WOL filters specifications.

[Paul Fox]

alberto wrote:
 > On mer, 2010-09-01 at 11:09 -0400, John W. Linville wrote:
 > > Adding S-o-b's...
 > > 
 > > This patch is nearly 2 years old -- is there a real problem here?
 > 
 > Well, about the nowadays driver, that piece of code is never used,
 > so it is hard that it was successfully tested..
 > I'm asking if there was a rationale behind that. (If anyone knows)
 > Because We are going to try to use that functionality but that
 > wide differences between specs and code makes me wonder that
 > the specs I have maybe are not right.
 > 
 > There had been also a try to use that functionality in the thread:
 > http://thread.gmane.org/gmane.linux.kernel.wireless.general/22563/focus=23312
 > 
 > But porting this patch to the nowadays driver results (in my case
 > 88W8686 based wifi chip connected via sdio) in deafness of the wifi
 > chip about the iwpriv calls that sets rules: iwpriv calls that get the 
 > rules results always in 0 rules set.

i've been told that the 5.1 spec is probably the most recent.  it's
also been my impression that the 8686 (which is used on the OLPC
XO-1.5) doesn't support the full suite of WOL filtering that was
present in, for instance, the 8388 as used on the XO-1.  i would love to
be wrong about this.

our most recent releases haven't been able to take advantage of
wake-on-arp, or on-multicast, because of the lack of firmware support
on the new laptops, and lack of driver/user infrastructure in recent
kernels for the old laptop.  (not to mention our lack of resources to
correct either issue.)

paul

 > 
 > 
 > > 
 > > On Wed, Sep 01, 2010 at 04:52:01PM +0200, Alberto Panizzo wrote:
 > > > The patch 582c1b538fb47a2d6a41dbdadb031086c49446c1 introduced
 > > > (other than the ability to wake on specific criteria) structures
 > > > that are supposed to be used to specify fine grained wol roules.
 > > > Those rules tells the wifi chip to filter the received Ethernet packets
 > > > to be used for the wake event.
 > > > 
 > > > The question is, why those structures are so different from the
 > > > v5.1 firmware specifications?(pag 131: MrvlIETypes_HostSleepFilterType1
 > > > linked from pag 85: 5.13.3.3 CMD_802_11_HOST_SLEEP_CFG)
 > > > 
 > > > Is the 5.1 fw specs too old? or are these structures not valid?
 > > > Looking at the Marvell driver the Ethernet rules shall be specified
 > > > as written in the 5.1 firmware specs.
 > > > 
 > > > Thanks in advance for all suggestions.
 > > > 
 > > > Best regards,
 > > > 
 > > > -- 
 > > > Alberto!
 > > > 
 > > >         Be Persistent!
 > > >                 - Greg Kroah-Hartman (FOSDEM 2010)
 > > > 
 > > > 
 > 
 > 
 > 
 > _______________________________________________
 > libertas-dev mailing list
 > libertas-dev at lists.infradead.org
 > http://lists.infradead.org/mailman/listinfo/libertas-dev

=---------------------
 paul fox, pgf at laptop.org