Hello! On Sat, Nov 17, 2007 at 03:31:04PM +0100, I wrote: > On Tue, Nov 13, 2007 at 05:33:31PM -0800, Roland McGrath wrote: > > > I.e., the usec / nsec value of variable `st' (``struct stat'') isn't set > > > at all. In my reading of the used ``struct ext2_inode'' (variable `di') > > > there indeed is nothing better than the sec value being provided. Should > > > we explicitly set the usec / nsec value of `st' to zero? > > > > Yes. diskfs_make_node does not zero everything first, so read_node should > > set every field. > > Like this?
As per suggestion of Samuel I installed a slightly enhanced patch that
better points out what to do in the future.
#v+
2007-12-11 Thomas Schwinge <[EMAIL PROTECTED]>
* inode.c (read_node, write_node): Use `st->st_atim', `st->st_mtim',
`st->st_ctim'. Suggest what needs to be done in the future.
(read_node): Correctly initialize those.
#v-
Regards,
Thomas
signature.asc
Description: Digital signature
_______________________________________________ Bug-hurd mailing list [email protected] http://lists.gnu.org/mailman/listinfo/bug-hurd
