Re: [PATCH 14/21] cifs: we do not allow changing username/password/unc/... during remount

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Wed, Dec 9, 2020 at 7:08 AM Pavel Shilovsky <piastryyy@xxxxxxxxx> wrote:
>
> пн, 7 дек. 2020 г. в 21:07, Steve French <smfrench@xxxxxxxxx>:
> >
> > Minor nits pointed out by checkpatch:
> >
> > 0015-cifs-we-do-not-allow-changing-username-password-unc-.patch
> > ---------------------------------------------------------------
> > WARNING: Missing commit description - Add an appropriate one
> >
> > WARNING: kfree(NULL) is safe and this check is probably not required
> > #76: FILE: fs/cifs/fs_context.c:673:
> > + if (ctx->field) { \
> > + kfree(ctx->field);
> >
> > On Mon, Dec 7, 2020 at 5:37 PM Ronnie Sahlberg <lsahlber@xxxxxxxxxx> wrote:
> > >
> > > Signed-off-by: Ronnie Sahlberg <lsahlber@xxxxxxxxxx>
> > > ---
> > >  fs/cifs/cifsfs.c     |  2 +-
> > >  fs/cifs/fs_context.c | 55 +++++++++++++++++++++++++++++++++++++++++++++++++---
> > >  fs/cifs/fs_context.h |  2 +-
> > >  3 files changed, 54 insertions(+), 5 deletions(-)
> > >
> > > diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
> > > index 80117e9d35f9..13d7f4a3c836 100644
> > > --- a/fs/cifs/cifsfs.c
> > > +++ b/fs/cifs/cifsfs.c
> > > @@ -490,7 +490,7 @@ cifs_show_options(struct seq_file *s, struct dentry *root)
> > >
> > >         if (tcon->no_lease)
> > >                 seq_puts(s, ",nolease");
> > > -       if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER)
> > > +       if (cifs_sb->ctx->multiuser)
> > >                 seq_puts(s, ",multiuser");
> > >         else if (tcon->ses->user_name)
> > >                 seq_show_option(s, "username", tcon->ses->user_name);
> > > diff --git a/fs/cifs/fs_context.c b/fs/cifs/fs_context.c
> > > index edfdea129fcc..542fa75b74aa 100644
> > > --- a/fs/cifs/fs_context.c
> > > +++ b/fs/cifs/fs_context.c
> > > @@ -629,10 +629,53 @@ static int smb3_verify_reconfigure_ctx(struct smb3_fs_context *new_ctx,
> > >                 cifs_dbg(VFS, "can not change sec during remount\n");
> > >                 return -EINVAL;
> > >         }
> > > +       if (new_ctx->multiuser != old_ctx->multiuser) {
> > > +               cifs_dbg(VFS, "can not change multiuser during remount\n");
> > > +               return -EINVAL;
> > > +       }
> > > +       if (new_ctx->UNC &&
> > > +           (!old_ctx->UNC || strcmp(new_ctx->UNC, old_ctx->UNC))) {
> > > +               cifs_dbg(VFS, "can not change UNC during remount\n");
> > > +               return -EINVAL;
> > > +       }
> > > +       if (new_ctx->username &&
> > > +           (!old_ctx->username || strcmp(new_ctx->username, old_ctx->username))) {
> > > +               cifs_dbg(VFS, "can not change username during remount\n");
> > > +               return -EINVAL;
> > > +       }
> > > +       if (new_ctx->password &&
> > > +           (!old_ctx->password || strcmp(new_ctx->password, old_ctx->password))) {
> > > +               cifs_dbg(VFS, "can not change password during remount\n");
> > > +               return -EINVAL;
> > > +       }
> > > +       if (new_ctx->domainname &&
> > > +           (!old_ctx->domainname || strcmp(new_ctx->domainname, old_ctx->domainname))) {
> > > +               cifs_dbg(VFS, "can not change domainname during remount\n");
> > > +               return -EINVAL;
> > > +       }
> > > +       if (new_ctx->nodename &&
> > > +           (!old_ctx->nodename || strcmp(new_ctx->nodename, old_ctx->nodename))) {
> > > +               cifs_dbg(VFS, "can not change nodename during remount\n");
> > > +               return -EINVAL;
> > > +       }
> > > +       if (new_ctx->iocharset &&
> > > +           (!old_ctx->iocharset || strcmp(new_ctx->iocharset, old_ctx->iocharset))) {
> > > +               cifs_dbg(VFS, "can not change iocharset during remount\n");
> > > +               return -EINVAL;
> > > +       }
> > >
> > >         return 0;
> > >  }
> > >
> > > +#define STEAL_STRING(cifs_sb, ctx, field)                              \
> > > +do {                                                                   \
> > > +       if (ctx->field) {                                               \
> > > +               kfree(ctx->field);                                      \
> > > +               ctx->field = cifs_sb->ctx->field;                       \
> > > +               cifs_sb->ctx->field = NULL;                             \
> > > +       }                                                               \
> > > +} while (0)
>
> If ctx->field is NULL we won't assign new value from
> cifs_sb->ctx->field and the procedure will become no-op. Is this an
> intent?

No, that is a bug. I will fix that.

Thanks.

>
> --
> Best regards,
> Pavel Shilovsky





[Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux