doc updates.
This commit is contained in:
parent
b9e74e7a7e
commit
cf38f98c1f
1
CHANGES
1
CHANGES
|
@ -41,6 +41,7 @@
|
|||
* Intersil Corporation as part of PRISM(R) chipset product development.
|
||||
*
|
||||
* --------------------------------------------------------------------
|
||||
- additions to the MIB.
|
||||
- Cleaned up wireless extension stuff a little bit.
|
||||
- Add the Microsoft MN510 device ID.
|
||||
- Finally fix the 'crash on unload' problem for 2.2.x kernels with PCI.
|
||||
|
|
19
README
19
README
|
@ -300,21 +300,18 @@ F) Run 'ifconfig wlan0 <your IP address>'
|
|||
NOTE: USB devices currently integrated into the hotplug subsystem, and
|
||||
use the /etc/wlan.conf configuration file.
|
||||
|
||||
IMPORTANT: Due to an issue with the current firmware in Prism USB
|
||||
devices, the driver must usually do a port reset (which is normally
|
||||
handled by usbcore).
|
||||
IMPORTANT: Due to an issue with some versions of the Prism USB firmware,
|
||||
the driver usually needs to perform a port reset.
|
||||
|
||||
Some combinations of usb low-level drivers, kernel releases, and
|
||||
hardware don't like this, and usually end up generating a kernel OOPS.
|
||||
newer kernels are much better in this regard. In particular, Intel usb
|
||||
controllers are the most trouble-prone.
|
||||
|
||||
Unfortunately, there's nothing we can do about it until new firmware
|
||||
becomes available. The good news is that the firmware developers are aware
|
||||
of the problem, they have a fix, and the new firmware is coming, we just
|
||||
don't know precisely when it will be available. It all depends on how quickly
|
||||
they can get it through their QA process. When it does become available,
|
||||
install it in your device using the Windows firmware update utility,
|
||||
remove the "prism2_doreset" on the modprobe command line above and everything
|
||||
will work just peachy.
|
||||
The OOPS is due to bugs in the linux USB core, and newer kernels
|
||||
(2.4.19 and later) behave much better in this regard.
|
||||
|
||||
However, the good news is that primary firmware 1.1.2 seems to resolve
|
||||
the need for the port reset to begin with. Contact your vendor to
|
||||
request this update.
|
||||
|
||||
|
|
1
TODO
1
TODO
|
@ -49,7 +49,6 @@ Linux WLAN NG TODO
|
|||
- Make the command-completion interrupt driven vs busywaiting.
|
||||
- Possibly hook up DMA on the Rx side for PCI devices
|
||||
- Further investigate lingering SMP issues
|
||||
- Add a query of the ifstate.
|
||||
- Move all of the wireless extensions work up to p80211 so that the new
|
||||
MACs will get wireless extensions support for free!
|
||||
- [USB] Remove driver's port-reset when new primary firmware becomes
|
||||
|
|
Loading…
Reference in a new issue