[PATCH 0/3] Move rtl8192e from staging

[Stefan Lippers-Hollmann]

Hi

On Tuesday 15 November 2011, Sean MacLennan wrote:
> This patchset is an attempt to move the rtl8192e wireless driver from
> staging to the drivers directory proper. The original staging driver
> was one driver in one directory. This patchset creates two new
> subdirectories:

I guess you really want to CC linux-wireless at vger.kernel.org

> * drivers/net/wireless/rtllib - common rtl819x driver code
> * drivers/net/wireless/rtllib/rtl8192e - rtl8192e driver code
> 
> There are now six drivers:
> 
> * rtllib
> * rtllib_crypt
> * rtllib_crypt_ccmp/tkip/wep

However looking at these, and select WIRELESS_EXT/ WEXT_PRIV strongly 
suggests that r8192e hasn't been ported to mac80211 yet, but still uses
yet another 2.6.16-era ieee80211softmac fork, which would - to the best
of my knowledge - make it unfit for mainline.

> * r8192e
[...]

Likewise (patches against linux-next, next-20111115, following).

--- /dev/null
+++ b/drivers/net/wireless/rtllib/Makefile
@@ -0,0 +1,41 @@
+ccflags-y += -DUSE_FW_SOURCE_IMG_FILE
always true, there are no firmware arrays anymore, neither does 
anything in rtl8192e evaluate it anymore.

+ccflags-y += -DCONFIG_PM_RTL
(k)config option or not (my guess is no, given that it has been enabled
unconditionally so far)?

+ccflags-y += -DCONFIG_PM
nothing in rtl8192e evaluates PM anymore.

+ccflags-y += -DHAVE_NET_DEVICE_OPS
always true in current kernels, just a few code paths to adapt

+ccflags-y += -DENABLE_DOT11D
nothing in rtl8192e evaluates ENABLE_DOT11D anymore, it's active 
unconditionally (as it should be).


Disclaimer: Please don't consider this to be a thorough review, these 
issues just caught my eye from the cover letter and the first few lines
of your patch.

Regards
	Stefan Lippers-Hollmann