Subject: rtw(4), work-in-progress Realtek 802.11b driver
To: None <>
From: David Young <>
List: tech-net
Date: 09/25/2004 22:11:42
I have checked-in rtw, an unfinished driver for Realtek 802.11b chips.
I don't think that it is far from completion, but I don't have much time
to work on it.  I hope that others will collaborate to finish it.

rtw is a significant driver because many cheap 802.11b Cardbus & PCI
adapters use Realtek's chip.  Also, there are cheap ($60) access points
available in the U.S. and Australia that contain the RTL8181, Realtek's
802.11b System on a Chip.  There is an effort underway to port NetBSD
to RTL8181, but it's stalled waiting for some GNU toolchain modifications.

The status of the driver is that the transmit section of the code is
mostly unwritten.  The receive section of code is finished, but I cannot
get the chip to give me an Rx interrupt.  I have made the chip give
me timeout interrupts, so I know that at least that much is working.
I'm not certain that I'm blasting the bits at the RF front-end properly.

I have seen Realtek-based PCI & Cardbus cards with three different
radio transceiver/synthesizers, the Maxim MAX2820, Philips SA2400,
and GCT GRF5101.  There may be a variety with the RFMD RF2958, too.
I have tried to abstract the functions of the transceiver/synthesizer;
see struct rtw_rf and struct rtw_phy.

Recently I got word of a Linux driver for the Realtek/Philips combo.
It is open source.  Apparently it works.  It is a good place to seek clue.

I had written the code intending to write a comprehensive test suite
for it.  For that reason, there are many small subroutines.  I intended
to write a test harness for each one.  I have had some thoughts about
simulating the hardware, too; I think I sent those to tech-kern@ many
months ago.


David Young             OJC Technologies      Urbana, IL * (217) 278-3933