Re: [PATCH] /dev/mem: Allow rewinding

[Wu Fengguang <fengguang.wu@intel.com>]

On Tue, Apr 06, 2010 at 01:30:29PM +0800, Américo Wang wrote:
> On Tue, Apr 6, 2010 at 5:19 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> > Le lundi 05 avril 2010 à 23:04 +0200, Eric Dumazet a écrit :
> >> Andrew, commit dcefafb6ac90ece8d68a6c203105f3d313e52da4 seems to be the
> >> problem.
> >>
> >> It breaks rewinds (negative offsets to lseek (... SEEK_CUR))
> >>
> >
> > Here is a patch to make rewind working again on /dev/mem
> >
> > [PATCH] /dev/mem: Allow rewinding
> >
> > commit dcefafb6 (/dev/mem: dont allow seek to last page)
> > inadvertently disabled rewinding on /dev/mem.
> >
> > This broke x86info for example.
> >
> > Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
> > ---
> > diff --git a/drivers/char/mem.c b/drivers/char/mem.c
> > index 1f3215a..3973a1d 100644
> > --- a/drivers/char/mem.c
> > +++ b/drivers/char/mem.c
> > @@ -710,11 +710,6 @@ static loff_t memory_lseek(struct file *file, loff_t offset, int orig)
> >        switch (orig) {
> >        case SEEK_CUR:
> >                offset += file->f_pos;
> > -               if ((unsigned long long)offset <
> > -                   (unsigned long long)file->f_pos) {
> > -                       ret = -EOVERFLOW;
> > -                       break;
> > -               }
> 
> Why completely dropping the overflow check? What you need to do is just
> adding the 'offset < 0' case check.

It will fall through to the next EOVERFLOW check in the SEEK_SET case :)

Thanks,
Fengguang