On Tue, Aug 15, 2017 at 02:03:43PM -0400, Jeremie Courreges-Anglas wrote: > On Tue, Aug 15 2017, Rob Pierce <r...@2keys.ca> wrote: > > [...] > > > I was able to reproduce the problem with a HISTSIZE of 100000 which at > > 125000 > > entries rendered my system unusable. With the patch I am running fine with a > > HISTSIZE of 120000 and have come back several times after hitting the 1.25x > > threshold. > > > > Regression tests pass. > > Thanks for your feedback. As I said privately, I did a bad job at > analyzing how to fix the slowness of afree(). With latest commit to > alloc.c, switching history lines allocation to strdup(3)/free(3) is not > needed any more; also it is probably better for us to use the same > allocation pattern everywhere.
Sounds good. Your latest commit passed my local tests with a large HISTSIZE. Regards, Rob > Thanks, > -- > jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
