Blame SOURCES/expat-2.1.0-CVE-2016-0718.patch

ed72f6
ed72f6
Patch as described below, plus a12e78cb1b290e26a2b669eff3664d76bb73a408.
ed72f6
ed72f6
>From cdfcb1b5c95e93b00ae9e9d25708b4a3bee72c15 Mon Sep 17 00:00:00 2001
ed72f6
From: Sebastian Pipping <sebastian@pipping.org>
ed72f6
Date: Mon, 2 May 2016 00:02:44 +0200
ed72f6
Subject: [PATCH] Address CVE-2016-0718 (/patch/ version 2.2.1)
ed72f6
ed72f6
* Out of bounds memory access when doing text conversion on malformed input
ed72f6
* Integer overflow related to memory allocation
ed72f6
ed72f6
Reported by Gustavo Grieco
ed72f6
ed72f6
Patch credits go to
ed72f6
* Christian Heimes
ed72f6
* Karl Waclawek
ed72f6
* Gustavo Grieco
ed72f6
* Sebastian Pipping
ed72f6
* Pascal Cuoq
ed72f6
---
ed72f6
 expat/lib/xmlparse.c    |  34 +++++++++-----
ed72f6
 expat/lib/xmltok.c      | 115 +++++++++++++++++++++++++++++++++++-------------
ed72f6
 expat/lib/xmltok.h      |  10 ++++-
ed72f6
 expat/lib/xmltok_impl.c |  62 +++++++++++++-------------
ed72f6
 4 files changed, 146 insertions(+), 75 deletions(-)
ed72f6
ed72f6
diff --git a/expat/lib/xmlparse.c b/expat/lib/xmlparse.c
ed72f6
index e308c79..13e080d 100644
ed72f6
diff -uap expat-2.1.0/lib/xmlparse.c.cve0718 expat-2.1.0/lib/xmlparse.c
ed72f6
--- expat-2.1.0/lib/xmlparse.c.cve0718
ed72f6
+++ expat-2.1.0/lib/xmlparse.c
ed72f6
@@ -2415,11 +2415,11 @@
ed72f6
           for (;;) {
ed72f6
             int bufSize;
ed72f6
             int convLen;
ed72f6
-            XmlConvert(enc,
ed72f6
+            const enum XML_Convert_Result convert_res = XmlConvert(enc,
ed72f6
                        &fromPtr, rawNameEnd,
ed72f6
                        (ICHAR **)&toPtr, (ICHAR *)tag->bufEnd - 1);
ed72f6
             convLen = (int)(toPtr - (XML_Char *)tag->buf);
ed72f6
-            if (fromPtr == rawNameEnd) {
ed72f6
+            if ((convert_res == XML_CONVERT_COMPLETED) || (convert_res == XML_CONVERT_INPUT_INCOMPLETE)) {
ed72f6
               tag->name.strLen = convLen;
ed72f6
               break;
ed72f6
             }
ed72f6
@@ -2640,11 +2640,11 @@
ed72f6
           if (MUST_CONVERT(enc, s)) {
ed72f6
             for (;;) {
ed72f6
               ICHAR *dataPtr = (ICHAR *)dataBuf;
ed72f6
-              XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd);
ed72f6
+              const enum XML_Convert_Result convert_res = XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd);
ed72f6
               *eventEndPP = s;
ed72f6
               charDataHandler(handlerArg, dataBuf,
ed72f6
                               (int)(dataPtr - (ICHAR *)dataBuf));
ed72f6
-              if (s == next)
ed72f6
+              if ((convert_res == XML_CONVERT_COMPLETED) || (convert_res == XML_CONVERT_INPUT_INCOMPLETE))
ed72f6
                 break;
ed72f6
               *eventPP = s;
ed72f6
             }
ed72f6
@@ -3248,11 +3248,11 @@
ed72f6
           if (MUST_CONVERT(enc, s)) {
ed72f6
             for (;;) {
ed72f6
               ICHAR *dataPtr = (ICHAR *)dataBuf;
ed72f6
-              XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd);
ed72f6
+              const enum XML_Convert_Result convert_res = XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd);
ed72f6
               *eventEndPP = next;
ed72f6
               charDataHandler(handlerArg, dataBuf,
ed72f6
                               (int)(dataPtr - (ICHAR *)dataBuf));
ed72f6
-              if (s == next)
ed72f6
+              if ((convert_res == XML_CONVERT_COMPLETED) || (convert_res == XML_CONVERT_INPUT_INCOMPLETE))
ed72f6
                 break;
ed72f6
               *eventPP = s;
ed72f6
             }
ed72f6
@@ -5329,6 +5329,7 @@
ed72f6
               const char *s, const char *end)
ed72f6
 {
ed72f6
   if (MUST_CONVERT(enc, s)) {
ed72f6
+    enum XML_Convert_Result convert_res;
ed72f6
     const char **eventPP;
ed72f6
     const char **eventEndPP;
ed72f6
     if (enc == encoding) {
ed72f6
@@ -5341,11 +5342,11 @@
ed72f6
     }
ed72f6
     do {
ed72f6
       ICHAR *dataPtr = (ICHAR *)dataBuf;
ed72f6
-      XmlConvert(enc, &s, end, &dataPtr, (ICHAR *)dataBufEnd);
ed72f6
+      convert_res = XmlConvert(enc, &s, end, &dataPtr, (ICHAR *)dataBufEnd);
ed72f6
       *eventEndPP = s;
ed72f6
       defaultHandler(handlerArg, dataBuf, (int)(dataPtr - (ICHAR *)dataBuf));
ed72f6
       *eventPP = s;
ed72f6
-    } while (s != end);
ed72f6
+    } while ((convert_res != XML_CONVERT_COMPLETED) && (convert_res != XML_CONVERT_INPUT_INCOMPLETE));
ed72f6
   }
ed72f6
   else
ed72f6
     defaultHandler(handlerArg, (XML_Char *)s, (int)((XML_Char *)end - (XML_Char *)s));
ed72f6
@@ -6148,8 +6149,8 @@
ed72f6
   if (!pool->ptr && !poolGrow(pool))
ed72f6
     return NULL;
ed72f6
   for (;;) {
ed72f6
-    XmlConvert(enc, &ptr, end, (ICHAR **)&(pool->ptr), (ICHAR *)pool->end);
ed72f6
-    if (ptr == end)
ed72f6
+    const enum XML_Convert_Result convert_res = XmlConvert(enc, &ptr, end, (ICHAR **)&(pool->ptr), (ICHAR *)pool->end);
ed72f6
+    if ((convert_res == XML_CONVERT_COMPLETED) || (convert_res == XML_CONVERT_INPUT_INCOMPLETE))
ed72f6
       break;
ed72f6
     if (!poolGrow(pool))
ed72f6
       return NULL;
ed72f6
@@ -6233,8 +6234,13 @@
ed72f6
     }
ed72f6
   }
ed72f6
   if (pool->blocks && pool->start == pool->blocks->s) {
ed72f6
-    int blockSize = (int)(pool->end - pool->start)*2;
ed72f6
-    BLOCK *temp = (BLOCK *)
ed72f6
+    BLOCK *temp;
ed72f6
+    int blockSize = (int)((unsigned)(pool->end - pool->start)*2U);
ed72f6
+
ed72f6
+    if (blockSize < 0)
ed72f6
+      return XML_FALSE;
ed72f6
+
ed72f6
+    temp = (BLOCK *)
ed72f6
       pool->mem->realloc_fcn(pool->blocks,
ed72f6
                              (offsetof(BLOCK, s)
ed72f6
                               + blockSize * sizeof(XML_Char)));
ed72f6
@@ -6249,6 +6255,10 @@
ed72f6
   else {
ed72f6
     BLOCK *tem;
ed72f6
     int blockSize = (int)(pool->end - pool->start);
ed72f6
+
ed72f6
+    if (blockSize < 0)
ed72f6
+      return XML_FALSE;
ed72f6
+
ed72f6
     if (blockSize < INIT_BLOCK_SIZE)
ed72f6
       blockSize = INIT_BLOCK_SIZE;
ed72f6
     else
ed72f6
diff -uap expat-2.1.0/lib/xmltok.c.cve0718 expat-2.1.0/lib/xmltok.c
ed72f6
--- expat-2.1.0/lib/xmltok.c.cve0718
ed72f6
+++ expat-2.1.0/lib/xmltok.c
ed72f6
@@ -318,39 +318,55 @@
ed72f6
   UTF8_cval4 = 0xf0
ed72f6
 };
ed72f6
 
ed72f6
-static void PTRCALL
ed72f6
+static enum XML_Convert_Result PTRCALL
ed72f6
 utf8_toUtf8(const ENCODING *enc,
ed72f6
             const char **fromP, const char *fromLim,
ed72f6
             char **toP, const char *toLim)
ed72f6
 {
ed72f6
+  enum XML_Convert_Result res = XML_CONVERT_COMPLETED;
ed72f6
   char *to;
ed72f6
   const char *from;
ed72f6
   if (fromLim - *fromP > toLim - *toP) {
ed72f6
     /* Avoid copying partial characters. */
ed72f6
+    res = XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
     for (fromLim = *fromP + (toLim - *toP); fromLim > *fromP; fromLim--)
ed72f6
       if (((unsigned char)fromLim[-1] & 0xc0) != 0x80)
ed72f6
         break;
ed72f6
   }
ed72f6
-  for (to = *toP, from = *fromP; from != fromLim; from++, to++)
ed72f6
+  for (to = *toP, from = *fromP; (from < fromLim) && (to < toLim); from++, to++)
ed72f6
     *to = *from;
ed72f6
   *fromP = from;
ed72f6
   *toP = to;
ed72f6
+
ed72f6
+  if ((to == toLim) && (from < fromLim))
ed72f6
+    return XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
+  else
ed72f6
+    return res;
ed72f6
 }
ed72f6
 
ed72f6
-static void PTRCALL
ed72f6
+static enum XML_Convert_Result PTRCALL
ed72f6
 utf8_toUtf16(const ENCODING *enc,
ed72f6
              const char **fromP, const char *fromLim,
ed72f6
              unsigned short **toP, const unsigned short *toLim)
ed72f6
 {
ed72f6
+  enum XML_Convert_Result res = XML_CONVERT_COMPLETED;
ed72f6
   unsigned short *to = *toP;
ed72f6
   const char *from = *fromP;
ed72f6
-  while (from != fromLim && to != toLim) {
ed72f6
+  while (from < fromLim && to < toLim) {
ed72f6
     switch (((struct normal_encoding *)enc)->type[(unsigned char)*from]) {
ed72f6
     case BT_LEAD2:
ed72f6
+      if (fromLim - from < 2) {
ed72f6
+        res = XML_CONVERT_INPUT_INCOMPLETE;
ed72f6
+        break;
ed72f6
+      }
ed72f6
       *to++ = (unsigned short)(((from[0] & 0x1f) << 6) | (from[1] & 0x3f));
ed72f6
       from += 2;
ed72f6
       break;
ed72f6
     case BT_LEAD3:
ed72f6
+      if (fromLim - from < 3) {
ed72f6
+        res = XML_CONVERT_INPUT_INCOMPLETE;
ed72f6
+        break;
ed72f6
+      }
ed72f6
       *to++ = (unsigned short)(((from[0] & 0xf) << 12)
ed72f6
                                | ((from[1] & 0x3f) << 6) | (from[2] & 0x3f));
ed72f6
       from += 3;
ed72f6
@@ -358,8 +374,14 @@
ed72f6
     case BT_LEAD4:
ed72f6
       {
ed72f6
         unsigned long n;
ed72f6
-        if (to + 1 == toLim)
ed72f6
+        if (toLim - to < 2) {
ed72f6
+          res = XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
           goto after;
ed72f6
+        }
ed72f6
+        if (fromLim - from < 4) {
ed72f6
+          res = XML_CONVERT_INPUT_INCOMPLETE;
ed72f6
+          goto after;
ed72f6
+        }
ed72f6
         n = ((from[0] & 0x7) << 18) | ((from[1] & 0x3f) << 12)
ed72f6
             | ((from[2] & 0x3f) << 6) | (from[3] & 0x3f);
ed72f6
         n -= 0x10000;
ed72f6
@@ -377,6 +399,7 @@
ed72f6
 after:
ed72f6
   *fromP = from;
ed72f6
   *toP = to;
ed72f6
+  return res;
ed72f6
 }
ed72f6
 
ed72f6
 #ifdef XML_NS
ed72f6
@@ -425,7 +448,7 @@
ed72f6
   STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)
ed72f6
 };
ed72f6
 
ed72f6
-static void PTRCALL
ed72f6
+static enum XML_Convert_Result PTRCALL
ed72f6
 latin1_toUtf8(const ENCODING *enc,
ed72f6
               const char **fromP, const char *fromLim,
ed72f6
               char **toP, const char *toLim)
ed72f6
@@ -433,30 +456,35 @@
ed72f6
   for (;;) {
ed72f6
     unsigned char c;
ed72f6
     if (*fromP == fromLim)
ed72f6
-      break;
ed72f6
+      return XML_CONVERT_COMPLETED;
ed72f6
     c = (unsigned char)**fromP;
ed72f6
     if (c & 0x80) {
ed72f6
       if (toLim - *toP < 2)
ed72f6
-        break;
ed72f6
+        return XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
       *(*toP)++ = (char)((c >> 6) | UTF8_cval2);
ed72f6
       *(*toP)++ = (char)((c & 0x3f) | 0x80);
ed72f6
       (*fromP)++;
ed72f6
     }
ed72f6
     else {
ed72f6
       if (*toP == toLim)
ed72f6
-        break;
ed72f6
+        return XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
       *(*toP)++ = *(*fromP)++;
ed72f6
     }
ed72f6
   }
ed72f6
 }
ed72f6
 
ed72f6
-static void PTRCALL
ed72f6
+static enum XML_Convert_Result PTRCALL
ed72f6
 latin1_toUtf16(const ENCODING *enc,
ed72f6
                const char **fromP, const char *fromLim,
ed72f6
                unsigned short **toP, const unsigned short *toLim)
ed72f6
 {
ed72f6
-  while (*fromP != fromLim && *toP != toLim)
ed72f6
+  while (*fromP < fromLim && *toP < toLim)
ed72f6
     *(*toP)++ = (unsigned char)*(*fromP)++;
ed72f6
+
ed72f6
+  if ((*toP == toLim) && (*fromP < fromLim))
ed72f6
+    return XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
+  else
ed72f6
+    return XML_CONVERT_COMPLETED;
ed72f6
 }
ed72f6
 
ed72f6
 #ifdef XML_NS
ed72f6
@@ -483,13 +511,18 @@
ed72f6
   STANDARD_VTABLE(sb_)
ed72f6
 };
ed72f6
 
ed72f6
-static void PTRCALL
ed72f6
+static enum XML_Convert_Result PTRCALL
ed72f6
 ascii_toUtf8(const ENCODING *enc,
ed72f6
              const char **fromP, const char *fromLim,
ed72f6
              char **toP, const char *toLim)
ed72f6
 {
ed72f6
-  while (*fromP != fromLim && *toP != toLim)
ed72f6
+  while (*fromP < fromLim && *toP < toLim)
ed72f6
     *(*toP)++ = *(*fromP)++;
ed72f6
+
ed72f6
+  if ((*toP == toLim) && (*fromP < fromLim))
ed72f6
+    return XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
+  else
ed72f6
+    return XML_CONVERT_COMPLETED;
ed72f6
 }
ed72f6
 
ed72f6
 #ifdef XML_NS
ed72f6
@@ -536,13 +569,14 @@
ed72f6
 }
ed72f6
 
ed72f6
 #define DEFINE_UTF16_TO_UTF8(E) \
ed72f6
-static void  PTRCALL \
ed72f6
+static enum XML_Convert_Result  PTRCALL \
ed72f6
 E ## toUtf8(const ENCODING *enc, \
ed72f6
             const char **fromP, const char *fromLim, \
ed72f6
             char **toP, const char *toLim) \
ed72f6
 { \
ed72f6
-  const char *from; \
ed72f6
-  for (from = *fromP; from != fromLim; from += 2) { \
ed72f6
+  const char *from = *fromP; \
ed72f6
+  fromLim = from + (((fromLim - from) >> 1) << 1);  /* shrink to even */ \
ed72f6
+  for (; from < fromLim; from += 2) { \
ed72f6
     int plane; \
ed72f6
     unsigned char lo2; \
ed72f6
     unsigned char lo = GET_LO(from); \
ed72f6
@@ -552,7 +586,7 @@
ed72f6
       if (lo < 0x80) { \
ed72f6
         if (*toP == toLim) { \
ed72f6
           *fromP = from; \
ed72f6
-          return; \
ed72f6
+          return XML_CONVERT_OUTPUT_EXHAUSTED; \
ed72f6
         } \
ed72f6
         *(*toP)++ = lo; \
ed72f6
         break; \
ed72f6
@@ -562,7 +596,7 @@
ed72f6
     case 0x4: case 0x5: case 0x6: case 0x7: \
ed72f6
       if (toLim -  *toP < 2) { \
ed72f6
         *fromP = from; \
ed72f6
-        return; \
ed72f6
+        return XML_CONVERT_OUTPUT_EXHAUSTED; \
ed72f6
       } \
ed72f6
       *(*toP)++ = ((lo >> 6) | (hi << 2) |  UTF8_cval2); \
ed72f6
       *(*toP)++ = ((lo & 0x3f) | 0x80); \
ed72f6
@@ -570,7 +604,7 @@
ed72f6
     default: \
ed72f6
       if (toLim -  *toP < 3)  { \
ed72f6
         *fromP = from; \
ed72f6
-        return; \
ed72f6
+        return XML_CONVERT_OUTPUT_EXHAUSTED; \
ed72f6
       } \
ed72f6
       /* 16 bits divided 4, 6, 6 amongst 3 bytes */ \
ed72f6
       *(*toP)++ = ((hi >> 4) | UTF8_cval3); \
ed72f6
@@ -580,7 +614,11 @@
ed72f6
     case 0xD8: case 0xD9: case 0xDA: case 0xDB: \
ed72f6
       if (toLim -  *toP < 4) { \
ed72f6
         *fromP = from; \
ed72f6
-        return; \
ed72f6
+        return XML_CONVERT_OUTPUT_EXHAUSTED; \
ed72f6
+      } \
ed72f6
+      if (fromLim - from < 4) { \
ed72f6
+        *fromP = from; \
ed72f6
+        return XML_CONVERT_INPUT_INCOMPLETE; \
ed72f6
       } \
ed72f6
       plane = (((hi & 0x3) << 2) | ((lo >> 6) & 0x3)) + 1; \
ed72f6
       *(*toP)++ = ((plane >> 2) | UTF8_cval4); \
ed72f6
@@ -596,20 +634,32 @@
ed72f6
     } \
ed72f6
   } \
ed72f6
   *fromP = from; \
ed72f6
+  if (from < fromLim) \
ed72f6
+    return XML_CONVERT_INPUT_INCOMPLETE; \
ed72f6
+  else \
ed72f6
+    return XML_CONVERT_COMPLETED; \
ed72f6
 }
ed72f6
 
ed72f6
 #define DEFINE_UTF16_TO_UTF16(E) \
ed72f6
-static void  PTRCALL \
ed72f6
+static enum XML_Convert_Result  PTRCALL \
ed72f6
 E ## toUtf16(const ENCODING *enc, \
ed72f6
              const char **fromP, const char *fromLim, \
ed72f6
              unsigned short **toP, const unsigned short *toLim) \
ed72f6
 { \
ed72f6
+  enum XML_Convert_Result res = XML_CONVERT_COMPLETED; \
ed72f6
+  fromLim = *fromP + (((fromLim - *fromP) >> 1) << 1);  /* shrink to even */ \
ed72f6
   /* Avoid copying first half only of surrogate */ \
ed72f6
   if (fromLim - *fromP > ((toLim - *toP) << 1) \
ed72f6
-      && (GET_HI(fromLim - 2) & 0xF8) == 0xD8) \
ed72f6
+      && (GET_HI(fromLim - 2) & 0xF8) == 0xD8) { \
ed72f6
     fromLim -= 2; \
ed72f6
-  for (; *fromP != fromLim && *toP != toLim; *fromP += 2) \
ed72f6
+    res = XML_CONVERT_INPUT_INCOMPLETE; \
ed72f6
+  } \
ed72f6
+  for (; *fromP < fromLim && *toP < toLim; *fromP += 2) \
ed72f6
     *(*toP)++ = (GET_HI(*fromP) << 8) | GET_LO(*fromP); \
ed72f6
+  if ((*toP == toLim) && (*fromP < fromLim)) \
ed72f6
+    return XML_CONVERT_OUTPUT_EXHAUSTED; \
ed72f6
+  else \
ed72f6
+    return res; \
ed72f6
 }
ed72f6
 
ed72f6
 #define SET2(ptr, ch) \
ed72f6
@@ -1288,7 +1338,7 @@
ed72f6
   return (c & ~0xFFFF) || checkCharRefNumber(c) < 0;
ed72f6
 }
ed72f6
 
ed72f6
-static void PTRCALL
ed72f6
+static enum XML_Convert_Result PTRCALL
ed72f6
 unknown_toUtf8(const ENCODING *enc,
ed72f6
                const char **fromP, const char *fromLim,
ed72f6
                char **toP, const char *toLim)
ed72f6
@@ -1299,21 +1349,21 @@
ed72f6
     const char *utf8;
ed72f6
     int n;
ed72f6
     if (*fromP == fromLim)
ed72f6
-      break;
ed72f6
+      return XML_CONVERT_COMPLETED;
ed72f6
     utf8 = uenc->utf8[(unsigned char)**fromP];
ed72f6
     n = *utf8++;
ed72f6
     if (n == 0) {
ed72f6
       int c = uenc->convert(uenc->userData, *fromP);
ed72f6
       n = XmlUtf8Encode(c, buf);
ed72f6
       if (n > toLim - *toP)
ed72f6
-        break;
ed72f6
+        return XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
       utf8 = buf;
ed72f6
       *fromP += (AS_NORMAL_ENCODING(enc)->type[(unsigned char)**fromP]
ed72f6
                  - (BT_LEAD2 - 2));
ed72f6
     }
ed72f6
     else {
ed72f6
       if (n > toLim - *toP)
ed72f6
-        break;
ed72f6
+        return XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
       (*fromP)++;
ed72f6
     }
ed72f6
     do {
ed72f6
@@ -1322,13 +1372,13 @@
ed72f6
   }
ed72f6
 }
ed72f6
 
ed72f6
-static void PTRCALL
ed72f6
+static enum XML_Convert_Result PTRCALL
ed72f6
 unknown_toUtf16(const ENCODING *enc,
ed72f6
                 const char **fromP, const char *fromLim,
ed72f6
                 unsigned short **toP, const unsigned short *toLim)
ed72f6
 {
ed72f6
   const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);
ed72f6
-  while (*fromP != fromLim && *toP != toLim) {
ed72f6
+  while (*fromP < fromLim && *toP < toLim) {
ed72f6
     unsigned short c = uenc->utf16[(unsigned char)**fromP];
ed72f6
     if (c == 0) {
ed72f6
       c = (unsigned short)
ed72f6
@@ -1340,6 +1390,11 @@
ed72f6
       (*fromP)++;
ed72f6
     *(*toP)++ = c;
ed72f6
   }
ed72f6
+
ed72f6
+  if ((*toP == toLim) && (*fromP < fromLim))
ed72f6
+    return XML_CONVERT_OUTPUT_EXHAUSTED;
ed72f6
+  else
ed72f6
+    return XML_CONVERT_COMPLETED;
ed72f6
 }
ed72f6
 
ed72f6
 ENCODING *
ed72f6
@@ -1503,7 +1558,7 @@
ed72f6
 {
ed72f6
   const ENCODING **encPtr;
ed72f6
 
ed72f6
-  if (ptr == end)
ed72f6
+  if (ptr >= end)
ed72f6
     return XML_TOK_NONE;
ed72f6
   encPtr = enc->encPtr;
ed72f6
   if (ptr + 1 == end) {
ed72f6
diff -uap expat-2.1.0/lib/xmltok.h.cve0718 expat-2.1.0/lib/xmltok.h
ed72f6
--- expat-2.1.0/lib/xmltok.h.cve0718
ed72f6
+++ expat-2.1.0/lib/xmltok.h
ed72f6
@@ -130,6 +130,12 @@
ed72f6
                                const char *,
ed72f6
                                const char **);
ed72f6
 
ed72f6
+enum XML_Convert_Result {
ed72f6
+  XML_CONVERT_COMPLETED = 0,
ed72f6
+  XML_CONVERT_INPUT_INCOMPLETE = 1,
ed72f6
+  XML_CONVERT_OUTPUT_EXHAUSTED = 2  /* and therefore potentially input remaining as well */
ed72f6
+};
ed72f6
+
ed72f6
 struct encoding {
ed72f6
   SCANNER scanners[XML_N_STATES];
ed72f6
   SCANNER literalScanners[XML_N_LITERAL_TYPES];
ed72f6
@@ -158,12 +164,12 @@
ed72f6
                             const char *ptr,
ed72f6
                             const char *end,
ed72f6
                             const char **badPtr);
ed72f6
-  void (PTRCALL *utf8Convert)(const ENCODING *enc,
ed72f6
+  enum XML_Convert_Result (PTRCALL *utf8Convert)(const ENCODING *enc,
ed72f6
                               const char **fromP,
ed72f6
                               const char *fromLim,
ed72f6
                               char **toP,
ed72f6
                               const char *toLim);
ed72f6
-  void (PTRCALL *utf16Convert)(const ENCODING *enc,
ed72f6
+  enum XML_Convert_Result (PTRCALL *utf16Convert)(const ENCODING *enc,
ed72f6
                                const char **fromP,
ed72f6
                                const char *fromLim,
ed72f6
                                unsigned short **toP,
ed72f6
diff -uap expat-2.1.0/lib/xmltok_impl.c.cve0718 expat-2.1.0/lib/xmltok_impl.c
ed72f6
--- expat-2.1.0/lib/xmltok_impl.c.cve0718
ed72f6
+++ expat-2.1.0/lib/xmltok_impl.c
ed72f6
@@ -93,13 +93,13 @@
ed72f6
 PREFIX(scanComment)(const ENCODING *enc, const char *ptr,
ed72f6
                     const char *end, const char **nextTokPtr)
ed72f6
 {
ed72f6
-  if (ptr != end) {
ed72f6
+  if (ptr < end) {
ed72f6
     if (!CHAR_MATCHES(enc, ptr, ASCII_MINUS)) {
ed72f6
       *nextTokPtr = ptr;
ed72f6
       return XML_TOK_INVALID;
ed72f6
     }
ed72f6
     ptr += MINBPC(enc);
ed72f6
-    while (ptr != end) {
ed72f6
+    while (ptr < end) {
ed72f6
       switch (BYTE_TYPE(enc, ptr)) {
ed72f6
       INVALID_CASES(ptr, nextTokPtr)
ed72f6
       case BT_MINUS:
ed72f6
@@ -147,7 +147,7 @@
ed72f6
     *nextTokPtr = ptr;
ed72f6
     return XML_TOK_INVALID;
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     case BT_PERCNT:
ed72f6
       if (ptr + MINBPC(enc) == end)
ed72f6
@@ -233,7 +233,7 @@
ed72f6
     *nextTokPtr = ptr;
ed72f6
     return XML_TOK_INVALID;
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     CHECK_NAME_CASES(enc, ptr, end, nextTokPtr)
ed72f6
     case BT_S: case BT_CR: case BT_LF:
ed72f6
@@ -242,7 +242,7 @@
ed72f6
         return XML_TOK_INVALID;
ed72f6
       }
ed72f6
       ptr += MINBPC(enc);
ed72f6
-      while (ptr != end) {
ed72f6
+      while (ptr < end) {
ed72f6
         switch (BYTE_TYPE(enc, ptr)) {
ed72f6
         INVALID_CASES(ptr, nextTokPtr)
ed72f6
         case BT_QUEST:
ed72f6
@@ -305,7 +305,7 @@
ed72f6
 PREFIX(cdataSectionTok)(const ENCODING *enc, const char *ptr,
ed72f6
                         const char *end, const char **nextTokPtr)
ed72f6
 {
ed72f6
-  if (ptr == end)
ed72f6
+  if (ptr >= end)
ed72f6
     return XML_TOK_NONE;
ed72f6
   if (MINBPC(enc) > 1) {
ed72f6
     size_t n = end - ptr;
ed72f6
@@ -348,7 +348,7 @@
ed72f6
     ptr += MINBPC(enc);
ed72f6
     break;
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
 #define LEAD_CASE(n) \
ed72f6
     case BT_LEAD ## n: \
ed72f6
@@ -391,11 +391,11 @@
ed72f6
     *nextTokPtr = ptr;
ed72f6
     return XML_TOK_INVALID;
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     CHECK_NAME_CASES(enc, ptr, end, nextTokPtr)
ed72f6
     case BT_S: case BT_CR: case BT_LF:
ed72f6
-      for (ptr += MINBPC(enc); ptr != end; ptr += MINBPC(enc)) {
ed72f6
+      for (ptr += MINBPC(enc); ptr < end; ptr += MINBPC(enc)) {
ed72f6
         switch (BYTE_TYPE(enc, ptr)) {
ed72f6
         case BT_S: case BT_CR: case BT_LF:
ed72f6
           break;
ed72f6
@@ -432,7 +432,7 @@
ed72f6
 PREFIX(scanHexCharRef)(const ENCODING *enc, const char *ptr,
ed72f6
                        const char *end, const char **nextTokPtr)
ed72f6
 {
ed72f6
-  if (ptr != end) {
ed72f6
+  if (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     case BT_DIGIT:
ed72f6
     case BT_HEX:
ed72f6
@@ -441,7 +441,7 @@
ed72f6
       *nextTokPtr = ptr;
ed72f6
       return XML_TOK_INVALID;
ed72f6
     }
ed72f6
-    for (ptr += MINBPC(enc); ptr != end; ptr += MINBPC(enc)) {
ed72f6
+    for (ptr += MINBPC(enc); ptr < end; ptr += MINBPC(enc)) {
ed72f6
       switch (BYTE_TYPE(enc, ptr)) {
ed72f6
       case BT_DIGIT:
ed72f6
       case BT_HEX:
ed72f6
@@ -464,7 +464,7 @@
ed72f6
 PREFIX(scanCharRef)(const ENCODING *enc, const char *ptr,
ed72f6
                     const char *end, const char **nextTokPtr)
ed72f6
 {
ed72f6
-  if (ptr != end) {
ed72f6
+  if (ptr < end) {
ed72f6
     if (CHAR_MATCHES(enc, ptr, ASCII_x))
ed72f6
       return PREFIX(scanHexCharRef)(enc, ptr + MINBPC(enc), end, nextTokPtr);
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
@@ -474,7 +474,7 @@
ed72f6
       *nextTokPtr = ptr;
ed72f6
       return XML_TOK_INVALID;
ed72f6
     }
ed72f6
-    for (ptr += MINBPC(enc); ptr != end; ptr += MINBPC(enc)) {
ed72f6
+    for (ptr += MINBPC(enc); ptr < end; ptr += MINBPC(enc)) {
ed72f6
       switch (BYTE_TYPE(enc, ptr)) {
ed72f6
       case BT_DIGIT:
ed72f6
         break;
ed72f6
@@ -506,7 +506,7 @@
ed72f6
     *nextTokPtr = ptr;
ed72f6
     return XML_TOK_INVALID;
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     CHECK_NAME_CASES(enc, ptr, end, nextTokPtr)
ed72f6
     case BT_SEMI:
ed72f6
@@ -529,7 +529,7 @@
ed72f6
 #ifdef XML_NS
ed72f6
   int hadColon = 0;
ed72f6
 #endif
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     CHECK_NAME_CASES(enc, ptr, end, nextTokPtr)
ed72f6
 #ifdef XML_NS
ed72f6
@@ -716,7 +716,7 @@
ed72f6
   hadColon = 0;
ed72f6
 #endif
ed72f6
   /* we have a start-tag */
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     CHECK_NAME_CASES(enc, ptr, end, nextTokPtr)
ed72f6
 #ifdef XML_NS
ed72f6
@@ -740,7 +740,7 @@
ed72f6
     case BT_S: case BT_CR: case BT_LF:
ed72f6
       {
ed72f6
         ptr += MINBPC(enc);
ed72f6
-        while (ptr != end) {
ed72f6
+        while (ptr < end) {
ed72f6
           switch (BYTE_TYPE(enc, ptr)) {
ed72f6
           CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr)
ed72f6
           case BT_GT:
ed72f6
@@ -785,7 +785,7 @@
ed72f6
 PREFIX(contentTok)(const ENCODING *enc, const char *ptr, const char *end,
ed72f6
                    const char **nextTokPtr)
ed72f6
 {
ed72f6
-  if (ptr == end)
ed72f6
+  if (ptr >= end)
ed72f6
     return XML_TOK_NONE;
ed72f6
   if (MINBPC(enc) > 1) {
ed72f6
     size_t n = end - ptr;
ed72f6
@@ -832,7 +832,7 @@
ed72f6
     ptr += MINBPC(enc);
ed72f6
     break;
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
 #define LEAD_CASE(n) \
ed72f6
     case BT_LEAD ## n: \
ed72f6
@@ -895,7 +895,7 @@
ed72f6
     *nextTokPtr = ptr;
ed72f6
     return XML_TOK_INVALID;
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     CHECK_NAME_CASES(enc, ptr, end, nextTokPtr)
ed72f6
     case BT_SEMI:
ed72f6
@@ -921,7 +921,7 @@
ed72f6
     *nextTokPtr = ptr;
ed72f6
     return XML_TOK_INVALID;
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     CHECK_NAME_CASES(enc, ptr, end, nextTokPtr)
ed72f6
     case BT_CR: case BT_LF: case BT_S:
ed72f6
@@ -941,7 +941,7 @@
ed72f6
                 const char *ptr, const char *end,
ed72f6
                 const char **nextTokPtr)
ed72f6
 {
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     int t = BYTE_TYPE(enc, ptr);
ed72f6
     switch (t) {
ed72f6
     INVALID_CASES(ptr, nextTokPtr)
ed72f6
@@ -973,7 +973,7 @@
ed72f6
                   const char **nextTokPtr)
ed72f6
 {
ed72f6
   int tok;
ed72f6
-  if (ptr == end)
ed72f6
+  if (ptr >= end)
ed72f6
     return XML_TOK_NONE;
ed72f6
   if (MINBPC(enc) > 1) {
ed72f6
     size_t n = end - ptr;
ed72f6
@@ -1141,7 +1141,7 @@
ed72f6
     *nextTokPtr = ptr;
ed72f6
     return XML_TOK_INVALID;
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     CHECK_NAME_CASES(enc, ptr, end, nextTokPtr)
ed72f6
     case BT_GT: case BT_RPAR: case BT_COMMA:
ed72f6
@@ -1204,10 +1204,10 @@
ed72f6
                           const char *end, const char **nextTokPtr)
ed72f6
 {
ed72f6
   const char *start;
ed72f6
-  if (ptr == end)
ed72f6
+  if (ptr >= end)
ed72f6
     return XML_TOK_NONE;
ed72f6
   start = ptr;
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
 #define LEAD_CASE(n) \
ed72f6
     case BT_LEAD ## n: ptr += n; break;
ed72f6
@@ -1262,10 +1262,10 @@
ed72f6
                        const char *end, const char **nextTokPtr)
ed72f6
 {
ed72f6
   const char *start;
ed72f6
-  if (ptr == end)
ed72f6
+  if (ptr >= end)
ed72f6
     return XML_TOK_NONE;
ed72f6
   start = ptr;
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
 #define LEAD_CASE(n) \
ed72f6
     case BT_LEAD ## n: ptr += n; break;
ed72f6
@@ -1326,7 +1326,7 @@
ed72f6
       end = ptr + n;
ed72f6
     }
ed72f6
   }
ed72f6
-  while (ptr != end) {
ed72f6
+  while (ptr < end) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     INVALID_CASES(ptr, nextTokPtr)
ed72f6
     case BT_LT:
ed72f6
@@ -1373,7 +1373,7 @@
ed72f6
 {
ed72f6
   ptr += MINBPC(enc);
ed72f6
   end -= MINBPC(enc);
ed72f6
-  for (; ptr != end; ptr += MINBPC(enc)) {
ed72f6
+  for (; ptr < end; ptr += MINBPC(enc)) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
     case BT_DIGIT:
ed72f6
     case BT_HEX:
ed72f6
@@ -1744,7 +1744,7 @@
ed72f6
                        const char *end,
ed72f6
                        POSITION *pos)
ed72f6
 {
ed72f6
-  while (ptr < end) {
ed72f6
+  while (end - ptr >= MINBPC(enc)) {
ed72f6
     switch (BYTE_TYPE(enc, ptr)) {
ed72f6
 #define LEAD_CASE(n) \
ed72f6
     case BT_LEAD ## n: \
ed72f6
@@ -1760,7 +1760,7 @@
ed72f6
     case BT_CR:
ed72f6
       pos->lineNumber++;
ed72f6
       ptr += MINBPC(enc);
ed72f6
-      if (ptr != end && BYTE_TYPE(enc, ptr) == BT_LF)
ed72f6
+      if (ptr < end && BYTE_TYPE(enc, ptr) == BT_LF)
ed72f6
         ptr += MINBPC(enc);
ed72f6
       pos->columnNumber = (XML_Size)-1;
ed72f6
       break;