MADWIFI

On Monday 28 January 2008 19:50:20 Kel Modderman wrote:
> On Monday 28 January 2008 15:13:38 Michael Renzmann wrote:
> > Hi.
> >
> > > As far as I know, MadWifi has no release engineer right now.
> >
> > We still have, he just steps aside for someone else who better fits the
> > "job description" :)
> >
> > The only thing I'm waiting for to make the 0.9.4 release is a "go!". I
> > couldn't follow the development process during the past weeks very
> > closely, so I'm not sure whether all desired changes are in the 0.9.4
> > branch or not. Which is one of the reasons why I intend to pass the
> > "release engineer" stick to someone else.
> >
> > So, is 0.9.4 ready to be released? Or is something still missing?
>
> I believe the 0.9.4 branch is ready to be released after communication on
> these lists with bruno and Scott, who were the main contributors to that
> branch.

not quite. there are two few patches pending review and approval, namely 
scotts: "[0.9.4 RFC] STA powersave notification fixes" and my "[0.9.4 RFC] 
allow retries be specified for injected frames". of course both are not too 
important so you could release without them, but it would be nice to include 
them, imho.

bruno

On 29/01/2008, at 1:40 AM, bruno randolf wrote:

> On Monday 28 January 2008 19:50:20 Kel Modderman wrote:
>> On Monday 28 January 2008 15:13:38 Michael Renzmann wrote:
>>> Hi.
>>>
>>>> As far as I know, MadWifi has no release engineer right now.
>>>
>>> We still have, he just steps aside for someone else who better  
>>> fits the
>>> "job description" :)
>>>
>>> The only thing I'm waiting for to make the 0.9.4 release is a  
>>> "go!". I
>>> couldn't follow the development process during the past weeks very
>>> closely, so I'm not sure whether all desired changes are in the  
>>> 0.9.4
>>> branch or not. Which is one of the reasons why I intend to pass the
>>> "release engineer" stick to someone else.
>>>
>>> So, is 0.9.4 ready to be released? Or is something still missing?
>>
>> I believe the 0.9.4 branch is ready to be released after  
>> communication on
>> these lists with bruno and Scott, who were the main contributors to  
>> that
>> branch.
>
> not quite. there are two few patches pending review and approval,  
> namely
> scotts: "[0.9.4 RFC] STA powersave notification fixes" and my  
> "[0.9.4 RFC]
> allow retries be specified for injected frames". of course both are  
> not too
> important so you could release without them, but it would be nice to  
> include
> them, imho.
>

Just ACKed your patch bruno. As far as the powersave fixes, I'd like  
them to go in but I'd feel more comfortable with someone taking a good  
look at the behaviour before and after. Essentially the problem is  
that before the patch, the power management bit is not set when a sta  
goes into powersave mode when it goes off scanning, so the AP  
continues to send frames to the sta while it's off channel. After the  
patch, the bit is correctly set in the Null data frame so the AP  
queues the traffic correctly.

Cheers all,


> bruno
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Madwifi-devel mailing list
> Madwifi-devel@...
> https://lists.sourceforge.net/lists/listinfo/madwifi-devel

--
Scott Raynel
WAND Network Research Group
Department of Computer Science
University of Waikato
New Zealand

Hi All,

On 29/01/2008, at 2:09 PM, Scott Raynel wrote:

>
> On 29/01/2008, at 1:40 AM, bruno randolf wrote:
>
>> On Monday 28 January 2008 19:50:20 Kel Modderman wrote:
>>> On Monday 28 January 2008 15:13:38 Michael Renzmann wrote:
>>>> Hi.
>>>>
>>>>> As far as I know, MadWifi has no release engineer right now.
>>>>
>>>> We still have, he just steps aside for someone else who better
>>>> fits the
>>>> "job description" :)
>>>>
>>>> The only thing I'm waiting for to make the 0.9.4 release is a
>>>> "go!". I
>>>> couldn't follow the development process during the past weeks very
>>>> closely, so I'm not sure whether all desired changes are in the
>>>> 0.9.4
>>>> branch or not. Which is one of the reasons why I intend to pass the
>>>> "release engineer" stick to someone else.
>>>>
>>>> So, is 0.9.4 ready to be released? Or is something still missing?
>>>
>>> I believe the 0.9.4 branch is ready to be released after
>>> communication on
>>> these lists with bruno and Scott, who were the main contributors to
>>> that
>>> branch.
>>
>> not quite. there are two few patches pending review and approval,
>> namely
>> scotts: "[0.9.4 RFC] STA powersave notification fixes" and my
>> "[0.9.4 RFC]
>> allow retries be specified for injected frames". of course both are
>> not too
>> important so you could release without them, but it would be nice to
>> include
>> them, imho.
>>
>
> Just ACKed your patch bruno. As far as the powersave fixes, I'd like
> them to go in but I'd feel more comfortable with someone taking a good
> look at the behaviour before and after. Essentially the problem is
> that before the patch, the power management bit is not set when a sta
> goes into powersave mode when it goes off scanning, so the AP
> continues to send frames to the sta while it's off channel. After the
> patch, the bit is correctly set in the Null data frame so the AP
> queues the traffic correctly.
>

Both those patches have been committed now. I'll do some testing with  
the current revision (r3302) and let you know how it goes.

Cheers,

--
Scott Raynel
WAND Network Research Group
Department of Computer Science
University of Waikato
New Zealand

Hi.

> Both those patches have been committed now. I'll do some testing with
> the current revision (r3302) and let you know how it goes.

Snapshot of v0.9.4 release candidate 1 is available for download here:

http://snapshots.madwifi.org/special/madwifi-0.9.4-rc1.tar.gz

Bye, Mike

Hello,

I've managed to compile & run madwifi 0.9.4 on a home router, however 
the tools in the "tools" directory does not work.

The became linked against glibc while I only use uClibc. I haven't 
managed to make them link against the correct c lib, does anyone have 
any tips?

The reason I need them is because ath0 has changed name to wifi0 and 
apparently you need to run some wlanconfig-command to activate the ath0 
interface again, is this correct?


regards
/Jocke

This email was Anti Virus checked by Astaro Security Gateway.

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

Joakim Wennergren a écrit :
| Hello,
|
| I've managed to compile & run madwifi 0.9.4 on a home router, however
| the tools in the "tools" directory does not work.
|
| The became linked against glibc while I only use uClibc. I haven't
| managed to make them link against the correct c lib, does anyone have
| any tips?
|
| The reason I need them is because ath0 has changed name to wifi0 and
| apparently you need to run some wlanconfig-command to activate the ath0
| interface again, is this correct?
|
|
| regards
| /Jocke

Correct. It properly compiles under OpenWRT which is using uclibc. You
can look at their Makefile to see how it works.

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

iD8DBQFHoOk0OR6EySwP7oIRAkcwAJ9UmbMK4CHbOQn4tAEa3suRRhU9yACfU4Uo
YVYVpHBmPZ0BL78i7Wfd/Ho=
=wNEC
-----END PGP SIGNATURE-----

Il Wednesday 30 January 2008 14:26:47 Joakim Wennergren ha scritto:
> Hello,
>
> I've managed to compile & run madwifi 0.9.4 on a home router, however
> the tools in the "tools" directory does not work.
>
> The became linked against glibc while I only use uClibc. I haven't
> managed to make them link against the correct c lib, does anyone have
> any tips?
>
> The reason I need them is because ath0 has changed name to wifi0 and
> apparently you need to run some wlanconfig-command to activate the ath0
> interface again, is this correct?
>
I have the same your problem... if you manage to work thinks out, please let 
me know.
In the meantime I'm using Openwrt, which compiles and runs without problem 
(but in the future, I plan to try snapgear, to see which is the faster).
Bye

Fair enough, is this already in the other branches? ACK.

On 31/01/2008, at 9:31 AM, Pavel Roskin wrote:

> Section mismatch checker in the forthcoming Linux 2.6.25 complains  
> about
> ath_pci_drv_id and recommends to rename it.  Rename ath_pci_drv_id to
> ath_pci_driver, which is consistent with other PCI drivers.
>
> Index: ath/if_ath_pci.c
> ===================================================================
> --- ath/if_ath_pci.c	(revision 3306)
> +++ ath/if_ath_pci.c	(working copy)
> @@ -325,7 +325,7 @@
>
> MODULE_DEVICE_TABLE(pci, ath_pci_id_table);
>
> -static struct pci_driver ath_pci_drv_id = {
> +static struct pci_driver ath_pci_driver = {
> 	.name		= "ath_pci",
> 	.id_table	= ath_pci_id_table,
> 	.probe		= ath_pci_probe,
> @@ -380,7 +380,7 @@
> {
> 	printk(KERN_INFO "%s: %s\n", dev_info, version);
>
> -	if (pci_register_driver(&ath_pci_drv_id) < 0) {
> +	if (pci_register_driver(&ath_pci_driver) < 0) {
> 		printk("ath_pci: No devices found, driver not installed.\n");
> 		return (-ENODEV);
> 	}
> @@ -393,7 +393,7 @@
> exit_ath_pci(void)
> {
> 	ath_sysctl_unregister();
> -	pci_unregister_driver(&ath_pci_drv_id);
> +	pci_unregister_driver(&ath_pci_driver);
>
> 	printk(KERN_INFO "%s: driver unloaded\n", dev_info);
> }
>
>
> -- 
> Regards,
> Pavel Roskin
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Madwifi-devel mailing list
> Madwifi-devel@...
> https://lists.sourceforge.net/lists/listinfo/madwifi-devel

--
Scott Raynel
WAND Network Research Group
Department of Computer Science
University of Waikato
New Zealand

On Thu, 2008-01-31 at 09:49 +1300, Scott Raynel wrote:
> Fair enough, is this already in the other branches? ACK.

I'll apply it to the trunk as well.  Thanks!

-- 
Regards,
Pavel Roskin

Don't see any problem with this going in. ACK.

On 31/01/2008, at 9:21 AM, Pavel Roskin wrote:

> Use zero-length arrays instead.  Flexible arrays are ISO C99, and
> zero-length arrays is an older gcc extension.  But apparently we still
> have users with gcc-2.95, and it's the only issue preventing MadWifi
> compilation, so it would be better to fix it for now.  We could switch
> back to C99 flexible arrays once Linux 2.4 support is dropped.
>
> Ref: r2256
> ---
> This has been for many months in the trunk, and needs to be in 0.9.4  
> and
> possibly in 0.9.3.3.
>
> Index: net80211/ieee80211_ioctl.h
> ===================================================================
> --- net80211/ieee80211_ioctl.h	(revision 3306)
> +++ net80211/ieee80211_ioctl.h	(working copy)
> @@ -634,7 +634,7 @@
> struct ieee80211req_getset_appiebuf {
> 	u_int32_t	app_frmtype;		/* management frame type for which buffer  
> is added */
> 	u_int32_t	app_buflen;		/* application-supplied buffer length */
> -	u_int8_t	app_buf[];		/* application-supplied IE(s) */
> +	u_int8_t	app_buf[0];		/* application-supplied IE(s) */
> };
>
> /* Flags ORed by application to set filter for receiving management  
> frames */
> Index: net80211/ieee80211.h
> ===================================================================
> --- net80211/ieee80211.h	(revision 3306)
> +++ net80211/ieee80211.h	(working copy)
> @@ -358,7 +358,7 @@
> struct ieee80211_ie {
> 	u_int8_t id;
> 	u_int8_t len;
> -	u_int8_t info[];
> +	u_int8_t info[0];
> } __packed;
>
> /*
>
>
>
> -- 
> Regards,
> Pavel Roskin
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Madwifi-devel mailing list
> Madwifi-devel@...
> https://lists.sourceforge.net/lists/listinfo/madwifi-devel

--
Scott Raynel
WAND Network Research Group
Department of Computer Science
University of Waikato
New Zealand

hi,
> I 
would 
like 
to 
implement 
a 
rate 
adaptation 
algorithm 
like 
SampleRate  
> and 
I 
want 
to 
enable 
RTS/CTS 
in 
a 
per 
packet 
basis 
(e.g 
in 
the 
first  
> 10 
packets 
I 
want 
RTS/CTS 
disabled 
and 
in 
the 
11th 
packet 
I 
want  
> RTS/CTS 
enabled). 
How 
can 
I 
do 
that?


Well, I had asked very similar query about 8 months back, please search through
the archives for developers list and you can get the solution.

Cheers,
Lochan.




      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ