Arek Jankowski wrote:
> Some of this is already known, but I hope this information may be still 
> useful 
> for some. My main question is why the zd1211rw performance is so low? Is 
> there a way to debug this problem?

You didn't state whether you were measuring RX or TX performance, or 
whether you were using 802.11b or g.

In general zd1211rw performance is not that bad, although I think we 
still only meet about half the RX speed of the vendor driver, unsure 
about TX. We're not sure why, but at the same time nobody has really 
investigated. Our driver is less efficient in that we allocate memory in 
the RX path rather than using some kind of circular buffer system, but 
this should be insignificant (unless we're pegging the CPU on your ARM?)

Also remember that the vendor driver dynamically adjust it's rate based 
on signal quality but zd1211rw defaults to 11m and relies on the user 
setting a sensible rate. If you set 54M but don't have a strong signal, 
then you'll get considerably worse performance than (e.g.) 24M.

To start with, use a vendor driver which does not have large packet 
support, otherwise it is not a fair comparison. (It is *possible* that 
zd1211rw not using jumbo frames is a reason for slowdown but I expect it 
will be an insignificant difference)

The next steps would be to find what's different. You could dump the 
controlset on each outgoing frame and check that the match for both 
drivers. You could also write a patch for both drivers to dump the 
contents of all control registers on-demand (possibly through the 
ethtool interface) and then compare results from the 2 drivers.

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

Reply via email to