On 12/15/2017 10:49 AM, Greg KH wrote:
On Fri, Dec 15, 2017 at 10:29:29AM +0100, Richard Schütz wrote:From: Ilan peer <ilan.peer@xxxxxxxxx> commit 57629915d568c522ac1422df7bba4bee5b5c7a7c upstream. The code was setting the capabilities byte to zero, after it was already properly set previously. Fix it. The bug was found while debugging hwsim mesh tests failures that happened since the commit mentioned below. Fixes: 76f43b4c0a93 ("mac80211: Remove invalid flag operations in mesh TSF synchronization") Signed-off-by: Ilan Peer <ilan.peer@xxxxxxxxx> Reviewed-by: Masashi Honma <masashi.honma@xxxxxxxxx> Signed-off-by: Johannes Berg <johannes.berg@xxxxxxxxx> Signed-off-by: Richard Schütz <rschuetz@xxxxxxxxxxxxxx> --- net/mac80211/mesh.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index cc2a63bd233f..9c23172feba0 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -279,8 +279,6 @@ int mesh_add_meshconf_ie(struct ieee80211_sub_if_data *sdata, /* Mesh PS mode. See IEEE802.11-2012 8.4.2.100.8 */ *pos |= ifmsh->ps_peers_deep_sleep ? IEEE80211_MESHCONF_CAPAB_POWER_SAVE_LEVEL : 0x00; - *pos++ = 0x00; - return 0; }--2.15.1I'm sorry, I don't understand what you want done with this patch. Why send it to stable@? confused, greg k-h
AUTOSEL picked up 76f43b4c0a93 for stable a short while ago, but not this accompanying fix. Currently wireless mesh networking is severely impacted in stable because of that.
-- Richard
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature