From: Paul J. Reder <rederpj@apache.org>
Date: Thu, 4 Apr 2002 21:19:32 +0000 (+0000)
Subject: This patch fixes a core dump that occurs in mod_include during tag parsing
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=888e4f5fb8fd43ccbd2eed191716fc2df1f92542;p=apache

This patch fixes a core dump that occurs in mod_include during tag parsing
if the starting sequence (<!--#) finishes at the end of a bucket and the
directive starts at the beginning of the next bucket.

There is still a bug in the handling of conditional directives that span
buckets, but the core dump is now fixed and the conditional bug should
be even rarer than the core dump was. I'm fixing the conditional bug
seperately.


git-svn-id: https://svn.apache.org/repos/asf/httpd/httpd/trunk@94436 13f79535-47bb-0310-9956-ffa450edef68
---

diff --git a/modules/filters/mod_include.c b/modules/filters/mod_include.c
index 5004c8fec5..58054fe253 100644
--- a/modules/filters/mod_include.c
+++ b/modules/filters/mod_include.c
@@ -319,14 +319,20 @@ static apr_size_t bndm(const char *n, apr_size_t nl, const char *h,
 /* We've now found a start sequence tag... */
 static apr_bucket* found_start_sequence(apr_bucket *dptr,
                                         include_ctx_t *ctx, 
-                                        int tagStart)
+                                        int tagStart,
+                                        apr_size_t len)
 {
     /* We want to split the bucket at the '<'. */
     ctx->state = PARSE_DIRECTIVE;
     ctx->tag_length = 0;
     ctx->parse_pos = 0;
-    ctx->tag_start_bucket = dptr;
-    ctx->tag_start_index = tagStart;
+
+    /* Don't set tag_start_bucket if tagStart indexes the end of the bucket. */
+    if (tagStart < len) {
+        ctx->tag_start_bucket = dptr;
+        ctx->tag_start_index = tagStart;
+    }
+
     if (ctx->head_start_index > 0) {
         apr_bucket *tmp_bkt;
 
@@ -334,7 +340,7 @@ static apr_bucket* found_start_sequence(apr_bucket *dptr,
         apr_bucket_split(ctx->head_start_bucket, ctx->head_start_index);
         tmp_bkt = APR_BUCKET_NEXT(ctx->head_start_bucket);
         /* If it was a one bucket match */
-        if (dptr == ctx->head_start_bucket) {
+        if ((tagStart < len) && (dptr == ctx->head_start_bucket)) {
             ctx->tag_start_bucket = tmp_bkt;
             ctx->tag_start_index = tagStart - ctx->head_start_index;
         }
@@ -440,7 +446,7 @@ static apr_bucket *find_start_sequence(apr_bucket *dptr, include_ctx_t *ctx,
             if (str[ctx->parse_pos] == '\0')
             {
                 ctx->bytes_parsed += c - buf;
-                return found_start_sequence(dptr, ctx, c - buf);
+                return found_start_sequence(dptr, ctx, c - buf, len);
             }
             else if (c == buf + tmpLen) {
                 dptr = APR_BUCKET_NEXT(dptr);
@@ -451,7 +457,16 @@ static apr_bucket *find_start_sequence(apr_bucket *dptr, include_ctx_t *ctx,
              */
             APR_BRIGADE_PREPEND(bb, ctx->ssi_tag_brigade);
 
+            /* We know we are at the beginning of this bucket so
+             *   we can just prepend the saved bytes from the
+             *   ssi_tag_brigade (which empties the ssi_tag_brigade)
+             *   and continue processing.
+             * We do not need to set do_cleanup beacuse the
+             *   prepend takes care of that.
+             */
             ctx->state = PRE_HEAD;
+            ctx->head_start_bucket = NULL;
+            ctx->head_start_index = 0;
         }
 
         if (len)
@@ -461,8 +476,10 @@ static apr_bucket *find_start_sequence(apr_bucket *dptr, include_ctx_t *ctx,
             {
                 ctx->head_start_bucket = dptr;
                 ctx->head_start_index = pos;
-                ctx->bytes_parsed += pos + slen;
-                return found_start_sequence(dptr, ctx, pos + slen);
+                ctx->bytes_parsed += pos + slen;       /* pjr - isn't this incrementing by too much?
+                                                        *    What if only 1 byte in this bucket?
+                                                        */
+                return found_start_sequence(dptr, ctx, pos + slen, len);
             }
         }
         
@@ -485,29 +502,35 @@ static apr_bucket *find_start_sequence(apr_bucket *dptr, include_ctx_t *ctx,
                     ctx->head_start_index = c - buf;
                 }
                 ctx->parse_pos++;
+                c++;
+                ctx->bytes_parsed++;
             }
             else if (ctx->parse_pos != 0) 
             {
-                /* The reason for this, is that we need to make sure 
-                 * that we catch cases like <<!--#.  This makes the 
-                 * second check after the original check fails.
-                 * If parse_pos was already 0 then we already checked this.
+                /* DO NOT INCREMENT c IN THIS BLOCK!
+                 * Don't increment bytes_parsed either.
+                 * This block is just to reset the indexes and
+                 *   pointers related to parsing the tag start_sequence.
+                 * The value c needs to be checked again to handle
+                 *   the case where we find "<<!--#". We are now
+                 *   looking at the second "<" and need to restart
+                 *   the start_sequence checking from parse_pos = 0.
+                 * do_cleanup causes the stored bytes in ssi_tag_brigade
+                 *   to be forwarded on and cleaned up. We may not be
+                 *   able to just prepend the ssi_tag_brigade because
+                 *   we may have advanced too far before we noticed this
+                 *   case, so just flag it and clean it up later.
                  */
-                /* FIXME: Why? */
                 *do_cleanup = 1;
-                if (*c == str[0]) {
-                    ctx->parse_pos = 1;
-                    ctx->head_start_index = c - buf;
-                }
-                else {
-                    ctx->parse_pos = 0;
-                    ctx->state = PRE_HEAD;
-                    ctx->head_start_bucket = NULL;
-                    ctx->head_start_index = 0;
-                }
+                ctx->parse_pos = 0;
+                ctx->state = PRE_HEAD;
+                ctx->head_start_bucket = NULL;
+                ctx->head_start_index = 0;
+            }
+            else {
+               c++;
+               ctx->bytes_parsed++;
             }
-            c++;
-            ctx->bytes_parsed++;
         }
         dptr = APR_BUCKET_NEXT(dptr);
     } while (dptr != APR_BRIGADE_SENTINEL(bb));