Enforce token characters for field-name (#700)

RFC 7230 defines field-name as a token. Request splitting and cache
poisoning attacks have used non-token characters to fool broken HTTP
agents behind or in front of Squid for years. This change should
significantly reduce that abuse.

If we discover exceptional situations that need special treatment, the
relaxed parser can allow them on a case-by-case basis (while being extra
careful about framing-related header fields), just like we already
tolerate some header whitespace (e.g., between the response header
field-name and colon).

diff --git a/src/HttpHeader.cc b/src/HttpHeader.cc
index dc6e0ffd635284952dd46fa61c794666d708989b..9e5e47fb3424e718e5232fcf4a513ae981134695 100644 (file)
--- a/src/HttpHeader.cc
+++ b/src/HttpHeader.cc
@@ -443,18 +443,6 @@ HttpHeader::parse(const char *header_start, size_t hdrLen)
             return 0;
         }
 
-        if (e->id == Http::HdrType::OTHER && stringHasWhitespace(e->name.termedBuf())) {
-            debugs(55, warnOnError, "WARNING: found whitespace in HTTP header name {" <<
-                   getStringPrefix(field_start, field_end-field_start) << "}");
-
-            if (!Config.onoff.relaxed_header_parser) {
-                delete e;
-                PROF_stop(HttpHeaderParse);
-                clean();
-                return 0;
-            }
-        }
-
         addEntry(e);
     }
 
@@ -1437,6 +1425,20 @@ HttpHeaderEntry::parse(const char *field_start, const char *field_end, const htt
         }
     }
 
+    /* RFC 7230 section 3.2:
+     *
+     *  header-field   = field-name ":" OWS field-value OWS
+     *  field-name     = token
+     *  token          = 1*TCHAR
+     */
+    for (const char *pos = field_start; pos < (field_start+name_len); ++pos) {
+        if (!CharacterSet::TCHAR[*pos]) {
+            debugs(55, 2, "found header with invalid characters in " <<
+                   Raw("field-name", field_start, min(name_len,100)) << "...");
+            return nullptr;
+        }
+    }
+
     /* now we know we can parse it */
 
     debugs(55, 9, "parsing HttpHeaderEntry: near '" <<  getStringPrefix(field_start, field_end-field_start) << "'");