Fix up logging for write-fail-5/collapsed-forwarding (#12050)

The logic here changed so that it no longer loops for all write retries before retrying
to the read. So the logging no longer worked correctly

(cherry picked from commit c82122e)

Author: ezelkow1

src/proxy/logging/LogAccess.cc

@@ -3028,8 +3028,8 @@ LogAccess::marshal_cache_collapsed_connection_success(char *buf)
     if (m_http_sm) {
       SquidLogCode code = m_http_sm->t_state.squid_codes.log_code;
 
-      // We increment open_write_tries beyond the max when we want to jump back to the read state for collapsing
-      if ((m_http_sm->get_cache_sm().get_open_write_tries() > (m_http_sm->t_state.txn_conf->max_cache_open_write_retries)) &&
+      // We attempted an open write, but ended up with some sort of HIT which means we must have gone back to the read state
+      if ((m_http_sm->get_cache_sm().get_open_write_tries() > (0)) &&
           ((code == SQUID_LOG_TCP_HIT) || (code == SQUID_LOG_TCP_MEM_HIT) || (code == SQUID_LOG_TCP_DISK_HIT) ||
            (code == SQUID_LOG_TCP_CF_HIT))) {
         // Attempted collapsed connection and got a hit, success