Jump to content

Intel Wireless 2100, 2200bg, 2915bg, 3945abg, 4965agn

jalavoui

By jalavoui,
in LAN and Wireless

 Share
3411 posts in this topic

Recommended Posts

Tin_man_0
Tin_man_0
Posted
Posted

So after, formatting, zeroing out and installing a fresh new iATKOS (i figured out why I wan't able to load the compiled kexts - I had opted to install the 10.4 wifi kexts durring install). I finally got my KPs! I only hope that my logs shed more light. :(

dmesg.txt

ioreg.txt

system.txt

Tin_man_0
Tin_man_0
Posted
Posted

so, I decide to load the driver using the kextload method this time arround. I saw that there was a new Eathernet interface in my network settings... I also notice that my existing Airport device that was already installed in my system turned into a standard NIC (still labeld airport but the settings options changed). I used the networkSelector to turn on the device... although I couldn't be sure if it was turning on the device or scanning because it gave no feedback when I select option 1 or 2. is this normal? based on the logs it looks like it turned my card on, but you're the gurus. My 'puter musta been on for a good 15min before I got the KP. I'll attache my new logs.

 

P.S. - jalavoui had mentioned some command or something to produce a more verbose loglevel and I can't seem to find it in the previous posts... anyone remember what it is?

dmesg.txt

ioreg.txt

system.txt

flynhawaiian
flynhawaiian
Posted
Posted

I saw something interesting in lastsoft and tin_man_0's system.txt

 

"error iwl3945: Failed to get channel info for channel 0 [3]"

 

Not sure if I'm correct but I don't think there is a channel 0, but I could be wrong.

 

Just saw something that might be causing problems not sure if it is or not :-)

MrPenguin_2
MrPenguin_2
Posted
Posted
I saw something interesting in lastsoft and tin_man_0's system.txt

 

"error iwl3945: Failed to get channel info for channel 0 [3]"

 

Not sure if I'm correct but I don't think there is a channel 0, but I could be wrong.

 

Just saw something that might be causing problems not sure if it is or not :-)

 

Yeah, I think you're right. It goes 1-11 doesn't it?

jalavoui
jalavoui
Posted
  • Author
Posted
so, I decide to load the driver using the kextload method this time arround. I saw that there was a new Eathernet interface in my network settings... I also notice that my existing Airport device that was already installed in my system turned into a standard NIC (still labeld airport but the settings options changed). I used the networkSelector to turn on the device... although I couldn't be sure if it was turning on the device or scanning because it gave no feedback when I select option 1 or 2. is this normal? based on the logs it looks like it turned my card on, but you're the gurus. My 'puter musta been on for a good 15min before I got the KP. I'll attache my new logs.

 

P.S. - jalavoui had mentioned some command or something to produce a more verbose loglevel and I can't seem to find it in the previous posts... anyone remember what it is?

 

best logs ever seen s far :)

you manage to make the driver start scanning correctly

looks like all it takes is a few reboots - i really need to find out why...

note: networkSelector is disabled don't use it

long system.log - use console - reload to see full system.log

glorybox
glorybox
Posted
Posted
best logs ever seen s far :)

you manage to make the driver start scanning correctly

looks like all it takes is a few reboots - i really need to find out why...

note: networkSelector is disabled don't use it

long system.log - use console - reload to see full system.log

 

 

i wonder is there a way to make driver write to system.log or some other log in -single user mode?

jalavoui
jalavoui
Posted
  • Author
Posted

iwi3945 rel 696

first version for today

if you get errors try a few reboots to check if the driver gets more stable - just see Tin_man_0 system.log as an example

 

glorybox,

if boot -s you need to wait for mac os finish sh etc/rc to make system.log works

sometimes the driver don't show initial log lines - don't know why

can you try to remove this lines from code?

//hack

ieee80211_sta_start_scan(local->mdev, NULL, 0);

the driver isn't expected to work this way

it should call the __ieee80211_rx() them ieee80211_invoke_rx_handlers(local, local->rx_handlers

if you manage to make this work it will help a lot

i don't know why local->rx_handlers aren't called

if the driver can make this call it will start scanning without hacks

a good way to see if this functions is called is to set the driver to give less debug msgs:

put static int iwl3945_param_debug = 0;//0xffffffff & ~(IWL_DL_IO | IWL_DL_ISR | IWL_DL_POWER | IWL_DL_TEMP | IWL_DL_INFO);

 

if you have the card and xcode please try to help me on this

i'm wasting to much time here trying to guess it without the card

 

if anyone as the iwl3945 logs from linux please post it here

i can't find the old 200 pages thread - it as this file

glorybox
glorybox
Posted
Posted
iwi3945 rel 696

first version for today

if you get errors try a few reboots to check if the driver gets more stable - just see Tin_man_0 system.log as an example

 

glorybox,

if boot -s you need to wait for mac os finish sh etc/rc to make system.log works

sometimes the driver don't show initial log lines - don't know why

can you try to remove this lines from code?

//hack

ieee80211_sta_start_scan(local->mdev, NULL, 0);

the driver isn't expected to work this way

it should call the __ieee80211_rx() them ieee80211_invoke_rx_handlers(local, local->rx_handlers

if you manage to make this work it will help a lot

i don't know why local->rx_handlers aren't called

if the driver can make this call it will start scanning without hacks

a good way to see if this functions is called is to set the driver to give less debug msgs:

put static int iwl3945_param_debug = 0;//0xffffffff & ~(IWL_DL_IO | IWL_DL_ISR | IWL_DL_POWER | IWL_DL_TEMP | IWL_DL_INFO);

 

if you have the card and xcode please try to help me on this

i'm wasting to much time here trying to guess it without the card

 

if anyone as the iwl3945 logs from linux please post it here

i can't find the old 200 pages thread - it as this file

 

i will surely try later today. today we have the first sunny day, so want to spend sometime outdoors ;)

Tin_man_0
Tin_man_0
Posted
Posted

he's my 696 logs...

 

Question.... does the kext get put into the extentions folder automaticaly after build?

 

I'm using the command "sudo xcodebuild"

 

this time arround my computer started KPing right after build and when I booted in -s I saw the iwi3945 in my extientions folder. So i deleted iwi3945.kext and Extentions.mkext rebooted and re-ran kext using the kextload command, is this normal behaviour?

dmesg.txt

ioreg.txt

system.txt

jalavoui
jalavoui
Posted
  • Author
Posted

Tin_man_0,

 

you can change the build script in xcode to put iwi3945.kext in /

the default is system/library/extensions

this doesn't help debugging because it make the kext to autoload after rebooting

better use kextload to do it

can you post your ifconfig?

 

iwi3945 rel 697

this is getting better

by some unknown reason Tin_man_0 shows the driver working... weird...

it's crucial to see the first driver log lines - i can't see it in the logs you've send me

 

this line is very important:

en1: DHCP duplicate configured service

are you guys testing iwi3945.kext with other network devices connected

disconnect ALL other devices or this will never work

you can add en1 to system preferences but don't change any parameter

the driver will work without any changes

one good way to do this:

remove iwi3945.kext

reboot

remove en1 device from system preferences - network

reboot

kextload /iwi3945.kext

goto system preferences and add en1 (it should be autodetected)

check system.log

Roal2488
Roal2488
Posted
Posted
his this original svn code or have you changed something?

 

No, it is the original SVN, i've noticed that i get different results every time i boot with -s, some times i just get a KP right away, other times i get it after what seems to be a scanning process. I've noticed too that my network name appears after the number chain in some of the logs for example here:

Apr 13 11:36:55 rodrigo-ascanios-mac-pro kernel[0]: IOPCCard debug:  iwl data: 0: 80 00 00 00 ff ff ff ff ff ff 00 14 a5 91 b7 2a  ...............*
Apr 13 11:36:54 rodrigo-ascanios-mac-pro rpc.statd[51]: statd.notify - no notifications needed
Apr 13 11:36:55 rodrigo-ascanios-mac-pro kernel[0]: IOPCCard debug:  iwl data: 10: 00 14 a5 91 b7 2a 50 00 ff 8e 93 01 00 00 00 00  .....*P.........
Apr 13 11:36:55 rodrigo-ascanios-mac-pro kernel[0]: IOPCCard debug:  iwl data: 20: 64 00 01 04 00 07 46 54 34 36 30 39 39 01 04 82  d.....FT46099...
Apr 13 11:36:55 rodrigo-ascanios-mac-pro kernel[0]: IOPCCard debug:  iwl data: 30: 84 8b 96 03 01 01 05 04 00 01 01 00 2a 01 00 2f  ............*../
Apr 13 11:36:55 rodrigo-ascanios-mac-pro kernel[0]: IOPCCard debug:  iwl data: 40: 01 00 32 08 0c 12 18 24 30 48 60 6c dd 05 00 10  ..2....$0H`l....
Apr 13 11:36:55 rodrigo-ascanios-mac-pro kernel[0]: IOPCCard debug:  iwl data: 50: 18 01 00                                         ...

 

FT46099 is my networks name.

 

i couldn't get the logs every time i booted but so far is the best thing i've gotten. i have some pics of the different KPs, hope u can find them some use for developing.

 

I also noticed that one time the driver made a complete scan of all 11 channels.

13042008383.jpg.zip

13042008384.jpg.zip

13042008385.jpg.zip

jalavoui
jalavoui
Posted
  • Author
Posted

iwi3945 rel 698

made a mistake on last rel

this should fix it

 

imwithalwood,

this is the source code

what i need is the system.log of the linux driver

the most recent: iwl3945

 

iwi3945 testing

please red previous system.log

the driver is unstable and you might need to do some reboots to get it work

i can't use logs full of bugs - check them before posting

JammerML
JammerML
Posted
Posted
iwi3945 rel 698

made a mistake on last rel

this should fix it

 

imwithalwood,

this is the source code

what i need is the system.log of the linux driver

the most recent: iwl3945

 

iwi3945 testing

please red previous system.log

the driver is unstable and you might need to do some reboots to get it work

i can't use logs full of bugs - check them before posting

 

 

Where can I access these drivers, I see they are not available from your google code site.

Tin_man_0
Tin_man_0
Posted
Posted

jalavoui - where in xcode can I change the build setting because every time I build I immediatly get a kp and have to boot in -s remove the kext and reboot again to do kextload.

 

P.S. here's my last logs (i think they're build 697) - I had a hard lock this time arround.

dmesg.txt

ioreg.txt

system.txt

jalavoui
jalavoui
Posted
  • Author
Posted
jalavoui i can get logs from my ubuntu live??? i know that the i3945 works.how can i get theme is it the same command???

i think they're in /var/log folder

 

jalavoui - where in xcode can I change the build setting because every time I build I immediatly get a kp and have to boot in -s remove the kext and reboot again to do kextload.P.S. here's my last logs (i think they're build 697) - I had a hard lock this time arround.

all developers should change the script

in xcodein the icons (expand)

Target

them

iwi3945

them you see Run Script

change the linecp -rf build/Debug/iwi3945.kext /System/Library/Extensions

to

cp -rf build/Debug/iwi3945.kext /

this will allow you to use

kextload /iwi3945.kext

after reboot to test the driver

 

iwi3945 rel 699

fix ieee80211_sta_timer

this can make the driver work much better

but i think you need a few reboots/kextunload to make it give good logs

at this time i suspect

stabssid=00:00:00:00:00:00

is the driver bug

Guest
This topic is now closed to further replies.
 Share
Go to topic listing
×
×
  • Create New...