Hi Reinette, > > This totally messes up iwlwifi: http://paste.pocoo.org/show/113789/ > > > > You'll see: > > [ 1455.006896] phy0: turning off radio - idle > > ... > > [ 1455.020180] phy0: turning on radio - scan > > [ 1455.020202] ieee80211 phy0: U iwl_mac_config enter to channel 1 changed 0x1 > > [ 1455.031724] ieee80211 phy0: U iwl_mac_config leave - RF-KILL - waiting for uCode > > [ 1455.031739] ieee80211 phy0: U iwl_mac_config leave > > [ 1455.031752] ieee80211 phy0: U iwl_mac_hw_scan enter > > [ 1455.031764] ieee80211 phy0: U iwl_mac_hw_scan leave - not ready or exit pending > > [ 1455.031798] phy0: turning off radio - idle > > > > at that point, things get so confused that you can't even scan any more > > because something isn't properly reporting to cfg80211 that the scan > > ended (but wasn't successful.) > > This seems related to the discussion about a patch we tried to submit. > See > http://thread.gmane.org/gmane.linux.kernel.wireless.general/30196/focus=30280 Related, but a different situation. Also see the patch I sent to iwlwifi. In the mail you quoted, mac80211 certainly handles an error if it's returned directly, but if ->hw_scan() returns 0 (and only then) you need to call the ieee80211_scan_completed() function at some point. johannes
Attachment:
signature.asc
Description: This is a digitally signed message part
