rtl-wifi-devel

Re: [Rtl-wifi-devel] Realtek rtl8187B wireless chip support for linux

[Hin-Tak L. <hin...@ya...>]

Hi, I thought I should mention that there were some 2.6.23<->26.24 
breakage and I have attached the change to the redhat bug report
I filed earlier - the redhat folks wanted to enhance the in-kernel 
realtek driver rather than talking this is directly, but they are 
not ready yet, so I have an interrim patch for for the driver to 
make it work in kernel 2.6.24:

https://bugzilla.redhat.com/show_bug.cgi?id=432280

--- On Sun, 10/2/08, Hin-Tak Leung <hin...@ya...> wrote:

> From: Hin-Tak Leung <hin...@ya...>
> Subject: Realtek rtl8187B wireless chip support for linux
> To: wl...@re...
> Cc: rtl...@li..., cu...@ze..., and...@ti..., fla...@so...
> Date: Sunday, 10 February, 2008, 8:59 PM
> One question: why is the official driver for 8187B not
> available for download?
> I have got it off a 3rd-party site:
> http://www.datanorth.net/~cuervo/rtl8187b/
> 
> I have reworked the modification made by Johnny Cuervo
> (attached as
> rtl8187-modified-diff) with a more useful kernel debug
> message and some 
> comments about why the drop through to default doesn't
> work
> (rtl8187B_linux_24...diff), and I am currently using the
> mod right now
> on the 8187B chip which has a product id 0x8197. 
> 
> I hope this change get more widely and officially
> adopted...
> 
> 
> 
>      
> ___________________________________________________________
> Yahoo! Answers - Got a question? Someone out there knows
> the answer. Try it
> now.
> http://uk.answers.yahoo.com/diff -ruBb
> old/rtl8187B_linux_24.6.1024.0822.2007/rtl8187/r8187_core.c
> rtl8187B_linux_24.6.1024.0822.2007/rtl8187/r8187_core.c
> ---
> old/rtl8187B_linux_24.6.1024.0822.2007/rtl8187/r8187_core.c	2007-08-22
> 08:54:58.000000000 +0100
> +++
> rtl8187B_linux_24.6.1024.0822.2007/rtl8187/r8187_core.c	2008-02-10
> 03:25:46.000000000 +0000
> @@ -93,6 +93,7 @@
>  static struct usb_device_id rtl8187_usb_id_tbl[] = {
>  	{USB_DEVICE(USB_VENDOR_ID_REALTEK, 0x8187)},
>  	{USB_DEVICE(USB_VENDOR_ID_REALTEK, 0x8189)},
> +	{USB_DEVICE(USB_VENDOR_ID_REALTEK, 0x8197)},
>  //	{USB_DEVICE_VER(USB_VENDOR_ID_REALTEK,
> 0x8187,0x0200,0x0200)},
>  	{USB_DEVICE(USB_VENDOR_ID_NETGEAR, 0x6100)},
>  	{USB_DEVICE(USB_VENDOR_ID_NETGEAR, 0x6a00)},
> @@ -2833,6 +2834,11 @@
>  	idProduct = le16_to_cpu(udev->descriptor.idProduct);
>  	
>  	switch (idProduct) {
> +		case 0x8197:
> +		  /* priv->rf_chip = 6 for this chip, >
> EPROM_RFCHIPID_RTL8225U = 5 */
> +		  /* rtl8225_is_V_z2(dev) return false, and drop through
> to default: does not work */
> +		  /* drop through to 0x8189 */
> +		  DMESG("Using configuration of Product id 0x8189
> for newer 0x8197");
>  		case 0x8189:
>  			/* check RF frontend chipset */
>  			priv->card_8187 = NIC_8187B;


      __________________________________________________________
Sent from Yahoo! Mail.
More Ways to Keep in Touch. http://uk.docs.yahoo.com/nowyoucan.html

Re: [Rtl-wifi-devel] Realtek rtl8187B wireless chip support for linux

[Hin-Tak L. <hin...@ya...>]

I just checked Johnny Cuervo's web site on the 2.6.24 change he received - it is actually
somewhat different from my 2.6.24-related diff I made - I used sg_init_one() rather than
sg_set_page(). Well, I haven't had any 2.6.24 crashes yet (I am using wireless while
writing this). Maybe you can have my diff for reference also. This is to be applied
*in addition* to the earlier vender-id/product-id related change. 

Can I make a suggestion: I prefer having the changes as diff's - rather than downloading 
a whole modified tarball, and a diff file isn't too much so maybe Johnny can try to provide both?



      ___________________________________________________________ 
Rise to the challenge for Sport Relief with Yahoo! For Good  

http://uk.promotions.yahoo.com/forgood/

Re: [Rtl-wifi-devel] Realtek rtl8187B wireless chip support for linux

[Andrea M. <and...@ti...>]

Can someone please summarize me what is going on there?
I had no time to follow things in driver development since quite a bit of time..
Is this to add support for rtl8187B?
Even if I had no time to study well the chip, it seems it is quite
different from the rtl8187L
(some registers changed offset, a lot of new, undocumented registers
appared, radio programming seems also very different, as well as the
RTX mechanism).
I wish to write a new driver for those device, but I have no time no.
furthermore someone told me that some card with Rtl8187B has PCI id
like rtl8187L.
I wonder if there are two different rtl8187B chip version.
I can ask to Realtek eventually, however I was waiting to have a free
time slot to evaluate things more carefully..
Also I have some pending patches to prepare for rtl8180 code but I
have had no time yet to prepare them

Andrea


On Mon, Mar 24, 2008 at 2:57 PM, Hin-Tak Leung <hin...@ya...> wrote:
> Hi, I thought I should mention that there were some 2.6.23<->26.24
>  breakage and I have attached the change to the redhat bug report
>  I filed earlier - the redhat folks wanted to enhance the in-kernel
>  realtek driver rather than talking this is directly, but they are
>  not ready yet, so I have an interrim patch for for the driver to
>  make it work in kernel 2.6.24:
>
>  https://bugzilla.redhat.com/show_bug.cgi?id=432280
>
>  --- On Sun, 10/2/08, Hin-Tak Leung <hin...@ya...> wrote:
>
>  > From: Hin-Tak Leung <hin...@ya...>
>  > Subject: Realtek rtl8187B wireless chip support for linux
>  > To: wl...@re...
>  > Cc: rtl...@li..., cu...@ze..., and...@ti..., fla...@so...
>  > Date: Sunday, 10 February, 2008, 8:59 PM
>  > One question: why is the official driver for 8187B not
>  > available for download?
>  > I have got it off a 3rd-party site:
>  > http://www.datanorth.net/~cuervo/rtl8187b/
>  >
>  > I have reworked the modification made by Johnny Cuervo
>  > (attached as
>  > rtl8187-modified-diff) with a more useful kernel debug
>  > message and some
>  > comments about why the drop through to default doesn't
>  > work
>  > (rtl8187B_linux_24...diff), and I am currently using the
>  > mod right now
>  > on the 8187B chip which has a product id 0x8197.
>  >
>  > I hope this change get more widely and officially
>  > adopted...
>  >
>  >
>  >
>  >
>  > ___________________________________________________________
>  > Yahoo! Answers - Got a question? Someone out there knows
>  > the answer. Try it
>  > now.
>  > http://uk.answers.yahoo.com/diff -ruBb
>  > old/rtl8187B_linux_24.6.1024.0822.2007/rtl8187/r8187_core.c
>  > rtl8187B_linux_24.6.1024.0822.2007/rtl8187/r8187_core.c
>  > ---
>  > old/rtl8187B_linux_24.6.1024.0822.2007/rtl8187/r8187_core.c   2007-08-22
>  > 08:54:58.000000000 +0100
>  > +++
>  > rtl8187B_linux_24.6.1024.0822.2007/rtl8187/r8187_core.c       2008-02-10
>  > 03:25:46.000000000 +0000
>  > @@ -93,6 +93,7 @@
>  >  static struct usb_device_id rtl8187_usb_id_tbl[] = {
>  >       {USB_DEVICE(USB_VENDOR_ID_REALTEK, 0x8187)},
>  >       {USB_DEVICE(USB_VENDOR_ID_REALTEK, 0x8189)},
>  > +     {USB_DEVICE(USB_VENDOR_ID_REALTEK, 0x8197)},
>  >  //   {USB_DEVICE_VER(USB_VENDOR_ID_REALTEK,
>  > 0x8187,0x0200,0x0200)},
>  >       {USB_DEVICE(USB_VENDOR_ID_NETGEAR, 0x6100)},
>  >       {USB_DEVICE(USB_VENDOR_ID_NETGEAR, 0x6a00)},
>  > @@ -2833,6 +2834,11 @@
>  >       idProduct = le16_to_cpu(udev->descriptor.idProduct);
>  >
>  >       switch (idProduct) {
>  > +             case 0x8197:
>  > +               /* priv->rf_chip = 6 for this chip, >
>  > EPROM_RFCHIPID_RTL8225U = 5 */
>  > +               /* rtl8225_is_V_z2(dev) return false, and drop through
>  > to default: does not work */
>  > +               /* drop through to 0x8189 */
>  > +               DMESG("Using configuration of Product id 0x8189
>  > for newer 0x8197");
>  >               case 0x8189:
>  >                       /* check RF frontend chipset */
>  >                       priv->card_8187 = NIC_8187B;
>
>
>
>       __________________________________________________________
>  Sent from Yahoo! Mail.
>  More Ways to Keep in Touch. http://uk.docs.yahoo.com/nowyoucan.html
>

Re: [Rtl-wifi-devel] Realtek rtl8187B wireless chip support for linux

[Hin-Tak L. <hin...@ya...>]

> On Tue, Mar 25, 2008 at 01:19:22PM +0100, Andrea Merello
> wrote:
> Can someone please summarize me what is going on
> there?
> I had no time to follow things in driver development
> since quite a bit of time..
> Is this to add support for rtl8187B?

Don't worry, It is all in good hands... I attached Johnny Cuervo's mod's
and and some of my enhancements to a redhat bug report (a dkms config file
to rebuild the kernel module automatically for a new redhat kernel release
in the next reboot, and also a 2.6.23<->2.6.24 patch I made recently) to a
redhat bug report. 

The redhat people said they don't have to incoporate my ideas but choose
to enhance the in-kernel realtek driver; so eventually that's going to happen, I believe; meanwhile, I am using my own mod with dkms (dynamic 
kernel module support) on my satellite A210 until I get an official redhat kernel update.


      __________________________________________________________
Sent from Yahoo! Mail.
More Ways to Keep in Touch. http://uk.docs.yahoo.com/nowyoucan.html

Re: [Rtl-wifi-devel] Realtek rtl8187B wireless chip support for linux

[Hin-Tak L. <hin...@ya...>]

Johnny Cuervo wrote:
<snipped>
> It's functional, just a little glitchy. I've had people tell me WEP/WPA
> breaks, and signal strength isn't reported. The on/off switch doesn't do
> much of anything, either, but I don't know if that's the device's
> problem or the rest of the system's.
<snipped>
> If I can help at all, let me know.

There is a new mac80211-based driver available from Herton of Mandriva,
and it needs a little change to get all functional, but it is working alright.
all the stuff, patches and discussions are in the bug report. Would be useful
if you have a go at it.

>>>  https://bugzilla.redhat.com/show_bug.cgi?id=432280

It is interesting to re-read what was written some weeks ago; yes, my motivation
for making the new driver work was that WEP causes kernel-oops - and
I have also observed that the on/off switch does nothing (unlike how it is under 
Vista). I have not tried encryption with the new driver yet, but it is likely
to be cleaner and probably would "just work". (I believe the old driver breaks
because of crypto changes in the kernel). Also the new driver is quite happy
to be unloaded (modprobe -r'ed) so one can change/mod within limits and just 
overwrite the old module and do depmod -a and reload again. The old driver
is quite buggy in that it sometimes(often?) locks up on unload.