Date: Fri, 27 Mar 2015 06:52:58 -0700 From: David Wolfskill <david@catwhisker.org> To: Adrian Chadd <adrian@freebsd.org> Cc: "net@freebsd.org" <net@freebsd.org> Subject: Re: iwn(4) works (mostly) in stable/10; fails to associate in head Message-ID: <20150327135258.GK7594@albert.catwhisker.org> In-Reply-To: <CAJ-VmokWhNi2EgtFJ7WfLYLi=FA20Q%2B1qYa=U8QvsSaQ8sDAgg@mail.gmail.com> References: <20150323131918.GN7594@albert.catwhisker.org> <CAJ-VmokLk952wf_K03c6hbR8C=RPoaL2UcXDzFQEo5s9=Lskjg@mail.gmail.com> <20150326022132.GI7594@albert.catwhisker.org> <CAJ-VmomiOT2u%2BFdhcYS_%2BY=Sg0HtBGeSAp%2BvktsuDsR5wUptaw@mail.gmail.com> <20150327023111.GB7594@albert.catwhisker.org> <CAJ-VmokWhNi2EgtFJ7WfLYLi=FA20Q%2B1qYa=U8QvsSaQ8sDAgg@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--+F2yqQIdYdj7GirX Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Thu, Mar 26, 2015 at 08:02:37PM -0700, Adrian Chadd wrote: > Ooooooooh. I think you're hitting the "multiple wpa_supplicant > processes are running at startup" bug. :( > ... During a bit of exploring this morning, I'm not seeing evidence that I recognize as corroborating that suggestion. For one thing, the messages logged by wpa_supplicant (in /var/log/messages) report just the one PID (until I manually wade in and start breaking things). For another, there is a /var/run/wpa_supplicant/wlan0.pid created & populated with the PID. (And yes, its contents match the PID logged in the above-cited messages.) During my experiments, I found some "interesting" behavior: * I set up the em0 & iwn0/wlan0 interfaces as described in the "laptop" example (31.3) at <https://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/network-aggre= gation.html> (subsitituting "em0" for "bge0"): g1-254(10.1-S)[5] egrep 'ifconfig|(em|lagg|wlan)0' rc.conf | grep -v '^#' ifconfig_em0=3D"up" ifconfig_iwn0=3D"ether 00:24:e8:b5:85:46" wlans_iwn0=3Dwlan0 ifconfig_wlan0=3D"WPA" cloned_interfaces=3D"lagg0" ifconfig_lagg0=3D"laggproto failover laggport em0 laggport wlan0 DHCP" g1-254(10.1-S)[6]=20 * Immediately subsequent to the transition from single- to multi-user mode: + em0 is active + WiFi LED is blinking rapidly (indicating failed attempts at association) + wlan0 shows "no carrier" + lagg0 shows "active", but the inet address is (still) 0.0.0.0. * If I: + "pgrep wpa_supplicant", only a single PID is returned. + "pkill wpa_supplicant", the WiFi LED goes out; if I then follow that with another "pgrep wpa_supplicant", no PID is returned. + Out of perversity, I then verified that "ifconfig wlan0" reported the status of wlan0 (implying tha wlan0 exists). + "service netif stop wlan0 iwn0", I get a whimper: wpa_supplicant not running? (check /var/run/wpa_supplicant/wlan0.pid). Stopping Network: wlan0 iwn0. ifconfig: interface wlan0 does not exist + An attempt to "ifconfig wlan0"confirms that it no longer exists. + "ifconfig" shows that em0 is still active, as is lagg0. lagg0 also still has no IP address assigned -- and it no longer has wlan0 as a component. + "service netif start iwn0" yields: Starting wpa_supplicant. Starting Network: iwn0. and wlan0 now exists -- and gets association. But wlan0 has not become part of lagg0 as a result of this. + "service netif restart lagg0" yields: Stopping dhclient. Waiting for PIDS: 1035. Stopping Network: lagg0. lagg0: flags=3D8802<BROADCAST,SIMPLEX,MULTICAST> metric 0 mtu 1500 options=3D4019b<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCS= UM,TSO4,VLAN_HWTSO> ether 34:e6:d7:3c:4a:93 nd6 options=3D29<PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL> media: Ethernet autoselect status: active groups: lagg=20 laggproto failover lagghash l2,l3,l4 laggport: em0 flags=3D5<MASTER,ACTIVE> Destroyed clone interfaces: lagg0. Created clone interfaces: lagg0. Starting dhclient. dhclient: /etc/dhclient-enter-hooks invoked with reason PREINIT dhclient: Leaving hostname set to=20 and the WiFi LED went from solid "on" to "blinking madly" -- and the system then locks up hard -- requiring a power-cycle to do anything. (The BIOS has an option for an "unobtrusive mode" -- if this is enabled (which I did), Fn+B will blank the screen and make all lights on the device go out. I had tried this earlier, and it seems to work as advertised; the Fn+B is a toggle. When the above-described "lock up" occurred, I tried "Fn+B".... no reaction at all. Yeah, I tried a few other things -- even switching the WiFi switch to "off" didn't stop the WiFi LED from blinking madly.) (I have a typescript showing that stuff....) Peace, david --=20 David H. Wolfskill david@catwhisker.org Those who murder in the name of God or prophet are blasphemous cowards. See http://www.catwhisker.org/~david/publickey.gpg for my public key. --+F2yqQIdYdj7GirX Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQJ8BAEBCgBmBQJVFWC5XxSAAAAAAC4AKGlzc3Vlci1mcHJAbm90YXRpb25zLm9w ZW5wZ3AuZmlmdGhob3JzZW1hbi5uZXQ4RThEMDY4QTIxMjc1MDZFRDIzODYzRTc4 QTY3RjlDOERFRjQxOTNCAAoJEIpn+cje9Bk7InMP/joB1F0GJvFt7o0RPv8+Mheh Nw7ByVVaH43bAczRZTg4Ef6esmQUubhr0hjDZlKjfzLW8pHcwvO9zwN6ce+pn0+M HJFK7x2v7bEnzDUg2fUUhk4zX0exhLPe9t6f5liv1DM6FwwIcUul3pcB4DkmFN49 qzhvcMY4PZDArWmlQIvXJ+rC2Nar2+2FfMaDHczcHy+FSHdbn4AXdVWy39OVIH7t lwpael0KkfIbL4XKkE9yr0l3Nmgfaf3U9id8PEwy0pgpk9VzVjJNicF+1X/oxuOI 54EGbWFkqr8/JXcxFQG/J97eJkcuQbH0uLJi00hpYZkIaja3DWVcYLEZwJBbHw8y V84mD6z6tTgvlRs7vrVrczaegOyo6Wwvn7jM+1Sa17gEspNvrzzsDWfNdCDGfZI8 kSh08ojMAujzeqNmQ0lNwdFbF2DlNm2fvp8oYIo9GpPxyiMCTqEczKQTEOLMn9GU 4rr6RMl0pwkeTPPUiPrxm2UKa7Y8bq4Oj/NMHe5I3iRpdiLuiRrsNjZfDwZS5DSS cD05sgPQ57r+kndpBZG162M37vBaH6GX7RtKhl+2Cz65qmeDcpF1NpCoy7KCo2q4 BANHY7M3aiPWifK/MNzmb/9DS9Z1kJ0sWNeTEZi5b3VkAYcjIaofRoRA+Jil97wW ikdtpHgBUMvm108y+Tiz =eIdI -----END PGP SIGNATURE----- --+F2yqQIdYdj7GirX--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20150327135258.GK7594>