Do not warn about Vary loops and mismatches. 

I have a feeling that a lot of Vary-handling code has too-high debugging
levels, but it is not clear to me whether those loops are dangerous 
enough to warrant level-1 debugging. This needs to be investigated 
before committing this change.

Index: src/client_side.cc
===================================================================
RCS file: /cvsroot/squid/squid3/src/client_side.cc,v
retrieving revision 1.779
diff -u -r1.779 client_side.cc
--- src/client_side.cc	26 Feb 2008 21:49:34 -0000	1.779
+++ src/client_side.cc	14 Mar 2008 21:11:52 -0000
@@ -3274,7 +3274,7 @@
             /* Oops.. we have already been here and still haven't
              * found the requested variant. Bail out
              */
-            debugs(33, 1, "varyEvaluateMatch: Oops. Not a Vary match on second attempt, '" <<
+            debugs(33, 2, "varyEvaluateMatch: Oops. Not a Vary match on second attempt, '" <<
                     entry->mem_obj->url << "' '" << vary << "'");
             return VARY_CANCEL;
         }
Index: src/client_side_reply.cc
===================================================================
RCS file: /cvsroot/squid/squid3/src/client_side_reply.cc,v
retrieving revision 1.154
diff -u -r1.154 client_side_reply.cc
--- src/client_side_reply.cc	16 Feb 2008 17:42:27 -0000	1.154
+++ src/client_side_reply.cc	14 Mar 2008 21:11:52 -0000
@@ -534,7 +534,7 @@
 
     case VARY_CANCEL:
         /* varyEvaluateMatch found a object loop. Process as miss */
-        debugs(88, 1, "clientProcessHit: Vary object loop!");
+        debugs(88, 2, "clientProcessHit: Vary object loop!");
         processMiss();
         return;
     }