haribabu.kommi, in WT-5264 you added two new fields to the session structure:

    /* Original transaction time pair to use for the history store inserts */
    uint64_t orig_txnid_to_las;
    wt_timestamp_t orig_timestamp_to_las;
 

And there's currently code in__wt_txn_modify() that uses those fields:

   /* Use the original transaction time pair for the history store inserts */
    if (WT_IS_HS(S2BT(session))) {
        upd->txnid = session->orig_txnid_to_las;
        upd->start_ts = session->orig_timestamp_to_las;
    } else {
        upd->txnid = txn->id;
        __wt_txn_op_set_timestamp(session, op);
    } 

sulabh.mahajan, in WT-5478 you removed the function that initialized those fields (a function that was at the time named __hs_store_time_pair()), but didn't remove the fields or the code that used them?

I suspect the fields can just go away, but I'm a little concerned that we're setting the update structure txnid/start_ts values to zero'd out memory, can that get us into trouble?