\n",
+			offset, synclen);
                 offset += synclen;
                 skip(synclen, fp);
             }
@@ -282,12 +283,12 @@ int main(int argc, char **argv)
         }
         while (p > q && (p - q >= INSN_MAX || lenread == 0)) {
             lendis =
-                disasm((uint8_t *) q, outbuf, sizeof(outbuf), bits, offset, autosync,
-                       prefer);
+                disasm((uint8_t *) q, outbuf, sizeof(outbuf), bits,
+		       offset, autosync, prefer);
             if (!lendis || lendis > (p - q)
                 || ((nextsync || synclen) &&
 		    (uint32_t)lendis > nextsync - offset))
-                lendis = eatbyte((uint8_t *) q, outbuf, sizeof(outbuf));
+                lendis = eatbyte((uint8_t *) q, outbuf, sizeof(outbuf), bits);
             output_ins(offset, (uint8_t *) q, lendis, outbuf);
             q += lendis;
             offset += lendis;