[Bug-gnulib] Re: Bug#202243: glibc utimes glitch with coreutils 'touch'

[GOTO Masanori]

At 18 Aug 2003 00:07:28 -0700,
Paul Eggert wrote:
> GOTO Masanori <address@hidden> writes:
> 
> > So I patched Jakub's modification for 2.3.2-3, ok?
> 
> If you installed the patch in
> <http://sources.redhat.com/ml/libc-alpha/2003-08/msg00062.html>
> then you should be OK, but if it's some other patch then I'd like to
> know what it was.

My previous mention is insufficient.  It's not your patch, Jakub's
2003-07-31 modification, rouding one (the current debian glibc 2.3.2
is based on Mid of July).  Because I found the below sentence in
SuSv3.  It says it may be rounding (or am I missing?):

48734 DESCRIPTION
48735              The utimes( ) function shall set the access and modification 
times of the file pointed to by the path
48736              argument to the value of the times argument. The utimes( ) 
function allows time specifications
48737              accurate to the microsecond.
48738              For utimes( ), the times argument is an array of timeval 
structures. The first array member
48739              represents the date and time of last access, and the second 
member represents the date and time
48740              of last modification. The times in the timeval structure are 
measured in seconds and
48741              microseconds since the Epoch, although rounding toward the 
nearest second may occur.

your reffered "Historically, utimes truncated when it couldn't store
the microseconds" is BSD's original implementation?  Yes, the latest
FreeBSD cvs says:

                tv[0].tv_sec = times->actime;
                tv[1].tv_sec = times->modtime;
                tv[0].tv_usec = tv[1].tv_usec = 0;

Regards,
-- gotom