Subject: Re: 11 Jun 2005 ath/net80211: fixes WEP
To: None <tech-net@NetBSD.org, current-users@NetBSD.org>
From: Daniel Carosone <dan@geek.com.au>
List: tech-net
Date: 06/13/2005 17:27:22
--xhJxprRH8Q4al5cy
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sun, Jun 12, 2005 at 04:24:09PM -0500, David Young wrote:
> [new versions with WEP glue updated]

Thanks for all your work on this!

> Tomorrow, I am going to commit all of this
> stuff to NetBSD-current.

Alas, you may want to hold off that just for a moment :-(

I did some testing with a mixture of cards:

 - ath(4) dlink DWL-AG660 (AR5004X) cardbus
 - ath(4) netgear WG311T (AR5002G) PCI
 - atw(4) no-name (ADM8211) PCI
 - wi(4) netgear MA401 (PRISM2.5) pcmcia
 - wi(4) lucent silver (Hermes) pcmcia

There were some rather mixed results

 - The DWL-AG660 had previously worked fine (despite being a higher
   rev chip than 'officially' supported by the older driver), although
   the 'turbo' 108Mbps modes hadn't been detected.  These are now
   detected and reported capabilities with 'ifconfig -m', but only
   with mode 11g, not 11a - the card should be capable of 108Mbps in
   either band.

 - The WG311T had previously not attached (HAL status 13). With this
   new code, it now attaches and seems to work - and with this latest
   update WEP also works. =20

   However, neither card seems to work with 'mediaopt turbo' with my
   Netgear WGU624 dual-108 AP; the cards drop link with this option
   set.  The AP is set to auto-108 on both bands.=20

   Apart from these nits, both ath cards seem to work fine with some
   basic testing and traffic.

 - the atw(4) card panics at probe time:

atw0 at pci2 dev 2 function 0: ADMtek ADM8211 802.11 MAC/BBP, revision 1.1
atw0: interrupting at irq 10
uvm_fault(0xc0604ce0, 0, 0, 1) -> 0xe
kernel: page fault trap, code=3D0
Stopped in pid 0.1 (swapper) at netbsd:atw_write_sram+0xde:     testb   $0x=
4,0x30(%eax)
db> tr
atw_write_sram(c1121000,0,c1122568,400,ffffffff) at netbsd:atw_write_sram+0=
xde
atw_clear_sram(c1121000,0,c350,40,1) at netbsd:atw_clear_sram+0x5a
atw_reset(c1121000,800,1,800,0) at netbsd:atw_reset+0x193
atw_attach(c1121000,c112101c,c05eef20,c01ed900,c1121000) at netbsd:atw_atta=
ch+0x27f
atw_pci_attach(c10cd000,c1121000,c0691940,1,0) at netbsd:atw_pci_attach+0x2=
f9
config_attach_loc(c10cd000,c059d140,c0691930,c0691940,c03d1130) at netbsd:c=
onfig_attach_loc+0x31c
pci_probe_device(c10cd000,80021000,0,0,0) at netbsd:pci_probe_device+0x260
pci_enumerate_bus(c10cd000,c04c0264,0,0,c10cd000) at netbsd:pci_enumerate_b=
us+0x18c
pcirescan(c10cd000,c053d4b1,c04c0264,37,0) at netbsd:pcirescan+0x24
config_attach_loc(c11125c0,c059d140,0,c0691ba0,c03d4740) at netbsd:config_a=
ttach_loc+0x31c
ppbattach(c1057000,c11125c0,c0691cc0,c030d2af,0) at netbsd:ppbattach+0x120
config_attach_loc(c1057000,c059d140,c0691cb0,c0691cc0,c03d1130) at netbsd:c=
onfig_attach_loc+0x31c
pci_probe_device(c1057000,80008800,0,0,0) at netbsd:pci_probe_device+0x260
pci_enumerate_bus(c1057000,c04c0264,0,0,c1057000) at netbsd:pci_enumerate_b=
us+0x18c
pcirescan(c1057000,c053d4b1,c04c0264,10,0) at netbsd:pcirescan+0x24
config_attach_loc(c103ef80,c059d140,0,c0691e30,c03d4740) at netbsd:config_a=
ttach_loc+0x31c=20
mainbus_attach(0,c103ef80,0,c0305b12,0) at netbsd:mainbus_attach+0x20c
config_attach_loc(0,c059d140,0,0,0) at netbsd:config_attach_loc+0x31c
config_rootfound(c053d1d9,0,c0691f58,c0305701,c054775b) at netbsd:config_ro=
otfound+0x4a   =20
cpu_configure(c054775b,c054822f,c05a5200,c0605080,c059d160) at netbsd:cpu_c=
onfigure+0x2c  =20
configure(c0604ec0,1,0,bfeff000,0) at netbsd:configure+0x41
main(0,0,0,0,0) at netbsd:main+0x30f
db>=20

 - as do both the wi cards on insertion (or at boot time if already
   inserted):

cbb1: wait took 0.009685s
wi0 at pcmcia1 function 0: <Lucent Technologies, WaveLAN/IEEE, Version 01.0=
1, >
wi0: 802.11 address 00:60:1d:f6:32:b7
wi0: using Lucent Technologies, WaveLAN/IEEE
wi0: Lucent Firmware: Station (8.10.1)
kernel: page fault trap, code=3D0
Stopped in pid 11.1 (cardslot1) at      netbsd:wi_attach+0x16f: movl    %ea=
x,0x14(%edx)
db> tr
wi_attach(c1192000,c1197160,c15e9540,0,c1192000) at netbsd:wi_attach+0x16f
wi_pcmcia_attach(c114a400,c1192000,cb33df04,0,0) at netbsd:wi_pcmcia_attach=
+0xc1
config_attach_loc(c114a400,c059d140,cb33def4,cb33df04,c0421f80) at netbsd:c=
onfig_attach_loc+0x31c
pcmcia_rescan(c114a400,c053d49e,c051b7a0,c111eb80,c111eb80) at netbsd:pcmci=
a_rescan+0xb4
pcmcia_card_attach(c114a400,20,c0558fd0,0,0) at netbsd:pcmcia_card_attach+0=
xa8
cardslot_event_thread(c111eb80,68d000,695000,0,c0100321) at netbsd:cardslot=
_event_thread+0x234
db>=20

The wi cards had been working fine with the previous tarball/patches,
except for WEP.  The atw I hadn't tried until today.

--
Dan.
--xhJxprRH8Q4al5cy
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (NetBSD)

iD8DBQFCrTVaEAVxvV4N66cRAqpOAKC5abHUqHSOqy6z2loxQVTo8HZqUQCghnPv
t1ohvfh1o2VeUpS9JBFNCGs=
=5S3T
-----END PGP SIGNATURE-----

--xhJxprRH8Q4al5cy--