[rt2x00-users] [PATCH] rt2x00: fix beacon reset on rt2800

[Helmut Schaa]

Am Dienstag 22 Juni 2010 schrieb Ivo Van Doorn:
> On Tue, Jun 22, 2010 at 10:14 AM, Helmut Schaa
> <helmut.schaa at googlemail.com> wrote:
> > When an interface is removed the according beacon entry should be reset.
> > The current approach to only clear the first word is not enough to stop
> > the device from sending out the beacon, hence resulting in beacons being
> > sent out for already removed interfaces.
> >
> > Fix this by invalidating the entire TXWI in front of the beacon instead
> > of only the first word.
> >
> > Signed-off-by: Helmut Schaa <helmut.schaa at googlemail.com>
> > ---
> >  drivers/net/wireless/rt2x00/rt2800lib.c |    8 +++++---
> >  1 files changed, 5 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/net/wireless/rt2x00/rt2800lib.c b/drivers/net/wireless/rt2x00/rt2800lib.c
> > index f25997e..d53f680 100644
> > --- a/drivers/net/wireless/rt2x00/rt2800lib.c
> > +++ b/drivers/net/wireless/rt2x00/rt2800lib.c
> > @@ -735,16 +735,18 @@ void rt2800_config_intf(struct rt2x00_dev *rt2x00dev, struct rt2x00_intf *intf,
> >  {
> >        unsigned int beacon_base;
> >        u32 reg;
> > +       int i;
> >
> >        if (flags & CONFIG_UPDATE_TYPE) {
> >                /*
> >                 * Clear current synchronisation setup.
> >                 * For the Beacon base registers we only need to clear
> > -                * the first byte since that byte contains the VALID and OWNER
> > -                * bits which (when set to 0) will invalidate the entire beacon.
> > +                * the whole TXWI which (when set to 0) will invalidate
> > +                * the entire beacon.
> >                 */
> 
> Hmm off course, rt2800 doesn't have a owner and valid field. :S
> 
> >                beacon_base = HW_BEACON_OFFSET(intf->beacon->entry_idx);
> > -               rt2800_register_write(rt2x00dev, beacon_base, 0);
> > +               for (i = 0; i < TXWI_DESC_SIZE; i += sizeof(__le32))
> > +                       rt2800_register_write(rt2x00dev, beacon_base + i, 0);
> 
> I think doing rt2800_register_multiwrite would be easier/faster. :)

Yes, but we would have to pass a 16byte long buffer to it. That's why I used
the loop. Or did I miss anything?

Helmut