Uli Kunitz wrote:
> we had wireless-dev for 2.6.19, but I had no idea about your work for
> 2.6.20. Actually I wanted to integrate Jon Smirl's patch, but there was
> already the fix in wireless-2.6.
> I would like to use the name wireless-2.6, because that indicates
> better, that it is against wireless-2.6. Notify that Jon has called the
> descape/d80211 tree wireless-dev. So wireless-dev might not be the most
> prudent choice.
Yep -- I didn't object to the name change, just advised that it would
help me out if you could use my patch instead, since I've already
ensured it matches upstream exactly. As I'm sure you know, you can
easily pull my wireless-dev branch into your aptly-named wireless-2.6
branch.
> I know that we have discussed already, but I suggest again to move also
> with the softmac-based driver to in-tree development. It would certainly
> save time. I would still create a tar tree, that could be compiled
> outside of the kernel.
Are you still planning on dropping the softmac work and switching to the
d80211 port? If so, I suggest you make that jump directly. When you have
created the tree I will clone it and we will go from there.
I will continue to maintain the softmac one in the current out-of-kernel
setup but wouldn't expect you to contribute there.
> Regarding the known issues of my master tree. I admit there are problems
> with the AL2230 RF, which doesn't work very well for me with or without
> the signal strength handling. There are no such issues with the RF2959.
> How about doing signal strength handling only for RF2959?
I can't comment as I don't have any RF2959 devices. If you've done some
kind of sensible testing, then I don't think I would object.
The kind of testing I would suggest is:
1. Move the system right next to the AP and measure transfer rates when
signal strength handling is disabled. Then enable signal strength
handling, it should go into strong signal (low power) mode, and then
repeat the transfer rate test.
2. Start with the system close to the AP and ensure it goes into strong
signal mode. Then move away from the AP and find the point where it
switches into 'normal' mode. Ensure that the range it flip-flops on
is relatively small and that the point at which it changes to normal
mode is something like what you might expect.
> 2.6.19.1 without softmac-patches is not a good setup, because
> deauthentication is not handled. It should also be noted, that calling
> ieee80211_rx in hard irq context might create problems in specific
> setups. This is fixed in my tree.
Ah yes. I will send in those patches for 2.6.19.2 if they are not
already queued.
Daniel
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Zd1211-devs mailing list - http://zd1211.ath.cx/
Unsubscribe: https://lists.sourceforge.net/lists/listinfo/zd1211-devs
