>
> On Tue, May 20, 2014 at 5:32 PM, John W. Linville <linville@xxxxxxxxxxxxx>
> wrote:
> > On Tue, May 20, 2014 at 01:37:18PM +0300, Emmanuel Grumbach wrote:
> >> From: Avri Altman <avri.altman@xxxxxxxxx>
> >>
> >> Enabling beacon filtering should be done only after a beacon has been
> >> received. Doing that too early will cause disconnections.
> >> This has already been fixed, but the fix didn't take care about the
> >> case where the beacon is received after the association, it waited
> >> only for association which is not enough.
> >>
> >> Signed-off-by: Avri Altman <avri.altman@xxxxxxxxx>
> >> Reviewed-by: Johannes Berg <johannes.berg@xxxxxxxxx>
> >> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@xxxxxxxxx>
> >> ---
> >> drivers/net/wireless/iwlwifi/mvm/mac80211.c | 13 +++---
> >> drivers/net/wireless/iwlwifi/mvm/mvm.h | 5 ---
> >> drivers/net/wireless/iwlwifi/mvm/power.c | 63 +++++++++--------------
> ------
> >> 3 files changed, 27 insertions(+), 54 deletions(-)
> >
> > It is very late in the release cycle for a fix of this size.
> > How serious is the issue? A random disconnect doesn't seem like a
> > huge problem so long as it isn't common...?
> >
> fwiw, it looks like you can simply workaround the issue by setting the
> IEEE80211_HW_NEED_DTIM_BEFORE_ASSOC flag.
>
That would slow down the connection. I'd rather disable the beacon filtering altogether.
The irony is that I sent a mail to Greg asking how to disable it in 3.13 / 3.14 without disabling it in 3.15 because I though this patch would make it to 3.15. Now I find myself disabling the feature for 3.15 too... Anyway...
��.n��������+%������w��{.n�����{���zW����ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f
