N900 as a Wifi Hotspot, part 2 - Page 33 - maemo.org

rambo | # 321 | 2010-11-30, 08:37 | Report

Originally Posted by JayBEE View Post

As promised is the output of MH when run from the command-line:

Code:

icd2_state_listener got args:

(dbus.String(u''), dbus.UInt32(0L), dbus.String(u''), dbus.String(u'GPRS'), dbus.UInt32(83886080L), dbus.ByteArray('T-Mobile Internet\x00'), dbus.String(u'com.nokia.icd.error.ppp_authentication_failed'), dbus.UInt32(0L))
state['icd']:

{}

Interesting, no doubt this is what causes the password dialog to pop up, does it ask password when you connect normally ?

Can you use DBUS-monitor to see what connection name is used when you connect normally (I wonder if MHS is trying to connect to wrong connection or something, remember tha fMMS and any other program that automagically tweaks your 3G connections' settings is going to play merry hell with MHS [I probably should add an option that disables the automatic 3G connection management])

Originally Posted by JayBEE View Post

Code:

icd2_state_listener got args:

(dbus.String(u''), dbus.UInt32(0L), dbus.String(u''), dbus.String(u'GPRS'), dbus.UInt32(83886080L), dbus.ByteArray('0000-0000-0000-0000\x00'), dbus.String(u''), dbus.UInt32(2L))
state['icd']:

{'old_connection_type': 'GPRS', 'old_connection_name': '0000-0000-0000-0000'}

And here it somehow manages to mess up the way we store the old connection so we can restore it (which is interesting since once the connection name is stored the backend no longer enters mainloop() so it should not be able to receive signals... [I'm trying to figure out how to unregister a signal since I would need that to be able to detect when the new connection is finalized reliably without making the single connection state signal processor overly complex)

JayBEE | # 322 | 2010-11-30, 17:34 | Report

Originally Posted by rambo View Post

Interesting, no doubt this is what causes the password dialog to pop up, does it ask password when you connect normally ?

Thanks for your response. I almost missed this. To answer your question, no, I'm never prompted for a password when normally connecting.

Originally Posted by

Can you use DBUS-monitor to see what connection name is used when you connect normally (I wonder if MHS is trying to connect to wrong connection or something, remember tha fMMS and any other program that automagically tweaks your 3G connections' settings is going to play merry hell with MHS [I probably should add an option that disables the automatic 3G connection management])

I'm not familiar with DBUS-monitor, so I'll look into it and let you know what I find. Yes, I am totally aware of and I'll quote Gollum, tricksey ways, fMMS does with the data connections. I too attempted to be tricksey and create a duplicate T-Mobile data connection using fAPN, but MHS wasn't fooled.

Originally Posted by

And here it somehow manages to mess up the way we store the old connection so we can restore it (which is interesting since once the connection name is stored the backend no longer enters mainloop() so it should not be able to receive signals... [I'm trying to figure out how to unregister a signal since I would need that to be able to detect when the new connection is finalized reliably without making the single connection state signal processor overly complex)

JayBEE | # 323 | 2010-12-01, 02:02 | Report

Here are some of the interesting events I gleamed from dbus-monitor when making a normal connection to my t-mobile data network:

Code:

signal sender=:1.20 -> dest=(null destination) serial=9242 path=/com/nokia/phone/net; interface=Phone.Net; member=registration_status_change
   byte 0
   uint16 40490
   uint32 85471277
   uint32 260
   uint32 310
   byte 0
   byte 27
signal sender=:1.20 -> dest=(null destination) serial=9243 path=/com/nokia/phone/net; interface=Phone.Net; member=operator_name_change
   byte 10
   string "T-Mobile"
   string ""
   uint32 260
   uint32 310
signal sender=:1.20 -> dest=(null destination) serial=9244 path=/com/nokia/phone/net; interface=Phone.Net; member=radio_info_change
   byte 0
   byte 0
   byte 0
signal sender=:1.87 -> dest=(null destination) serial=546 path=/com/nokia/icd; interface=com.nokia.icd; member=status_changed
   string "0000-0000-0000-0000"
   string "GPRS"
   string "CONNECTED"
   string ""
signal sender=:1.87 -> dest=(null destination) serial=547 path=/com/nokia/icd2; interface=com.nokia.icd2; member=state_sig
   string ""
   uint32 0
   string ""
   string "GPRS"
   uint32 83886080
   array [
      byte 48
      byte 48
      byte 48
      byte 48
      byte 45
      byte 48
      byte 48
      byte 48
      byte 48
      byte 45
      byte 48
      byte 48
      byte 48
      byte 48
      byte 45
      byte 48
      byte 48
      byte 48
      byte 48
      byte 0
   ]
   string ""
   uint32 2
signal sender=:1.40 -> dest=(null destination) serial=307 path=/com/nokia/alarmd; interface=com.nokia.alarmd; member=queue_status_ind
   int32 0
   int32 2147483647
   int32 1291169400
   int32 1291168200

rambo, I've PM'd you a link to the full log.

Hope this helps.

Last edited by JayBEE; 2010-12-01 at 03:12. Reason: added note for rambo

rambo | # 324 | 2010-12-01, 07:26 | Report

Originally Posted by JayBEE View Post

Here are some of the interesting events I gleamed from dbus-monitor when making a normal connection to my t-mobile data network:

"0000-0000-0000-0000" in an "interesting" name for a connection but it is technically a valid one (is this part of your/fMMS' "tricksey ways" ?)

Can you get DBUS log for when you start the backend as well.

scuddy | # 325 | 2010-12-02, 09:48 | Report

Ive tried to get mobile hotspot to work a few times but still to no avail, when I run the backend in terminal it outputs this

/home/user # /usr/sbin/mobilehotspot_backend start
DEBUG: executing uname -r
DEBUG: executing lsmod
DEBUG: executing insmod /lib/modules/2.6.28.10power46/nf_conntrack.ko
DEBUG: executing lsmod
DEBUG: executing insmod /lib/modules/2.6.28.10power46/nf_defrag_ipv4.ko
DEBUG: executing lsmod
DEBUG: executing insmod /lib/modules/2.6.28.10power46/nf_conntrack_ipv4.ko
DEBUG: executing lsmod
DEBUG: executing lsmod
DEBUG: executing lsmod
DEBUG: executing insmod /lib/modules/2.6.28.10power46/nf_nat.ko
DEBUG: executing lsmod
DEBUG: executing insmod /lib/modules/2.6.28.10power46/iptable_nat.ko
DEBUG: executing lsmod
DEBUG: executing insmod /lib/modules/2.6.28.10power46/ipt_MASQUERADE.ko
ERROR:dbus.proxies:Introspect error on :1.85:/com/nokia/icd2: dbus.exceptions.DBusException: org.freedesktop.DBus.Error.NotSupported: Unsupported interface or method
DEBUG: executing ifconfig wlan0 down
DEBUG: executing iwconfig wlan0 mode ad-hoc
DEBUG: executing ifconfig wlan0 up
DEBUG: executing iwconfig wlan0 essid "MobileHotspot"
DEBUG: executing ifconfig wlan0 10.7.160.1 netmask 255.255.255.0 up
DEBUG: executing /usr/sbin/dnsmasq -i wlan0 -a 10.7.160.1 -I lo -z -x /var/run/dnsmasq.wlan0.pid --dhcp-range=10.7.160.10,10.7.160.100,6h --dhcp-option=3,10.7.160.1 --dhcp-option=6,10.7.160.1
DEBUG: executing iptables --flush
DEBUG: executing iptables --flush -t nat
DEBUG: executing iptables -P FORWARD ACCEPT
DEBUG: executing iptables -P INPUT ACCEPT
DEBUG: executing iptables -P OUTPUT ACCEPT
DEBUG: executing iptables --table nat --append POSTROUTING --out-interface gprs0 -j MASQUERADE

any ideas as to why, i installed the kernel as per instructions, I also changed the backend.py file for the one that was on this post. could someone point me in the right direction? i have got fmms installed aswell, is that still a problem?

rambo | # 326 | 2010-12-02, 10:09 | Report

Originally Posted by scuddy View Post

could someone point me in the right direction? i have got fmms installed aswell, is that still a problem?

fMMS (and anything else that mucks about with the connections behind MHS' back is still a problem, I have an idea for fixing that [basically a flag that disables all interactions with IcD, which means you are responsible for making sure the right GPRS connection is up before starting MHS])

However it's not neccessarily fMMS that is the problem in this case, the backend starts fine so you're down to basic network troubleshooting (start with the "can you ping device" -posts earlier in the thread, if those shed no light to the issue you will need wireshark and possibly kismet [and knowhow]).

ifreeman | # 327 | 2010-12-06, 00:19 | Report

Originally Posted by ifreeman View Post

Using 2.6.28.10power37 kernel and with mobilehotspot from the app manager, [...] no dhcp offer is received from connecting clients.

After pr1.3 update and updated kernel, mobilehotspot works as if nothing was ever wrong. DHCP, DNS, and everything. Cool beans!

KPAH | # 328 | 2010-12-06, 09:30 | Report

First. It works.

Would prefer WAP though. Anyway, WEP is better than nothing.

The problem is the device turns VERY hot. The CPU usage is not over 15%, but the device gets so hot I can actually smell formaldehyde (that "new stereo" smell).

I assume the problem is in Wifi output power which is set to too high.
How to lower it?

rambo | # 329 | 2010-12-06, 19:01 | Report

Originally Posted by KPAH View Post

First. It works.

Would prefer WAP though. Anyway, WEP is better than nothing.

WPA: See the first post.

Originally Posted by KPAH View Post

The problem is the device turns VERY hot. The CPU usage is not over 15%, but the device gets so hot I can actually smell formaldehyde (that "new stereo" smell).

Running wifi and cellular at full throttle takes some power (which equals heat), if you're using BlueTooth on top of that (for example stereo headset) then it's even more.

Originally Posted by KPAH View Post

I assume the problem is in Wifi output power which is set to too high.
How to lower it?

Ad-Hoc wifi has no power management, in theory there could be some way to adjust it in the driver but it's a binary blob so we can't know.

KPAH | # 330 | 2010-12-07, 08:27 | Report

Originally Posted by rambo View Post

WPA: See the
Ad-Hoc wifi has no power management, in theory there could be some way to adjust it in the driver but it's a binary blob so we can't know.

What about reversing the connection?
I.e.Ad-Hoc connection is hosted by the PC, rather than N900.
Will it make any difference?