Can the devs give a short/expanded overview of WCN3660 and how it communicates with the driver?

Farm Dve farmdve at data.bg
Thu Feb 19 01:28:24 PST 2015


>>Can you share the changes?
The changes are very buggy, as I am still experimenting.

I have so far identified several hurdles.
1) The Wi-Fi chip is sending either stripped frames, or only very
specific packets. After a period of inactivity, that is no TX/Commands
being sent to the chip, it stops sending us any data.
2) The fw will have to be modified to make the chip not stop
broadcasting frames, but also send us EVERYTHING, not just
cherry-picked data.
3) The firmware has authentication and verification due to Trust Zone,
this can be disabled only if peripheral-loader.c and/or pil-riva.c are
modified to use the non-secure boot option, I managed to insert my
modified firmware this way.
3.1) The problem here is, even if we could enable monitor mode, it
would require recompiling the kernel as modules are built-in, or if we
tried to compile the prima driver as a loadable kernel module(.ko),
this would still require the kernel to be compiled with loadable
kernel module support, which I believe default FLO kernels are
without. The only option I see here, is to binary patch the kernel, OR
do some other kernel routine redirection.
3.2) The firmware file is huge, 2.4 megabytes of ARM code, this is
partially because the firmware implements 802.11, Bluetooth and FM
radio in one fw, AND because of poor code design, the prima driver is
but an illustration of the atrocities the Qualcomm developers did just
so they could keep their jobs(google the term Job 'Security through
Obscurity' if you don't believe me), sorry for being a douche about
this. I've looked at the various debug strings in the firmware file,
and I quickly understood it's in the same shape as the prima driver.

I am in no way saying these are not solvable, just sharing my findings so far.

On Tue, Feb 17, 2015 at 1:06 PM, Mihir Shete <mihirsht at gmail.com> wrote:
>> Unfortunately, there seems to be an issue where WDTS_RxPacket stops
>> being called after a while for some reason, but till I figure out how
>> to get 802.11 frames, I won't be bothered by that.
>>
> Can you share the changes?
>
>
> --
> ~Mihir~



More information about the wcn36xx mailing list