Re: spy.test in 1.9.3 release under solaris

>>> "Andy" == Andy Shelley <andy@xxxxxxxxxxx> writes:

[...]
 Andy> doing a stat of "a" just before the last $sleep made for
 Andy> a successful test.  Only for "a", and only at that point
 Andy> in the test.
[...]
 Andy> None of this resolves any root cause for my environment, though it
 Andy> does seem to trip some trigger.  
[...]

Frankly, if

 stat a
 sleep 2
 whatever

does not behave the same as

 sleep 2
 stat a
 whatever

I'm tempted to blame the file system.  `stat a' must cause some
timestamp to be updated.  What filesystem is this?  Is it
mounted with some funky option?  (I see Solaris has a dfratime
option to defer atime updates on UFS; and your case let me
feel like mtime updates was differed until the next stat.)

-- 
Alexandre Duret-Lutz