Hi!

This is driver for w35und usb wifi -- also in kohjinsha
subnotebook. It should work well enough to associate and ping, but it
obviously needs to be rewritten two more times...

OTOH worst horrors (like embedded wifi stack) should have been fixed
already...

(against 2.6.27-rc8).

Signed-off-by: Pavel Machek <pavel@suse.cz>

diff --git a/drivers/net/wireless/Kconfig b/drivers/net/wireless/Kconfig
index 9931b5a..3929a5d 100644
--- a/drivers/net/wireless/Kconfig
+++ b/drivers/net/wireless/Kconfig
@@ -702,5 +702,6 @@ source "drivers/net/wireless/b43/Kconfig
 source "drivers/net/wireless/b43legacy/Kconfig"
 source "drivers/net/wireless/zd1211rw/Kconfig"
 source "drivers/net/wireless/rt2x00/Kconfig"
+source "drivers/net/wireless/winbond/Kconfig"
 
 endmenu
diff --git a/drivers/net/wireless/Makefile b/drivers/net/wireless/Makefile
index 59aa89e..f1435e5 100644
--- a/drivers/net/wireless/Makefile
+++ b/drivers/net/wireless/Makefile
@@ -48,6 +48,8 @@ obj-$(CONFIG_USB_NET_RNDIS_WLAN)	+= rndi
 obj-$(CONFIG_USB_ZD1201)	+= zd1201.o
 obj-$(CONFIG_LIBERTAS)		+= libertas/
 
+obj-$(CONFIG_W35UND)		+= winbond/
+
 rtl8180-objs		:= rtl8180_dev.o rtl8180_rtl8225.o rtl8180_sa2400.o rtl8180_max2820.o rtl8180_grf5101.o
 rtl8187-objs		:= rtl8187_dev.o rtl8187_rtl8225.o
 
diff --git a/drivers/net/wireless/winbond/Kconfig b/drivers/net/wireless/winbond/Kconfig
new file mode 100644
index 0000000..94499d3
--- /dev/null
+++ b/drivers/net/wireless/winbond/Kconfig
@@ -0,0 +1,6 @@
+config W35UND
+	tristate "Winbond driver"
+	depends on MAC80211 && WLAN_80211 && EXPERIMENTAL && !4KSTACKS
+	---help---
+	  This is highly experimental driver for winbond wifi card on some Kohjinsha notebooks
+	  Check http://code.google.com/p/winbondport/ for new version
diff --git a/drivers/net/wireless/winbond/Makefile b/drivers/net/wireless/winbond/Makefile
new file mode 100644
index 0000000..4682dd1
--- /dev/null
+++ b/drivers/net/wireless/winbond/Makefile
@@ -0,0 +1,17 ...
[ message continues ]

winbond is a little too generic, please use the same name as the driver

Again, please give the option a useful name, and a useful description
including mentioning the actual chip name in there.


This is not a proper Makefile, and having these files in subdirectories
is not really a good idea either.  So move all files into one directory
and write a proper Makefile for it:

w35und-y	+= wbusb.o wb35reg.o wb35rx.o wb35tx.o mds.o mlmetxrx.o \
		   mto.o phy_calibration.o reg.o rxisr.o sme_api.o wbhal.o \
		   wblinux.o




neither.


No comments on the code now, I don't think I want to look at it..

--

[ message continues ]

Whatever in the driver uses lots of stack space needs to be fixed, plain
an simple...  no excuse for that.


--

[ message continues ]

No, you don't want to :-).
								Pavel

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
--

[ message continues ]

Quick review just looking at only a few files.











Documentation/networking/ somewhere, but this documentation is not


Please consider using cfg80211 and new regulatory framework for this.
Then you can refer to iw:

http://wireless.kernel.org/en/users/Documentation/iw

Regulatory stuff:

http://wireless.kernel.org/en/developers/Regulatory



Please install sparse and then compile the driver using C=1, and correct
the warnings.

Ok well that should take a while..

  Luis
--

[ message continues ]

Heh, in the mean time, I'll suck it into drivers/staging/, moving the
files around so that it is self-contained and remove the extra crud and
fix up the build stuff so that it plays nice with the rest of the
kernel.

I'll post the updated patch when it goes into my tree.

thanks,

greg k-h
--

[ message continues ]

On Wed, 1 Oct 2008 14:36:56 +0200

why this !4KSTACKS? At minimum that should be in the "to fix" list ;)

-- 
Arjan van de Ven 	Intel Open Source Technology Centre
For development, discussion and tips for power savings, 
visit http://www.lesswatts.org
--

[ message continues ]

Yeah seriously, the driver isn't going anywhere unless it works with 4k
stacks...

Dan


--

[ message continues ]

Well, it's going into the staging tree, where it can now be fixed up :)

And yes, I'll go add that to the TODO list for this driver, thanks.

greg k-h
--

[ message continues ]