mount doesn't complain about missing "mount directory" any more?

Kazuhiro Fujieda fujieda@jaist.ac.jp
Tue Jun 13 21:07:00 GMT 2000


>>> On Tue, 13 Jun 2000 22:58:20 -0400
>>> Chris Faylor <cgf@cygnus.com> said:

> Ok.  I'm sorry.  I didn't remember that -f avoided the error.  Is this
> new or has it always been like that?

The latter. The `mount' with -f had avoided the error before my
patch was applied.

> Btw, it sure would be nice if you could do this:
> 
> mount c:/foo /foo
> umount c:/foo	#doesn't work (but would be nice)
> umount /foo	#work
> 
> If you have an idle moment it would be nice to add this functionality to
> mount to make it behave more like the UNIX version.

`umount' has already reject win32 paths in del_reg_mount().
Anyway I removed the meaningless comparison in mount_info::del_item().

ChangeLog:
2000-06-14  Kazuhiro Fujieda <fujieda@jaist.ac.jp>
	* path.cc (mount_info::del_item): Eliminate the comparison between 
	the target mount point and native paths of mount entries.

Index: path.cc
===================================================================
RCS file: /cvs/src/src/winsup/cygwin/path.cc,v
retrieving revision 1.30
diff -u -p -r1.30 path.cc
--- path.cc	2000/06/13 16:48:37	1.30
+++ path.cc	2000/06/14 03:48:51
@@ -1696,9 +1696,7 @@ mount_info::del_item (const char *path, 
 
   for (int i = 0; i < nmounts; i++)
     {
-      /* Delete if paths and mount locations match. */
-      if ((strcasematch (mount[i].posix_path, pathtmp)
-	   || strcasematch (mount[i].native_path, pathtmp)) &&
+      if (strcasematch (mount[i].posix_path, pathtmp) &&
 	  (mount[i].flags & MOUNT_SYSTEM) == (flags & MOUNT_SYSTEM))
 	{
 	  nmounts--;		/* One less mount table entry */

____
  | AIST      Kazuhiro Fujieda <fujieda@jaist.ac.jp>
  | HOKURIKU  School of Information Science
o_/ 1990      Japan Advanced Institute of Science and Technology


More information about the Cygwin-developers mailing list