[libxml2] Detect excessive entities expansion upon replacement

[Daniel Veillard <veillard src gnome org>]

commit 23f05e0c33987d6605387b300c4be5da2120a7ab
Author: Daniel Veillard <veillard redhat com>
Date:   Tue Feb 19 10:21:49 2013 +0800

    Detect excessive entities expansion upon replacement
    
    If entities expansion in the XML parser is asked for,
    it is possble to craft relatively small input document leading
    to excessive on-the-fly content generation.
    This patch accounts for those replacement and stop parsing
    after a given threshold. it can be bypassed as usual with the
    HUGE parser option.

 include/libxml/parser.h |    1 +
 parser.c                |   44 ++++++++++++++++++++++++++++++++++++++------
 parserInternals.c       |    2 ++
 3 files changed, 41 insertions(+), 6 deletions(-)
---
diff --git a/include/libxml/parser.h b/include/libxml/parser.h
index e1346e4..3f5730d 100644
--- a/include/libxml/parser.h
+++ b/include/libxml/parser.h
@@ -310,6 +310,7 @@ struct _xmlParserCtxt {
     xmlParserNodeInfo *nodeInfoTab;   /* array of nodeInfos */
 
     int                input_id;      /* we need to label inputs */
+    unsigned long      sizeentcopy;   /* volume of entity copy */
 };
 
 /**
diff --git a/parser.c b/parser.c
index 91f8c90..ddf3b5b 100644
--- a/parser.c
+++ b/parser.c
@@ -122,7 +122,7 @@ xmlCreateEntityParserCtxtInternal(const xmlChar *URL, const xmlChar *ID,
  */
 static int
 xmlParserEntityCheck(xmlParserCtxtPtr ctxt, size_t size,
-                     xmlEntityPtr ent)
+                     xmlEntityPtr ent, size_t replacement)
 {
     size_t consumed = 0;
 
@@ -130,7 +130,24 @@ xmlParserEntityCheck(xmlParserCtxtPtr ctxt, size_t size,
         return (0);
     if (ctxt->lastError.code == XML_ERR_ENTITY_LOOP)
         return (1);
-    if (size != 0) {
+    if (replacement != 0) {
+       if (replacement < XML_MAX_TEXT_LENGTH)
+           return(0);
+
+        /*
+        * If the volume of entity copy reaches 10 times the
+        * amount of parsed data and over the large text threshold
+        * then that's very likely to be an abuse.
+        */
+        if (ctxt->input != NULL) {
+           consumed = ctxt->input->consumed +
+                      (ctxt->input->cur - ctxt->input->base);
+       }
+        consumed += ctxt->sizeentities;
+
+        if (replacement < XML_PARSER_NON_LINEAR * consumed)
+           return(0);
+    } else if (size != 0) {
         /*
          * Do the check based on the replacement size of the entity
          */
@@ -176,7 +193,6 @@ xmlParserEntityCheck(xmlParserCtxtPtr ctxt, size_t size,
          */
         return (0);
     }
-
     xmlFatalErr(ctxt, XML_ERR_ENTITY_LOOP, NULL);
     return (1);
 }
@@ -2743,7 +2759,7 @@ xmlStringLenDecodeEntities(xmlParserCtxtPtr ctxt, const xmlChar *str, int len,
                    while (*current != 0) { /* non input consuming loop */
                        buffer[nbchars++] = *current++;
                        if (nbchars + XML_PARSER_BUFFER_SIZE > buffer_size) {
-                           if (xmlParserEntityCheck(ctxt, nbchars, ent))
+                           if (xmlParserEntityCheck(ctxt, nbchars, ent, 0))
                                goto int_error;
                            growBuffer(buffer, XML_PARSER_BUFFER_SIZE);
                        }
@@ -2785,7 +2801,7 @@ xmlStringLenDecodeEntities(xmlParserCtxtPtr ctxt, const xmlChar *str, int len,
                    while (*current != 0) { /* non input consuming loop */
                        buffer[nbchars++] = *current++;
                        if (nbchars + XML_PARSER_BUFFER_SIZE > buffer_size) {
-                           if (xmlParserEntityCheck(ctxt, nbchars, ent))
+                           if (xmlParserEntityCheck(ctxt, nbchars, ent, 0))
                                goto int_error;
                            growBuffer(buffer, XML_PARSER_BUFFER_SIZE);
                        }
@@ -7203,7 +7219,7 @@ xmlParseReference(xmlParserCtxtPtr ctxt) {
            xmlFreeNodeList(list);
            return;
        }
-       if (xmlParserEntityCheck(ctxt, 0, ent)) {
+       if (xmlParserEntityCheck(ctxt, 0, ent, 0)) {
            xmlFreeNodeList(list);
            return;
        }
@@ -7361,6 +7377,13 @@ xmlParseReference(xmlParserCtxtPtr ctxt) {
                xmlNodePtr nw = NULL, cur, firstChild = NULL;
 
                /*
+                * We are copying here, make sure there is no abuse
+                */
+               ctxt->sizeentcopy += ent->length;
+               if (xmlParserEntityCheck(ctxt, 0, ent, ctxt->sizeentcopy))
+                   return;
+
+               /*
                 * when operating on a reader, the entities definitions
                 * are always owning the entities subtree.
                if (ctxt->parseMode == XML_PARSE_READER)
@@ -7400,6 +7423,14 @@ xmlParseReference(xmlParserCtxtPtr ctxt) {
            } else if ((list == NULL) || (ctxt->inputNr > 0)) {
                xmlNodePtr nw = NULL, cur, next, last,
                           firstChild = NULL;
+
+               /*
+                * We are copying here, make sure there is no abuse
+                */
+               ctxt->sizeentcopy += ent->length;
+               if (xmlParserEntityCheck(ctxt, 0, ent, ctxt->sizeentcopy))
+                   return;
+
                /*
                 * Copy the entity child list and make it the new
                 * entity child list. The goal is to make sure any
@@ -14767,6 +14798,7 @@ xmlCtxtReset(xmlParserCtxtPtr ctxt)
     ctxt->catalogs = NULL;
     ctxt->nbentities = 0;
     ctxt->sizeentities = 0;
+    ctxt->sizeentcopy = 0;
     xmlInitNodeInfoSeq(&ctxt->node_seq);
 
     if (ctxt->attsDefault != NULL) {
diff --git a/parserInternals.c b/parserInternals.c
index 02032d5..f8a7041 100644
--- a/parserInternals.c
+++ b/parserInternals.c
@@ -1719,6 +1719,8 @@ xmlInitParserCtxt(xmlParserCtxtPtr ctxt)
     ctxt->charset = XML_CHAR_ENCODING_UTF8;
     ctxt->catalogs = NULL;
     ctxt->nbentities = 0;
+    ctxt->sizeentities = 0;
+    ctxt->sizeentcopy = 0;
     ctxt->input_id = 1;
     xmlInitNodeInfoSeq(&ctxt->node_seq);
     return(0);