OESF Portables Forum
Everything Else => Zaurus - Everything Development => Distros, Development, and Model Specific Forums => Archived Forums => pdaXrom Development => Topic started by: Allelectrix on August 23, 2008, 09:59:27 pm
-
Its incredible.. cant get kernel modules to download, dont know how to modify vanilla kernel code for Zaurus C1000 because no one
else knows it seems ? I hope I am wrong ?
And all I want is:
rt73.o
yes, that single kernel module compiled for kernel 2.4.20 and to be used on a Zaurus SL-C1000 and pdaxii13 OS.
just make sure the source code is from the ASPj rt73-k2wrlz source code and you use GCC 3.4.5 which is the default GCC for
pdaxii13.
Its got to be the all time easiest $150 US your ever going to make.
someone, somewhere out there knows exactly what I'm talking about and want, if thats you, come get your money ! its ready !
create for me rt73.o and its all yours !
contact me on +44 775 7627096 on my Cellular phone daylight GMT . if no reply, please leave a message I will get back to you
quickly.
please also contact me at: allelectrix2@yahoo.co.uk
your efforts to help me will be well rewarded. thanks so much !
-
as of Monday September 29th 2008, this still applies: $150 US to anyone who can supply me the kernel 2.4.20 module in ARM machine language:
rt73.o
this will need to run on a Sharp SL-C1000 so the module needs to be compiled for XScale ARM.
it will usually be resident in:
/lib/modules/2.4.20/kernel/net/wireless/extra/
but you can execute in a bash shell:
find / -name rt73.o
the module should be compiled from ASPj's source code on his web site.
if you have it, I would greatly appreciate you contacting me at my email address and attaching it to a mail. if you send it to me, I
will host it on my website so others can benefit from this.
I have found out since the last post, above, you need to use GCC 2.95.3 which is the same version the kernel is compiled with.
COME ON! someone, somewhere with a Z has got that module ! I WANT IT! its got to be the easiest $150 US your ever going to make man!
I'm getting closer to solving this riddle myself, but its still a way off from me yet, I know I am getting closer. however, for now still, its
up to everyone else to help. please help !
-
Looking on Google for "rt73.o zaurus" or "rt73.o arm" gives interesting results in Chinese and Korean. I don't understand these languages but the web pages contain instructions to compile the kernel and the rt73 module. Given that your environment is probably already set up to compile it, maybe you could find some interesting clues there.
I remember my previous experiences when building my own kernel for the "fun" of it, and well, to tell the truth, it was not fun at all... Anyway, good luck for your quest! :-)
-
Looking on Google for "rt73.o zaurus" or "rt73.o arm" gives interesting results in Chinese and Korean. I don't understand these languages but the web pages contain instructions to compile the kernel and the rt73 module. Given that your environment is probably already set up to compile it, maybe you could find some interesting clues there.
I remember my previous experiences when building my own kernel for the "fun" of it, and well, to tell the truth, it was not fun at all... Anyway, good luck for your quest! :-)
Thank you Frederic so much for your reply, alas, I already know about the requirement for the existence of the directories mentioned there. happily, I managed
to cross compile a successful build of zImage.bin that the Z could actually run, and did run.
So, I am 50% there now. I need to get to know how to get my x86 to cross compile kernel modules for ARM namely, kernel 2.4.20, so I am still seeking
the illusive:
rt73.o
so, do you know anything else please ? anything would be greatly appreciated, all the very best.
-
Two thoughts. I know how frustrating what you are going through is, after a friend compiled serial modules for me for the Collie. The compilation was a bear. Here is my page about the adventure with 2.4.6 Kernel source:
http://sdjf.esmartdesign.com/kernel.html (http://sdjf.esmartdesign.com/kernel.html)
So yeah, building is hard, especially if there are errors in the kernel source itself, or you need a patch to make it work on your ROM, which is what he found for the 2.4.6 kernel.
I suggest using google to help you find someone who knows how to do what you want but who isn't reading the forum. Or check out the Zaurus channels at irc.freenode.net for help. That's how we located the patch we needed, I emailed developers of applications for my ROM (who I found using google or dogpile), and one kindly helped guide us to the necessary files and sent us the one that was not on the web.
Even #linux might have someone who could look at your output and help figure out the problem, if you don't find someone able to rescue you.
Developers are usually off doing other things. But here is the website of one who compiled the module for 2.4.23 for some other machine, and maybe he can give you guidance. Nothing to lose in sending him an email.
http://homepages.tu-darmstadt.de/%7Ep_larbig/wlan/ (http://homepages.tu-darmstadt.de/%7Ep_larbig/wlan/)
And look at results of this search:
http://www.google.com/search?hl=en&ie=...G=Google+Search (http://www.google.com/search?hl=en&ie=ISO-8859-1&q=rt73.o+arm+compiled+OR+module&btnG=Google+Search)
Add 2.4.20 to above search to narrow down results for your kernel.
Lots of guidance next url, and someone offering compilation help for a fee, but again, not on a zaurus:
http://rt2x00.serialmonkey.com/phpBB/viewt...?f=8&t=4925 (http://rt2x00.serialmonkey.com/phpBB/viewtopic.php?f=8&t=4925)
And a number of sites mention needing to add an alias rausb0=rt73
I have never compiled anything so this is the best I can do to help.
sdjf
-
Two thoughts. I know how frustrating what you are going through is, after a friend compiled serial modules for me for the Collie. The compilation was a bear. Here is my page about the adventure with 2.4.6 Kernel source:
http://sdjf.esmartdesign.com/kernel.html (http://sdjf.esmartdesign.com/kernel.html)
So yeah, building is hard, especially if there are errors in the kernel source itself, or you need a patch to make it work on your ROM, which is what he found for the 2.4.6 kernel.
I suggest using google to help you find someone who knows how to do what you want but who isn't reading the forum. Or check out the Zaurus channels at irc.freenode.net for help. That's how we located the patch we needed, I emailed developers of applications for my ROM (who I found using google or dogpile), and one kindly helped guide us to the necessary files and sent us the one that was not on the web.
Even #linux might have someone who could look at your output and help figure out the problem, if you don't find someone able to rescue you.
Developers are usually off doing other things. But here is the website of one who compiled the module for 2.4.23 for some other machine, and maybe he can give you guidance. Nothing to lose in sending him an email.
http://homepages.tu-darmstadt.de/%7Ep_larbig/wlan/ (http://homepages.tu-darmstadt.de/%7Ep_larbig/wlan/)
And look at results of this search:
http://www.google.com/search?hl=en&ie=...G=Google+Search (http://www.google.com/search?hl=en&ie=ISO-8859-1&q=rt73.o+arm+compiled+OR+module&btnG=Google+Search)
Add 2.4.20 to above search to narrow down results for your kernel.
Lots of guidance next url, and someone offering compilation help for a fee, but again, not on a zaurus:
http://rt2x00.serialmonkey.com/phpBB/viewt...?f=8&t=4925 (http://rt2x00.serialmonkey.com/phpBB/viewtopic.php?f=8&t=4925)
And a number of sites mention needing to add an alias rausb0=rt73
I have never compiled anything so this is the best I can do to help.
sdjf
-
thanks for your help guys, keep this stuff coming in. I do find it very hard to get compiling done myself, so if anyone else can
offer to do it, I will pay them a great deal of money to get it done.
The links so far suggested are good, and I will be following them up and give them a good try. I will return results from these back here.
thanks again, this thread is not by any means closed, I need a Zaurus specialist to do this, is this you ?? if so, I need your services please.
-
Just out of curiosity, why don't you just buy another wireless adapter that is already supported by your distro? Maybe it would be easier and cheaper in terms of sweat, time, and money...
-
I've only got Hostap for 802.11b CF Cards and the kernel Module:
rt2570.o
this is for 2 Cards:
a) D-Link DWL-G122 B1
Linksys WUSB54G
Which both turn out to be a load of crap. They run for a few times, then mysteriously just die, and
become about as much use as a worn out shoe. While I am investigating this great phenomena, the real
powerhouse card:
D-Link DWL-G122 C1
is just waiting for me to slap it into its USB slot and let it do its great stuff, alas, my Computer wont know
what to do with it, as it does not (as yet) have the kernel module:
rt73.o
Alas, your thought is good, but for the application I want, there is only one choice available:
Any rt73 chipset based USB wifi card
so, role up, role up, all you Unixheads, payment to whatever level I pledge to the first person to compile
for me:
rt73.o in ARM Machine Code
your payment awaits.
in the meantime, keep your ideas rolling in, I want to hear all.
-
Out of curiousity, what does "works for a few times and then just dies" mean?
May be a stupid question, from someone who knows nearly zilch about hostap and wireless, but it sounds suspiciously to me like there may just be some wrong settings somewhere in /etc/. I found Opera was resetting my config and ppp file settings to the wrong baud rate if I let it.
May be totally unrelated, you seem sure it's the driver you need, but what the heck. I just got my post up about that a few days ago (again, sl6000, Sharp ROM):
http://sdjf.esmartdesign.com/blog/settings.html (http://sdjf.esmartdesign.com/blog/settings.html)
Maybe the new card would do the same thing, because some application is messing up your settings?
-
Out of curiousity, what does "works for a few times and then just dies" mean?
May be a stupid question, from someone who knows nearly zilch about hostap and wireless, but it sounds suspiciously to me like there may just be some wrong settings somewhere in /etc/. I found Opera was resetting my config and ppp file settings to the wrong baud rate if I let it.
May be totally unrelated, you seem sure it's the driver you need, but what the heck. I just got my post up about that a few days ago (again, sl6000, Sharp ROM):
http://sdjf.esmartdesign.com/blog/settings.html (http://sdjf.esmartdesign.com/blog/settings.html)
Maybe the new card would do the same thing, because some application is messing up your settings?
Its a good thought sdjf, but I have confirmed that it is a Hardware problem with these 2 Cards.
The Cards main IC has no heatsink on it, and this is whats making these cards perform in the way I mentioned.
I put my finger on the top of the main IC on the card while connected to the USB system, as soon as I issued the
wireless control commands:
ifconfig rausb0 up
iwconfig rausb0 mode monitor channel 2
the main IC on the card heats up to such a degree that I can't keep my finger on the top of the IC anymore.
So, I'm hoping that by adding some refrigeration to the IC that the card continues to run indefinitely without
burning itself out.
Thats a real interesting problem you had there with Opera and your CF Modem ! nice that you sorted
it out !
If you know anyone that has a kernel 2.4.20 toolchain up and running and can make .o kernel modules
please let me know everyone.
In the meantime, please keep this thread running, I want to hear from all of you, thanks to all
so far that have tried to assist mw with this problem ;-)
-
OK, I'm getting Cacko dev stuff downloaded. We'll see how this turns out
-
Try this (It's likely broken in some way or another)
Remove the .txt and add .o
-
Try this (It's likely broken in some way or another)
Remove the .txt and add .o
Darn!... you were right, there are faults, but a REALLY nice try !
We're getting so warm with this, I can now feel the heat on my Face !
I'm getting the Segmentation Fault now when I try to bring the Interface up.
We nearly had it !
this is what I have tested it with:
SL-C1000 Computer
pdaXii13
D-Link DWL G122 C1
USB 2.0 Hub = dont work at all
USB 1.0 Hub = Segmentation Fault
Direct connection to Computer = Segmentation Fault
I presume that if the module will run the D-Link DWL G122 C1 it will run all the other rt73 Cards ?
Here's the system output:
[ You are not allowed to view attachments ]
Isn't it normally rausb0 ?
would it be possible to see if you could compile this code to ARM rt73.o ?:
[ You are not allowed to view attachments ]
1 Million Thank You's from me for trying this, your the Man!
I REALLY look forward to your next post here !
-
I tried 5 versions of the code (the one you had, a newer version, a CVS, and some others), but the one that built had the fewest errors.
The others gave weird "couldn't parse something before this," where the code looked fine to me.
I think I know what is causing the segfault, too...I had to change a line of code that I didn't really know what it did.
If somebody's familiar with kernel stuff, I'd appreciate you looking at this line:
struct usb_device *dev = pAd->pUsb_Dev;
...
//struct device *udevice = &dev->dev;
struct device *udevice = dev;
The third, commented out one is the original. The compiler claims that there's no such member "dev" in "&dev," so I hacked it together so it compiled
Basically, if somebody could tell me what to change that to, I should have a working module. I believe that usb_device is defined in the kernel header include/linux/usb.h. I haven't worked with structs, so this all seems like a jumbel/excruciatingly messy coding to me (which it likely is).
-
struct usb_device *dev = pAd->pUsb_Dev;
...
//struct device *udevice = &dev->dev;
struct device *udevice = dev;
The third, commented out one is the original. The compiler claims that there's no such member "dev" in "&dev," so I hacked it together so it compiled
The code is trying to assign udevice as a pointer to the dev member of dev struct and the compiler is complaining there is no member dev in dev struct so you need to find the definition of dev struct and find why there is no dev member.
The change you made assigned udevice as a pointer to whole dev struct which will overflow as soon as the code tries to use that pointer, if you find and post the definition of dev struct may be we can give you a hand (or maybe no )
As a stupid example the code is pointing to a book on a shelf and the compiler can not find that book while your edition in code pointed to the bookshelf itself not a single book and once compiled it brokes its back trying to pick the bookshelf .
-
struct usb_device *dev = pAd->pUsb_Dev;
...
//struct device *udevice = &dev->dev;
struct device *udevice = dev;
The third, commented out one is the original. The compiler claims that there's no such member "dev" in "&dev," so I hacked it together so it compiled
The code is trying to assign udevice as a pointer to the dev member of dev struct and the compiler is complaining there is no member dev in dev struct so you need to find the definition of dev struct and find why there is no dev member.
The change you made assigned udevice as a pointer to whole dev struct which will overflow as soon as the code tries to use that pointer, if you find and post the definition of dev struct may be we can give you a hand (or maybe no )
As a stupid example the code is pointing to a book on a shelf and the compiler can not find that book while your edition in code pointed to the bookshelf itself not a single book and once compiled it brokes its back trying to pick the bookshelf .
Oh.
Talk about doing a Rubiks Cube ! This riddle makes Rubiks Cube look like play dough. Some people in here are going to get seriously rewarded for
solving it ! thats my personal promise !
Keep it coming. we want to know how to resolve this. Gracious thanks to everyone that's helped so far.
-
struct usb_device *dev = pAd->pUsb_Dev;
...
//struct device *udevice = &dev->dev;
struct device *udevice = dev;
The third, commented out one is the original. The compiler claims that there's no such member "dev" in "&dev," so I hacked it together so it compiled
The code is trying to assign udevice as a pointer to the dev member of dev struct and the compiler is complaining there is no member dev in dev struct so you need to find the definition of dev struct and find why there is no dev member.
The change you made assigned udevice as a pointer to whole dev struct which will overflow as soon as the code tries to use that pointer, if you find and post the definition of dev struct may be we can give you a hand (or maybe no )
As a stupid example the code is pointing to a book on a shelf and the compiler can not find that book while your edition in code pointed to the bookshelf itself not a single book and once compiled it brokes its back trying to pick the bookshelf .
I guess I knew that much already, but when I looked up the usd_device struct, I didn't find any lines with members in it. While I'm OK at coding, all of these structs are pushing my limits.
I found a line
struct usb_device;
in the kernel headers, but I don't know where it would be defined later, or what the best way to find the definition would be.
EDIT: Does the module need to have an actual kernel to link against? That may account for it claiming there's no such member...
-
I had similar problems building zd1211 for 2.6.16.
The problem in my case was that the newer versions of the driver were for newer version so kernel (2.6.20+).
I solved my trial and error while going back with the driver releases till I found one that built for 2.6.16.
This may not be applicable in this case but it may be worth while having a go (but don't count on me as I gave up on the Z platform).
Another thing to keep in mind ...
Cx000 devices have USB OTG output that cannot drive the usb wireless card directly (that will require the full 500mA or something close to that ... that is also related to the reception quality).
You will need an externally powered hub to get anywhere. Alternatively you can try a hardware mod to the usb power circuitary (like I did) but the latter will can have trouble with some kernels that are aware of the OTG capabilities of the Z hardware (I had trouble while testing 2.6.2x kernels).
-
Come on everybody ! Please help Capn_Fish solve this problem !!
Capn_Fish, I want you to know that you have my respect and brotherly love for undertaking this for me, thank you so very much again !
-
I compiled the module against my kernel branch (see sig). No changes required in the USB struct. Although I had to remove one union from a struct since the compiler was complaining.
Anyway, the compiled module + modified sourcecode is attached. Let me know if that works. If it does I'll try to integrate the module into my branch.
-
I compiled the module against my kernel branch (see sig). No changes required in the USB struct. Although I had to remove one union from a struct since the compiler was complaining.
Anyway, the compiled module + modified sourcecode is attached. Let me know if that works. If it does I'll try to integrate the module into my branch.
eviLjazz,
Guess what ?
YOU DID IT !! YOU DID IT !! YOU DID IT !! YOU DID IT !! YOU DID IT !! YOU DID IT !! YOU DID IT !! YOU DID IT !! YOU DID IT !! YOU DID IT !! YOU DID IT !!
You and Capn_Fish have made not only my day, but I think my whole year too!!
WOW! your both amazing !
But, you eviLjazz, your our inspiration ! your my inspiration !
I am absolutely blown away by it all ! WOW! WOW! WOW!
I really didn't think that there was anyone out there that was going to fulfill my begging pleas ! but you have fulfilled all my Z
dreams and more !
I have waited SO very long for this moment... Hey ! this is beginning to sound like a love rant !! LOL!
But thats how happy you have made me feel today ! I am completely ecstatic !
when I saw the interface come up:
RT73 WLAN
Seriously, my heart was beating so hard, it nearly jumped out of my chest ! LOL!
I really don't think I have ever been so excited in all my life !
All Card functions are now fully enabled thanks to your kindness, and your awesome minds.
So, what now ?
Now that I am a little calmer ... After the excitement has abated somewhat..
I have a pot of 200 Pounds UK available for this special occasion, please don't tell me your not going to accept as I
I WILL NOT accept that !
I would like to share this between yourself and Capn_Fish, could you both chat and agree to how much you would
both like then let me know ? note that if one of you subsists from receiving the split, the full amount of 200 UK Pounds
will go to the other one of you! so, its up to you both to decide who gets what. I just want this to be as fare as possible.
Yes, I really want to do this, as I feel like rolling out the Champagne today, and this is the only way I can show
you both how elated I feel, so, please accept my gift to you both
I require 2 paypal accounts, thats 2 EMail addresses to deposit the payments.
All that remains for me to do now is to thank the following persons for their kind assistance in arriving at this point:
For giving directions on encouraging me to try to do this myself:
Krasimir Kazakov
David Louigi
And my special thanks to :
Capn_Fish
Who started everybody on this game and gave it a good go, your an inspiration to us Capn_Fish
And finally, to the man of the moment:
eviLjazz
you get the gold cup and the big big deep thank you from the bottom of my heart !
FINALLY
---------
eviLjazz, can you please tell us all the following:
1) What is your compiler ? and can you post it here for us all to download ? are there any special library's or add on's that have to be added to the
compiler to get this to work ? if so, can you allow us to download them too ?
2) Can you package up the kernel source code you used to compile the rt73-k2wrlz against, then allow us to download it ?
I ask the questions because I and I am sure, we, would all like the ability to help someone else with a similar enquiry in the future. Certainly,I
would love the opportunity myself if someone requested my services in this manner as I have of you both.
I have eviLjazz's modified rt73-k2wrlz-2.0.1 source code if anyone wishes to compare it to the original, if thats OK with you eviLjazz
to release this modification for others to see ?
And once again, peace, love and most importantly, good health be with all of you who have helped me in any way with this, Bye for now
and have a lot of fun with your beautiful Zaurus Computers !
-
The Beautiful Sight
[ You are not allowed to view attachments ]
-
Beautiful indeed! You could almost hang the picture in an arts museum :-)
I enjoyed "watching" this nice community effort.
Good work!
-
Glad the module is working.
1) What is your compiler ? and can you post it here for us all to download ? are there any special library's or add on's that have to be added to the
compiler to get this to work ? if so, can you allow us to download them too ?
Actually, it is all pretty simple. The compiler I used is the default compiler used to compile the Cacko and pdaXrom kernels. Since I did not compile the build environment myself, I can't state on what special patches or add ons have been added.
My mirror copy is located here:
http://katastrophos.net/zaurus/kernels/compiler/ (http://katastrophos.net/zaurus/kernels/compiler/)
2) Can you package up the kernel source code you used to compile the rt73-k2wrlz against, then allow us to download it ?
I ask the questions because I and I am sure, we, would all like the ability to help someone else with a similar enquiry in the future. Certainly,I
would love the opportunity myself if someone requested my services in this manner as I have of you both.
The kernel source code is available via my site:
http://katastrophos.net/zaurus/kernels/ (http://katastrophos.net/zaurus/kernels/)
http://katastrophos.net/zaurus/kernels/v55....20-v55.tar.bz2 (http://katastrophos.net/zaurus/kernels/v55/zaurus-kernel-2.4.20-v55.tar.bz2)
You'll find all instructions on how to build a kernel image inside the tar ball in the README file. I think it is pretty well documented.
Actually everything is also available via a Subversion repository. That way it is a lot more comprehensible from the change log what actually has changed compared to the original Sharp source code.
SVN repository: https://svn.katastrophos.net/zaurus-kernel-2.4.20/trunk/ (https://svn.katastrophos.net/zaurus-kernel-2.4.20/trunk/)
README file: https://svn.katastrophos.net/zaurus-kernel-...20/trunk/README (https://svn.katastrophos.net/zaurus-kernel-2.4.20/trunk/README)
To build the module sourcecode, prepare and initialize the kernel sourcecode with either the makeall.sh script as described in the README file or manually be copying the config files to the kernel source directory and then doing make oldconfig && make dep. Also set KERNDIR envvar to whereever the kernel sourcecode is located, like so:
export KERNDIR=/blah/blah/kernel/linux
Finally
make
I have eviLjazz's modified rt73-k2wrlz-2.0.1 source code if anyone wishes to compare it to the original, if thats OK with you eviLjazz
to release this modification for others to see ?
Sure, it's GPL and open source. Like I said, I'll also try to put the sourcecode into my kernel branch perhaps along with some other drivers. Let me know what else you would like to try.
It is nice to see the community work on stuff again even if it was quiet for quite a while.
-
I really don't think I have ever been so excited in all my life !
I won't say the obvious.
Anyway, congratulations to all of you who had, not only the power but also the will!!!
-
Thanks evilJazz for that information, I'll have a go at zd1211b.o myself now with this and let you all know how it goes.
Not that it matters evilJazz, but what was the Linux x86 OS you used ? I guess most x86 Linux Boxes will do ?
A special thanks also to :
Frederic Bergero
Iaan
sdjf
for their contributions to this project.
I was so over the moon the other day I nearly forgot to mention you all too.
In fact, if there is anyone else that helped here I left out, well, thank you too !
You people rock my world !
Now for something a little different.. just continuing on with the rt73 theme, I think it would be really nice if we could get
the rt73 Card to give us an Internet connection in pdaXii13 from using a BASH Shell and a few BASH commands,
So, how do we do this ? has anyone tried it before ? - that is get a Wifi Card other than a CF to give them an Internet connection
in PdaXii13 or PdaXrom ?
-
I've had an internet connection from a USB-Ethernet card (not Wi-Fi, though). I assume it would be a similar sort of process to get yours running:
ifconfig INTERFACE up
iwconfig INTERFACE mode managed essid YOURSSID key YOURKEY
dhclient INTERFACE
-
I've had an internet connection from a USB-Ethernet card (not Wi-Fi, though). I assume it would be a similar sort of process to get yours running:
ifconfig INTERFACE up
iwconfig INTERFACE mode managed essid YOURSSID key YOURKEY
dhclient INTERFACE
Allelectrix replies using the SL-C1000...
Ah yes, I seem to remember issuing several iwconfig commands in BT3 then finally issuing:
dhcpcd -h zaurus -G 192.168.1.254 rausb0
this command fetches the clients new IP Addr from the AP where the IP address = the address of the AP on the network.
Thanks for the reminder Capn_Fish :-)
getting back to the rt73 module, it seems to lack a lot of sensitivity on all AP's .
I find that using my D-Link DWL G122 C1 while not at home is very difficult to get a connection, at the same time my DCF-650W sees networks with ease, and easily picks them up.
I will see if we can get an update to rt73-k2wrlz . The source code that I gave evilJazz was not the most update. I only selected it because I thought it was the last version of this code available for kernel 2.4 .
I guess were not out of the woods yet !
It appears there is a fresher version:
http://homepages.tu-darmstadt.de/~p_larbig...z-3.0.0.tar.bz2 (http://homepages.tu-darmstadt.de/~p_larbig/wlan/rt73-k2wrlz-3.0.0.tar.bz2)
lets all have a go at compiling this for our 2.4.20 kernel this time ! I wonder if I can follow your advice evilJazz on setting up an xcompiler ?
now were going to find out !
I take it you can step in again if we don't succeed ?
thanks for eveything !
-
Attached is the latest version 3.0.1 of the rt73 driver module from http://homepages.tu-darmstadt.de/~p_larbig/wlan/ (http://homepages.tu-darmstadt.de/~p_larbig/wlan/).
Enjoy.
-
Attached is the latest version 3.0.1 of the rt73 driver module from http://homepages.tu-darmstadt.de/~p_larbig/wlan/ (http://homepages.tu-darmstadt.de/~p_larbig/wlan/).
Enjoy.
Man, you are on fire !
I can't keep up with you !
I was going to put this in before you posted this version of rt73.o .
We got Rx Sensitivity issues with the other module.
thus we need to do:
add this patch to the final 3.0.1 code. Before adding it, I think you should check the source code in 3.0.1
to see if its been added already. the patch concerns the file :
rtmp_init.c
basically, as far as I can see this statement at line 2038 needs to go from:
pAd->BbpTuning.bEnable = TRUE;
to this:
pAd->BbpTuning.bEnable = FALSE;
and these two commands need to be at their respective line positions in the same file:
pAd->BbpTuning.R17LowerBoundG 0x00 (line 2009)
pAd->BbpTuning.R17UpperBoundG 0x00 (line 2010)
root :-) make
and lets load this puppy up and see what gives ;-)
thanks so much for everything evilJazz, I await this 3rd module now.
In the meantime, I will load up the 2nd module you made of the 3.0.1 code and see if that has resolved this sensitivity issue already.
-
Attached is the latest version 3.0.1 of the rt73 driver module from http://homepages.tu-darmstadt.de/~p_larbig/wlan/ (http://homepages.tu-darmstadt.de/~p_larbig/wlan/).
Enjoy.
Well, that was a good try, but were back to the Segmentation Fault again.
Perhaps 3.0.0 and 3.0.1 are biased more towards the 2.6 kernel architecture ?
Anyway, its was a nice try evilJazz ;-)
I notice that the module has been reduced in size from a working 2.0.1 equaling 458,752 Bytes
down to the 3.0.1 version equaling 327,680 Bytes..
thats over 130,000 Bytes missing from the working version.
It also looks very much like the shape and size of the version Capn_Fish produced earlier with
the same fault showing.
perhaps we could just add the sensitivity patch to the working 2.0.1 source code and try that ?
All in all, I'll leave it for you to decide the next move
Thanks a million
-
also, someone made a sensitivity patch for 2.0.x. I,m now hosting it:
http://allelectrix.bravehost.com/rt73-k2wr...nsitivity.patch (http://allelectrix.bravehost.com/rt73-k2wrlz-2.0.x-rx_sensitivity.patch)
you probably know the rest evilJazz, but for those that don't the rest is:
$ cd rt73-k2wrlz-2.0.0 (your driver source directory)
$ patch -p0 < /path/to/rt73-k2wrlz-2.0.x-rx_sensitivity.patch
$ make
Thanks again
-
OK, with all of your fine help, I just compiled it myself now ! I DONT believe I did it !
well, here it is:
[ You are not allowed to view attachments ]
delete .zip and replace with .o
This is not a ZIP file it is a .o file . I had to change the suffix as OESF don't allow the modules to upload.
If someone can test this with any rt73 WiFi Card, but especially D-Link DWL G122 C1 I would appreciate it as I am using
Angstrom as a test for something else at the moment.
Also, the Sensitivity patch above, was applied before compilation. so, this is a different module to the first working
module listed in this thread which was eviLjazz's rt73-k2wrlz-2.0.1 . But, it is the same source code used.
Thanks everyone !
-
This post got me really excited as I have a usb card that this driver runs. I just tried both your new one and eviljazz'. Both load fine but neither work for my card, unless I am missing something. Here is what I get when the card is plugged into my powered hub with the module already loaded:
hub.c: USB new device connect on bus1, assigned device number 9
usb.c: kmalloc IF c3d348c0, numif 1
usb.c: new device strings: Mfr=1, Product=2, SerialNumber=0
usb.c: USB device number 9 default language ID 0x409
Manufacturer: Ralink
Product: 802.11 bg WLAN
Device Descriptor not matching
usb.c: unhandled interfaces on device
usb.c: USB device 9 (vend/prod 0x7d1/0x3c04) is not claimed by any active driver.
Length = 18
DescriptorType = 01
USB version = 2.00
Vendor:Product = 07d1:3c04
MaxPacketSize0 = 64
NumConfigurations = 1
Device version = 0.01
Device Class:SubClass:Protocol = 00:00:00
Per-interface classes
Configuration:
bLength = 9
bDescriptorType = 02
wTotalLength = 0020
bNumInterfaces = 01
bConfigurationValue = 01
iConfiguration = 00
bmAttributes = 80
MaxPower = 300mA
Interface: 0
Alternate Setting: 0
bLength = 9
bDescriptorType = 04
bInterfaceNumber = 00
bAlternateSetting = 00
bNumEndpoints = 02
bInterface Class:SubClass:Protocol = ff:ff:ff
iInterface = 00
Endpoint:
bLength = 7
bDescriptorType = 05
bEndpointAddress = 81 (in)
bmAttributes = 02 (Bulk)
wMaxPacketSize = 0040
bInterval = 00
Endpoint:
bLength = 7
bDescriptorType = 05
bEndpointAddress = 01 (out)
bmAttributes = 02 (Bulk)
wMaxPacketSize = 0040
bInterval = 00
usb.c: kusbd: /sbin/hotplug add 9
The rt73 driver works for this card on my linux box and on debian on this zaurus. Maybe this older version of the driver doesn't support my card? It's a d-link WUA-1340 card.
-
xamindar wrote about his ralink rt2571 chipset card...
Hi there xamindar, thanks for your input into this very exciting thread.
eviLjazz and Capn_Fish still have my head reeling from all the big buzz they caused !
OK, first things first, this card is an rt2571 chipset card. It just so turns out that I also have an rt2571 Card too !
When I try to load this card with the rt73.o module installed, it will not load at all.
When I try to load this card with the rt2570.o module, everything works.
When I try to load my rt73 chipset card with the rt2570.o module, it starts up, but when I issue wireless-tools commands, it freezes
the system up or makes it run very slowly.
The solution therefore is a simple one:
I need the rt73.o module listed as I have an rt73 based card
You need the rt2570.o module as you have an rt2571 based card !
And here is the said module, which, in fact, I compiled yesterday from rt2570-k2wrlz-1.6.0 sources:
[ You are not allowed to view attachments ]
same as before, change .zip to .o then load to:
/lib/modules/2.4.20/kernel/drivers/net/wireless
delete the rt73.o module now, or move it out of the ..../wireless directory to say /mnt/card
then do:
depmod -a
modprobe rt2570
insert the card, the card fires up immediately.
execute:
iwconfig
you should see the card listed under the RT2500 driver and NOT the rt73 driver.
BTW All the modules in this thread feature packet injection in Monitor mode, and this one is no exception
let us know how you get on with this Module, all the best from all of us !
-
Yes, after applying the Sensitivity patch to the Module, the rt73 cards work very well indeed, picking up everything in sight, it gives even better
reporting than my trusty, but rusty D-Link 802.11b CF Card.
I therefore recommend the module I compiled.
The rt2570 module also works well with the rt2570 cards and I report no sensitivity reception problems with this module.
-
You need the rt2570.o module as you have an rt2571 based card !
Just thought I would add a reply to this as I never did before. I just gave up. But to save anyone any trouble if they come across this. Allelectrix is incorrect. rt2570 is not the same as rt2571. This module is NOT for my card. As I stated in my previous post my card (d-link WUA-1340) requires the rt73 module. Weird it never worked correctly for me on my Zaurus. Might be because I am on the Sharp rom and not pdaXrom.
I just boot the Zaurus into debian anyway where it works correctly and it's fine that way.