> Hence I suggest that you test the following change, although I don't > see why it would work. Why it would work or why it would not work? :-) The below patch works fine for dropping the -d time back to sensible levels, to 0.73 (u+s) or so (5.6.1 had 0.61 and 5.8.0 has 4.09). > - if (PL_op->op_private || SvIV(PL_DBsingle) || SvIV(PL_DBsignal) || SvIV(PL_DBtrace)) > + if (PL_op->op_private & HINT_BYTES || SvIV(PL_DBsingle) || SvIV(PL_DBsignal) || SvIV(PL_DBtrace)) -- Jarkko Hietaniemi <jhi@iki.fi> http://www.iki.fi/jhi/ "There is this special biologist word we use for 'stable'. It is 'dead'." -- Jack CohenThread Previous | Thread Next