sidux.com

All the news about sidux - wlan compatibility
slh - Oct 29, 2007 - 03:57 PM
Post subject: wlan compatibility
In order to ensure the best driver compatibility for wlan devices in sidux 2004-04 Έρως, we need your feedback to decide which of the new mac80211 alternatives are stable enough or working better than their legacy counterparts.

The polls are now open for several RaLink legacy drivers (rt2400, rt2500, rt2570, rt61 and rt73) in comparison to their new rt2x00 2.0.12 counterparts (rt2400pci, rt2500pci, rt2500usb, rt61pci, rt73usb). To switch between the alternatives, "/etc/modprobe.d/mac80211" needs to blacklist the unwanted alternative (to revert back to safe defaults, use "dpkg --purge udev-config-sidux ; apt-get update && apt-get install udev-config-sidux).

udev might get confused by driver changes, if you encounter unusable "wlan0_rename" devices, please delete "/etc/udev/rules.d/z25_persistent-net.rules".

rt2x00 2.0.12 is present in kernel >= 2.6.23.1-slh*-smp-18, tests with earlier kernels would blur the polls. Please explain what doesn't work with rt2x00 or add any comments you like.
seagull - Oct 31, 2007 - 12:28 AM
Post subject: new rt61 driver with sidux prerelease works so far
Hi all

i downloaded + installed the new prerelease, copied the firmware from
my last sidux installation.
configured /etc/networks/interfaces:

Code:

# The loopback network interface
auto lo wlan0
iface lo inet loopback


iface wlan0 inet static
        wpa-conf /etc/wpa_supplicant/wpa_supplicant.conf
   address 192.168.2.20
   netmask 255.255.255.0
   network 192.168.2.0
   broadcast 192.168.2.255
   gateway 192.168.2.1


created wpa_supplicant.conf:

Code:

# Static WEP keys

ctrl_interface=/var/run/wpa_supplicant

network={
   ssid="xxxxx"
   key_mgmt=NONE
   wep_key0="MyPassword..."
   wep_key1=""
   wep_tx_keyidx=0
}


stopped and started the network, downloaded 140 MB while downloading watched a youtube video and wrote this short report.
Have a nice day! Wink
slh - Oct 31, 2007 - 01:31 AM
Post subject: RE: new rt61 driver with sidux prerelease works so far
Thanks a lot for the feedback, it's invaluable for choosing which drivers will be default on Ερως (the alternatives will remain available of course, but defaults matter quite a lot).

And the preview should make it rather easy to confirm what works and what doesn't.
mylo - Oct 31, 2007 - 10:54 AM
Post subject: RE: new rt61 driver with sidux prerelease works so far
I am not using wlan on sidux for now as wpaII-encryption is not available. Therefore I ask myself, is it or will it be available in the near future?
slh - Oct 31, 2007 - 12:32 PM
Post subject: RE: new rt61 driver with sidux prerelease works so far
WPA2 is available for most concurrent wlan drivers, using the new rt2x00 drivers also through standard means for RaLink based wlan cards. The legacy driver variants also support wpa2, but of varying stability (rt61 doesn't work longer than 20 minutes, rt2570 doesn't auth at all).
mrsarmitage - Nov 02, 2007 - 12:01 PM
Post subject: RE: new rt61 driver with sidux prerelease works so far
So far I am using the rt61 legacy-drivers (WEP, I know it is not safe) and this works fine for me. I tested R2x00 drivers on the eros live cd and copied firmware from usb stick into lib/firmware. Card is recognised but after configuration looses connection almost immediately. So no go?
Greetings UB
Chrisux - Nov 02, 2007 - 06:35 PM
Post subject:
rt2500 pci läuft wunderbar!!!!!!!!!
das erste mal das es sofort geht!
Super arbeit leute!!!!

Laughing
daldred - Nov 02, 2007 - 09:57 PM
Post subject:
Can I just check this: in order to provide any useful feedback on whether the drivers work in Eros, I need to do a clean install of Eros from CD, yes?

(They don't seem to work properly in my Gaia install, updated fully a couple of days ago, but it may be that I've messed something up in the past trying to get them to go!)
op4latino - Nov 03, 2007 - 04:44 AM
Post subject:
daldred, I'm guessing that an old install might need kernel 2.6.23.1-xx for the new drivers to work, idk.
bfree - Nov 03, 2007 - 09:31 AM
Post subject:
No need to install to test this, you can try it out live.
daldred - Nov 04, 2007 - 12:10 AM
Post subject:
op4latyino: I'm on 2.6.23.1-slh-smp-18 and it's not working, but I think that may be because I've been messing around for a while trying to make it work - and having to drop back to ndiswrapper every time. On kernels before about 2.6.21 the Linux driver worked, and worked better than it ever has with ndiswrapper!


bfree: good point! I'll grab the image and try it on liveCD.
daldred - Nov 04, 2007 - 06:48 PM
Post subject:
Ok, tried running from LiveCD. The card (RT2500 based) appears to come up, both lights are on (thouhg constantly - one of them should flash if it's connecting). I've run NetCardConfig to set it all up as it should be for my network.

But: it fails to find any access point. Since it's sitting under a metre from the wireless router, it's not signal! iwconfig shows correctly, just "Access Point: Not-Associated". ifdown/ifup doesn't change anything.

This is the same behaviour that I've been seeing under Gaia (fully updated). NDISwrapper will drive it under Gaia (though when I had the RT2500 drivers working pre kernel 2.6.21 its ability to connect to my wireless network at the fringes of its availability was far better).

I don't know quite how to vote on this as I don't know whether the legacy driver would work any better!

If anyone's got any ideas.....
mrsarmitage - Nov 07, 2007 - 11:30 AM
Post subject:
After yersterdays d-u I was confronted with the fact that the rt61 legacy driver which I'd always used before wasn't loaded. Instead the rt61pci was used. Ok I thought lets give it a go. After some ifups and ifdown wlan1 the rt61 suddenly connected with low speed (11Mb, 5,5MB etc). Several times it led to a complete freeze of the system. I was using 2.6.22 kernel. I managed then to download the latest kernel and there it is : rt61 wlan card connects to the network and works! In etc/modprobe.d./mac80211 the old serialmonkey driver rt61 driver is blacklisted.

Greetings UB
daldred - Nov 07, 2007 - 06:27 PM
Post subject:
daldred wrote:
Ok, tried running from LiveCD. The card (RT2500 based) appears to come up, both lights are on (thouhg constantly - one of them should flash if it's connecting). I've run NetCardConfig to set it all up as it should be for my network.
This is the same behaviour that I've been seeing under Gaia (fully updated).


Update on this: after more tinkering, I found a thread suggesting removing linux-wlan-ng from the machine. Did so and now it connects using rt2500pci.

However, this doesn't explain why it won't connect on Eros LiveCD - or perhaps it does, if that has linux-wlan-ng installed. If that is the case, then either it needs removing (if that doesn't break other things!) or we need to find a way to stop it doing whatever it's doing which confuses the rt2500pci module!
kelmo - Nov 08, 2007 - 12:52 PM
Post subject:
Null pointer with b43 in latest patchset:

Code:

BUG: unable to handle kernel NULL pointer dereference at virtual address 00000001
 printing eip:
c020fd3e
*pde = 00000000
Oops: 0000 [#1]
PREEMPT SMP
Modules linked in: radeon drm cpufreq_ondemand cpufreq_powersave ipv6 af_packet fuse dm_snapshot dm_mirror dm_mod acpi_cpufreq freq_table capability commoncap arc4 ecb blkcipher rc80211_simple joydev b43 snd_intel8x0 rfkill mac80211 snd_intel8x0m snd_ac97_codec cfg80211 ac97_bus snd_pcm_oss snd_mixer_oss led_class input_polldev battery snd_pcm container snd_timer ac yenta_socket parport_pc parport wbsd mmc_core rsrc_nonstatic button pcspkr snd evdev serio_raw intel_agp soundcore snd_page_alloc iTCO_wdt i2c_i801 iTCO_vendor_support psmouse shpchp i2c_core pci_hotplug ext3 jbd sg sr_mod sd_mod cdrom firewire_ohci firewire_core crc_itu_t ata_piix ohci1394 ieee1394 8139too mii ssb pcmcia pcmcia_core ata_generic ehci_hcd uhci_hcd libata usbcore thermal processor fan
CPU:    0
EIP:    0060:[<c020fd3e>]    Not tainted VLI
EFLAGS: 00010296   (2.6.23.1-slh-smp-22 #1)
EIP is at strcmp+0xe/0x30
eax: 00000001   ebx: f71ce2c0   ecx: 00000000   edx: dff2ddb4
esi: 00000001   edi: dff2ddb4   ebp: 00000000   esp: f71f5d48
ds: 007b   es: 007b   fs: 00d8  gs: 0033  ss: 0068
Process ifconfig (pid: 3137, ti=f71f4000 task=f59aaac0 task.ti=f71f4000)
Stack: f716c2b0 f716c2d0 c029bc8f f716c2b0 00000000 00000001 f8d27219 c19afc00
       dfe28600 00000000 f8d27299 f716c2ea f716c2ea f716c2ac f8e19068 ffffffff
       f8e1cb94 dff2d1c0 f71f5dcc f716c000 dff2d1c0 f716c000 00000001 00000000
Call Trace:
 [<c029bc8f>] led_trigger_set_default+0x5f/0x90
 [<f8d27219>] led_classdev_register+0xb9/0x100 [led_class]
 [<f8e19068>] b43_register_led+0xa8/0xe0 [b43]
 [<f8e1921f>] b43_leds_init+0x17f/0x2c0 [b43]
 [<f8bd23e2>] ssb_pci_switch_core+0x42/0x60 [ssb]
 [<f8e08e8d>] b43_chip_init+0x2fd/0x870 [b43]
 [<f8bd3a0a>] ssb_pcicore_dev_irqvecs_enable+0x11a/0x330 [ssb]
 [<f8e095ca>] b43_wireless_core_init+0x1ca/0x6f0 [b43]
 [<c01978d3>] mntput_no_expire+0x13/0x80
 [<c018a69f>] __link_path_walk+0x1df/0xde0
 [<f8e0a1d7>] b43_start+0x77/0x90 [b43]
 [<f8de6b17>] ieee80211_open+0x1a7/0x440 [mac80211]
 [<c01608f5>] filemap_fault+0x215/0x470
 [<c02b3904>] dev_open+0x34/0x80
 [<c02b1822>] dev_change_flags+0x82/0x190
 [<c02fe605>] devinet_ioctl+0x505/0x690
 [<c02b32da>] dev_ifsioc+0x13a/0x370
 [<c02b291a>] dev_load+0x2a/0x60
 [<c02a5d2f>] sock_ioctl+0xbf/0x210
 [<c011dc44>] do_page_fault+0x484/0x670
 [<c02a5c70>] sock_ioctl+0x0/0x210
 [<c018d7bb>] do_ioctl+0x2b/0x90
 [<c018da4e>] vfs_ioctl+0x22e/0x2b0
 [<c018db2d>] sys_ioctl+0x5d/0x70
 [<c010440a>] syscall_call+0x7/0xb
 [<c0320000>] sonet_copy_stats+0x30/0x40
 =======================
Code: 75 f7 31 c0 aa 8b 04 24 8b 5c 24 04 8b 74 24 08 8b 7c 24 0c 8b 6c 24 10 83 c4 14 c3 90 83 ec 08 89 34 24 89 c6 89 7c 24 04 89 d7 <ac> ae 75 08 84 c0 75 f8 31 c0 eb 04 19 c0 0c 01 8b 34 24 8b 7c
EIP: [<c020fd3e>] strcmp+0xe/0x30 SS:ESP 0068:f71f5d48
note: ifconfig[3137] exited with preempt_count 2
BUG: scheduling while atomic: ifconfig/0x10000003/3137
 [<c0324d64>] schedule+0x5d4/0x6e0
 [<c0292851>] i8042_interrupt+0x111/0x260
 [<c02691c7>] vt_console_print+0x67/0x2b0
 [<c020f265>] vsnprintf+0x2d5/0x610
 [<c0269160>] vt_console_print+0x0/0x2b0
 [<c0123968>] __cond_resched+0x18/0x30
 [<c032532a>] cond_resched+0x2a/0x40
 [<c016b399>] unmap_vmas+0x519/0x5b0
 [<c016e56e>] exit_mmap+0x7e/0x120
 [<c012585e>] mmput+0x1e/0x80
 [<c012ae90>] do_exit+0x120/0x840
 [<c0105a1d>] die+0x23d/0x250
 [<c011da5f>] do_page_fault+0x29f/0x670
 [<c020aa8b>] ida_get_new_above+0x7b/0x180
 [<c011d7c0>] do_page_fault+0x0/0x670
 [<c0326e22>] error_code+0x72/0x78
 [<c01c007b>] amiga_partition+0x29b/0x3e0
 [<c020fd3e>] strcmp+0xe/0x30
 [<c029bc8f>] led_trigger_set_default+0x5f/0x90
 [<f8d27219>] led_classdev_register+0xb9/0x100 [led_class]
 [<f8e19068>] b43_register_led+0xa8/0xe0 [b43]
 [<f8e1921f>] b43_leds_init+0x17f/0x2c0 [b43]
 [<f8bd23e2>] ssb_pci_switch_core+0x42/0x60 [ssb]
 [<f8e08e8d>] b43_chip_init+0x2fd/0x870 [b43]
 [<f8bd3a0a>] ssb_pcicore_dev_irqvecs_enable+0x11a/0x330 [ssb]
 [<f8e095ca>] b43_wireless_core_init+0x1ca/0x6f0 [b43]
 [<c01978d3>] mntput_no_expire+0x13/0x80
 [<c018a69f>] __link_path_walk+0x1df/0xde0
 [<f8e0a1d7>] b43_start+0x77/0x90 [b43]
 [<f8de6b17>] ieee80211_open+0x1a7/0x440 [mac80211]
 [<c01608f5>] filemap_fault+0x215/0x470
 [<c02b3904>] dev_open+0x34/0x80
 [<c02b1822>] dev_change_flags+0x82/0x190
 [<c02fe605>] devinet_ioctl+0x505/0x690
 [<c02b32da>] dev_ifsioc+0x13a/0x370
 [<c02b291a>] dev_load+0x2a/0x60
 [<c02a5d2f>] sock_ioctl+0xbf/0x210
 [<c011dc44>] do_page_fault+0x484/0x670
 [<c02a5c70>] sock_ioctl+0x0/0x210
 [<c018d7bb>] do_ioctl+0x2b/0x90
 [<c018da4e>] vfs_ioctl+0x22e/0x2b0
 [<c018db2d>] sys_ioctl+0x5d/0x70
 [<c010440a>] syscall_call+0x7/0xb
 [<c0320000>] sonet_copy_stats+0x30/0x40
 =======================


Earlier (~slh-smp-Cool were ok, iirc.
daldred - Nov 08, 2007 - 09:53 PM
Post subject:
daldred wrote:
daldred wrote:
Ok, tried running from LiveCD. The card (RT2500 based) appears to come up, both lights are on (thouhg constantly - one of them should flash if it's connecting). I've run NetCardConfig to set it all up as it should be for my network.
This is the same behaviour that I've been seeing under Gaia (fully updated).


Update on this: after more tinkering, I found a thread suggesting removing linux-wlan-ng from the machine. Did so and now it connects using rt2500pci.



Uurgh. Or not.

Having started the machine again today, using rt2500pci, the card failed to associate to an access point (as before).

I was going to reverse the changes (reinstall linux-wlan-ng and switch back to ndiswrapper sot he thing was usable!). Did the install, and accidentally did ifup wlan0 at that stage (hit enter on the wrong line on a bash history!). It complained about a missing p80211.ko (which was the clue which led me to uninstall linux-wlan-ng in the first place) - then the card came up and associated correctly!

However after restarting from scratch, we're back to a non-associating card.

Something is flaky here. This is (I repeat) on Gaia fully updated rather than on an actual Eros install; I think I'll leave things now until Eros release and try again with a clean install. I've mucked about with it so much I've probably broken something!
op4latino - Nov 08, 2007 - 10:00 PM
Post subject:
I can confirm the null pointer that kelmo posted using b43
Thanks
houms - Dec 03, 2007 - 11:57 PM
Post subject: Re: RE: new rt61 driver with sidux prerelease works so far
mylo wrote:
I am not using wlan on sidux for now as wpaII-encryption is not available. Therefore I ask myself, is it or will it be available in the near future?


I have wpa2 working, and have not had any connectivity issues... connects automagically during login, and i have been very pleased with it over all. Though i'm not using an rt based card (ipw2100). The only catch is you have to set a static ip address in your /etc/network/interfaces.
All times are GMT - 1 Hours
Powered by PNphpBB2 © 2003-2007 The PNphpBB Group
Credits