Mail Archives: cygwin/2007/03/26/10:29:09
On Mar 26 14:10, James Youngman wrote:
> On 3/26/07, Eric Blake <ebb9 AT byu DOT net> wrote:
> >Corinna only replied to the cygwin list. It looks like Windows will
> >populate st_birthtime with st_ctime when reading filesystems that don't
> >support birthtime.
Er, sorry, but I didn't say that. I said that all Windows functions
returning filetimes will return a creation time. We can only work with
the values given. I never found a situation in which the creation time
is not set. And if it's set, we must trust the value.
Native Windows file systems always maintain a real creation time and
Samba fakes it. Samba sets the creation time to the minimum value of
the stat members st_ctime, st_mtime and st_atime.
The only timestamp I ever found to be 0 is the ChangeTime. This is the
fourth timestamp maintained by the SMB protocol as well as by native NT
functions and it's meaning is equivalent to st_ctime. This timestamp is
not returned by Win32 funtions, but Cygwin uses the native NT functions
so we get a valid st_ctime as long as the file system supports it. NTFS
returns a valid ChangeTime. FAT returns 0, in which case Cygwin sets
st_ctime = st_mtime.
Unfortunately Samba sets the ChangeTime not to st_ctime, but to
st_mtime, even if the underlying filesystem supports st_ctime. However,
as I mentioned above, if the timestamp is valid, we have to trust it.
We have no other way to evaluate a value.
Corinna
--
Corinna Vinschen Please, send mails regarding Cygwin to
Cygwin Project Co-Leader cygwin AT cygwin DOT com
Red Hat
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
- Raw text -