Bad file access on the rise

Miloslav Trma─Ź mitr at
Fri Jun 7 20:14:36 UTC 2013

On Fri, Jun 7, 2013 at 10:05 PM, Colin Walters <walters at> wrote:
> On Fri, 2013-06-07 at 20:42 +0100, Matthew Garrett wrote:
>> Without further analysis, it doesn't tell us much. Does the code attempt
>> to open a file O_NOATIME and then fall back to trying it without?
> It's likely:

Is there any more rationale for the change available?

(IMHO only very special applications should use O_NOATIME; if it is
not predictable which accesses do/don't update atime, the field
completely loses its value.  The definition we have is "time of last
access" (see (man stat.h) for the POSIX wording), not "time of last
user-initiated access" or "time used for $specific_purpose"; given the
lack of $specific_purpose it's difficult to say what accesses
should/should not be excluded but that's, I think, even more of a
reason to stick to the formal definition.  OTOH mlocate does use
O_NOATIME so that atime isn't updated for every modified directory,
and the line in the sand between updatedb and thumbnail generation
admittedly isn't all that clear.)

More information about the devel mailing list