Bug#276500: #276500: utimensat added as syscall

[Eric Blake <ebb9@byu.net>]

For what it's worth, coreutils 6.12 and newer now use utimensat if
available; and when coupled with a new enough kernel, no longer
demonstrate the original issue reported against touch(1).  However, in
testing the coreutils patch to use utimensat, it was discovered that at
least some versions of the kernel have a buggy utimensat syscall that
return a bogus positive number instead of -1 on failure:

http://git.savannah.gnu.org/gitweb/?p=gnulib.git;a=commitdiff;h=9840199

--
Don't work too hard, make some time for fun as well!

Eric Blake             ebb9@byu.net