Re: noatime mounts inhibiting atime updates via utime()

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Tue, 4 Dec 2018 20:23:20 -0800
    From:        Jason Thorpe <thorpej%me.com@localhost>
    Message-ID:  <0DF612FF-9DD1-4B5E-9256-33C13730D2E1%me.com@localhost>

  | Honestly, I think atime is one of the dumbest thing ever.

That I don't agree with, I use it all the time.   The (similar) thing that is
a much dumber idea is birthtime - that's a waste of space and time
(it just has no I/O costs so mostly people just don't care.)

  | But if one must support it, then I think the following is a reasonable
  | compromise:

What you say is almost exactly what I meant, except perhaps ...

  | If one has a need to have super-strict atime updates, then I think *that*
  | should be the optional behavior.

Which is the default, and which requires config, is not something I
worry about very much - deciding upon the optimal default config
is a difficult problem, as we need to guess at what the majority of
users' needs are going to be.

But I would tend to err on the side of stability (if yesterday's config
did things one way, unless there is a very good reason, so should
tomorrow's).

As long as it can be configired is all that matters really.

kre