Re: [Madwifi-devel] Enable WEP Software Encryption

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 454-5900

>I think it's pretty easy. The driver checks if the HAL/hardware is
>capable of doing WEP hardware encryption or not. Just locate this
>function in order to return that the hardware is not capable of doing
>WEP. This will enable software WEP.

You are right:

I've tried to change the capabilities flags around the code in multiple places, but it didn't seem to have effect.
The modification should be done in ieee80211.c , ieee80211_vap_setup(). One can do any changes they like to vap->iv_caps successfully and these will stay permanent. 

Best,
Yara

________________________________
From: "mad...@li..." <mad...@li...>
To: mad...@li...
Sent: Monday, November 24, 2008 4:46:20 AM
Subject: Madwifi-devel Digest, Vol 30, Issue 11

Send Madwifi-devel mailing list submissions to
    mad...@li...

To subscribe or unsubscribe via the World Wide Web, visit
    https://lists.sourceforge.net/lists/listinfo/madwifi-devel
or, via email, send a message with subject or body 'help' to
    mad...@li...

You can reach the person managing the list at
    mad...@li...

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Madwifi-devel digest..."

Today's Topics:

   1. compiling help (Justin Hadella)
   2. Re: compiling help (Pavel Roskin)
   3. Re: Enable WEP Software Encryption (Benoit PAPILLAULT)
   4. Re: About transmission feedback (Benoit PAPILLAULT)
   5. R: Re:  About transmission feedback (Roberto Riggio)
   6. Kernel panic after 6 days of uptime (Andrej Podzimek)

----------------------------------------------------------------------

Message: 1
Date: Wed, 19 Nov 2008 21:34:30 -0700
From: "Justin Hadella" <pit...@gm...>
Subject: [Madwifi-devel] compiling help
To: mad...@li...
Message-ID:
    <d14...@ma...>
Content-Type: text/plain; charset="iso-8859-1"

I think I have finally figured out how to do my noise measurements I talked
about on the dev list some time ago. However, now I am having trouble
compiling my program. Could someone help me come up a command to do this or
help out with a makefile?

Here is an example of what I would like to see (as far as usage goes) plus
the relevant code:

$show-noise ath0
Channel: 1     2     3    4   ...
             -95  -94  -80  -96 ...

/* show-noise.c - written by Justin Hadella 11/18/2008 */

#include <stdio.h>
#include <string.h>

#include <linux/netdevice.h>    /* struct net_device, dev_get_by_name()
        */
#include "../ath/if_athvar.h"    /* struct ath_softc,
ath_hal_get_channel_noise()    */
#include "../hal/ah.h"        /* struct ath_hal, struct HAL_CHANNEL
    */

/* Note: device "ath0" uses channels 1-11    (2.4 GHz radio)        */
/*       device "ath1" uses channels 3-6    (900 MHz radio)        */

int main(int argc, char **argv)
{
    struct net_device *dev;
    struct ath_softc *sc;
    struct ath_hal *ah;
    struct HAL_CHANNEL channel;
    int i;
    int noise[11];
    /*  stock data for testing the print capability */
    /*
    int noise[11] = { -96, -95, -96, -84, -95, -96, -95, -95, -94, -92, -95
};
    */

    /* initialize device */
    /*

    /* get pointer to device from device name (ie "ath0") */
    dev = dev_get_by_name(argv[1]);

    /* code like this used throughout madwifi to get the ah structure */
    sc = dev->priv;
    ah = sc->sc_ah;

    i = ath_hal_get_channel_noise(ah, &channel);
    printf("%d\n", i);

    /* get noise values for each channel */
    for (i = 0; i <= 10; i++)
    {
        channel.channel = i + 1;
        noise[i] = ath_hal_get_channel_noise(ah, &channel);
    }

    /* print out what was measured */
    printf("Channel: ");
    for (i = 0; i <= 10; i++)
    {
        printf("%5d ", i + 1);
    }
    printf("\n  Noise: ");

    for (i = 0; i <= 10; i++)
    {
        printf("%5d ", noise[i]);
    }
    printf("dBm\n");

    return 0;
}

I am having trouble getting this to compile. The errors I am seeing have to
do with various things in if_athvar.h and ah.h files. However, I have no
trouble compiling madwifi itself. So I don't know what to do.
I do not know if I have to go the whole device driver type route, or if this
simple program would do.
-------------- next part --------------
An HTML attachment was scrubbed...

------------------------------

Message: 2
Date: Thu, 20 Nov 2008 00:30:03 -0500
From: Pavel Roskin <pr...@gn...>
Subject: Re: [Madwifi-devel] compiling help
To: Justin Hadella <pit...@gm...>
Cc: mad...@li...
Message-ID:
    <200...@we...>
Content-Type: text/plain;    charset=ISO-8859-1;    DelSp="Yes";
    format="flowed"

Quoting Justin Hadella <pit...@gm...>:

> I think I have finally figured out how to do my noise measurements I talked
> about on the dev list some time ago. However, now I am having trouble
> compiling my program. Could someone help me come up a command to do this or
> help out with a makefile?
>
> Here is an example of what I would like to see (as far as usage goes) plus
> the relevant code:
>
> $show-noise ath0
> Channel: 1     2     3    4   ...
>              -95  -94  -80  -96 ...
>
> /* show-noise.c - written by Justin Hadella 11/18/2008 */
>
> #include <stdio.h>
> #include <string.h>
>
> #include <linux/netdevice.h>    /* struct net_device, dev_get_by_name()

You are trying to mix kernel and userspace includes.  If you are  
compiling a module, don't use libc header.  If you are compiling a  
userspace program, avoid kernel includes and don't use any kernel  
structures.

> int main(int argc, char **argv)
> {
>     struct net_device *dev;

You are using the main() entry point used by userspace software and a  
kernel structure net_device.

> I am having trouble getting this to compile. The errors I am seeing have to
> do with various things in if_athvar.h and ah.h files. However, I have no
> trouble compiling madwifi itself. So I don't know what to do.
> I do not know if I have to go the whole device driver type route, or if this
> simple program would do.

Every source file in madwifi compiles either into a module or into a  
userspace program.  You program has features of both, so it cannot  
compile in either way.

-- 
Regards,
Pavel Roskin

------------------------------

Message: 3
Date: Fri, 21 Nov 2008 21:25:39 +0100
From: Benoit PAPILLAULT <ben...@fr...>
Subject: Re: [Madwifi-devel] Enable WEP Software Encryption
To: y omar <y_o...@ya...>
Cc: madwifi devel <mad...@li...>
Message-ID: <492...@fr...>
Content-Type: text/plain; charset=ISO-8859-1

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

y omar a ?crit :
> Hello,
> 
> I'm working on a project that intends to modify the code of
> wep_encrypt() in ieee80211_crypto_wep.c. For right now, madwifi does all
> the encryption via hardware, and never calls this function. Anybody
> knows how to disable the hardware encryption?
> 
> I'm also a newbie, I'm having some difficulties tracing back function
> calls. If anybody has a good suggestion to a debugging tool that enables
> me to trace function calls in the 80211 module would be great!
> 
> Thank you,
> Yara

I think it's pretty easy. The driver checks if the HAL/hardware is
capable of doing WEP hardware encryption or not. Just locate this
function in order to return that the hardware is not capable of doing
WEP. This will enable software WEP.

Regards,
Benoit
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFJJxlDOR6EySwP7oIRAjWvAKDwsiUcCAz2pgQxDhGGA0B8WawvxgCeO7YQ
+ir8s0khRHymahM4oq37zCI=
=x+xb
-----END PGP SIGNATURE-----

------------------------------

Message: 4
Date: Fri, 21 Nov 2008 21:30:54 +0100
From: Benoit PAPILLAULT <ben...@fr...>
Subject: Re: [Madwifi-devel] About transmission feedback
To: rob...@cr...
Cc: mad...@li...
Message-ID: <492...@fr...>
Content-Type: text/plain; charset=ISO-8859-1

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

rob...@cr... a ?crit :
> Hi,
> 
> in the old madwifi.stripped developed by J. Bicket, packets that are sent 
> actually come back from the interface the same way received packets come
> back. 
> 
> However the current madwifi does not have such a feature anymore. I'm using 
> click to generate some traffic but transmitted packet are not sent back.
> 
> Was the feature removed?
> 
> 

Are you talking about monitor VAP features? Please details your need.

Regards,
Benoit
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFJJxp+OR6EySwP7oIRAlunAJ0ex9EQR8d5MntGflE6F7nYuHsy5wCfTd4E
LA3fcKXEe9hGBp2vnBOApvg=
=IHbP
-----END PGP SIGNATURE-----

------------------------------

Message: 5
Date: Sat, 22 Nov 2008 18:00:07 +0100 (CET)
From: Roberto Riggio <rob...@cr...>
Subject: [Madwifi-devel] R: Re:  About transmission feedback
To: Benoit PAPILLAULT <ben...@fr...>
Cc: mad...@li...
Message-ID: <12149913.461051227373207212.JavaMail.root@aurora>
Content-Type: text/plain; charset=utf-8

----- "Benoit PAPILLAULT" <ben...@fr...> ha scritto:
> Are you talking about monitor VAP features? Please details your need.

I'm aware of the monitor mode. However as far as I know madwifi once had 
this transmission feedback feature. This was a feature of the monitor mode
where packets just transmitted were sent back to the interface the same
way received packets come back.

Feedback packets had the madwifi descriptors filled with the information 
of the packet just transmitted (tx attemp, tx rate, rssi) which are very 
useful to evaluate the channel (in wireless mesh networks for example)

> Regards,
> Benoit

Bye
R.

------------------------------

Message: 6
Date: Mon, 24 Nov 2008 03:06:03 +0100
From: Andrej Podzimek <an...@po...>
Subject: [Madwifi-devel] Kernel panic after 6 days of uptime
To: mad...@li...
Message-ID: <492...@po...>
Content-Type: text/plain; charset=UTF-8; format=flowed

Hello,

I use the current revision of MadWifi with hostapd. The driver causes a kernel panic regularly. This occurs both with trunk and with branches/madwifi-hal-testing. The stack trace shows lots of stuff around the ath0 interface and madwifi driver functions. I have seen the panic more than four times already, each time after about six days of uptime.

A panic is still a happy-end, as it causes the machine to reboot. In some cases, however, memory corruption leads to an inconsistent state with many zombie processes that cannot be collected. Disk and network access is either limited or blocked. The machine cannot be rebooted (gracefully), as any process that deals with disk access or logging simply dies and hangs forever. This has never happened before I installed the WiFi card. That's why I think the driver could be causing this. The machine has ECC memory chips, so I don't think this could be a physical memory failure.

That said, the current revision of MadWifi is useless on my hardware. The last revision that seemed to work was 3866. (Well, in fact it reached some 30 days of uptime, which could be just a piece of luck...)

Is this a known issue? Is there a workaround? Is there a branch/revision worth trying? Will it help if I reload the module once a day? I got a kernel panic after the third reload last time I tried this, but many things could have changed sice then.

Just any piece of advice would be very helpful.

Regards,

Andrej Podzimek

------------------------------

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/

------------------------------

_______________________________________________
Madwifi-devel mailing list
Mad...@li...
https://lists.sourceforge.net/lists/listinfo/madwifi-devel

End of Madwifi-devel Digest, Vol 30, Issue 11
*********************************************