On Tuesday 24 April 2007, Karel Zak wrote:
> On Fri, Apr 13, 2007 at 07:29:37AM -0400, Mike Frysinger wrote:
> > --- a/mount/umount.c
> > +++ b/mount/umount.c
> > @@ -349,7 +349,7 @@ umount_one (const char *spec, const char *node, const
> > char *type, remnt.mnt_type = remnt.mnt_fsname = NULL;
> > remnt.mnt_dir = xstrdup(node);
> > remnt.mnt_opts = xstrdup("ro");
> > - if (!nomtab)
> > + if (!nomtab && mtab_is_writable())
> > update_mtab(node, &remnt);
>
> See ftab.c:
>
> void
> update_mtab (const char *dir, struct my_mntent *instead) {
> mntFILE *mfp, *mftmp;
> const char *fnam = MOUNTED;
> struct mntentchn mtabhead; /* dummy */
> struct mntentchn *mc, *mc0, *absent = NULL;
>
> if (mtab_does_not_exist() || !mtab_is_writable())
> ^^^^^^^^^^^^^^^^^^^
> return;
>
>
> It means the patch duplicates the check. Unfortunately, the same
> duplication is already in mount.c (I will fix it).
>
> I think better will be keep the mtab_is_writable() in
> update_mtab() only.
the patch was to fix the mentioned use case ... if the use case is fixed, then
it doesnt matter to me how ... old version of util-linux lacked that check
you pointed out in the current git, so indeed my patch would double up the
writable check in current git
cheers
-mike
Attachment:
signature.asc
Description: This is a digitally signed message part.
