W dniu 4 stycznia 2010 21:55 użytkownik Larry Finger <larry.fin...@lwfinger.net> napisał: > On 01/04/2010 02:36 PM, Rafał Miłecki wrote: >> I try to update >> int b43_phy_initn(struct b43_wldev *dev) >> to match current specs. >> >> Currently it contains many operations/calls that I can not find in >> http://bcm-v4.sipsolutions.net/802.11/PHY/Init/N >> >> First I've just posted patch that removed "//TODO: Spectral >> management". I don't see anything about "spectral management" in >> specs. >> >> Next there is a lot of code after "b43_nphy_workarounds(dev);" call >> that I can not recognize. Let's just take some lines for example: >> b43_nphy_reset_cca(dev); >> ssb_write32(dev->dev, SSB_TMSLOW, ...); >> b43_nphy_force_rf_sequence(...); >> b43_phy_read(dev, B43_NPHY_CLASSCTL); /* dummy read */ >> >> I don't see connection between that code and specs. >> >> Michael: I guess you're author of this code (phy_n.c came from nphy.c, >> not sure how to check log for not-existing file). Could you help me >> and explain this, please? > > I can answer this question. The original N PHY specs and the code generated > from > them were for a driver that is now 2 versions old. The main changes are due to > later revisions of the N PHY, but there also seem to be some refactoring. To > keep it straight, I am evolving the current specs to match the latest driver > that we have available. Whenever you see code like the above that does not > match > the specs, you should probably rip it out. If you have any questions, please > ask > me if a given spec page is current. > > I have been trying to get initn and all the routines that it calls up to date, > but this RE process is so mind-numbing that I can only work on it for so long.
I'm afraid it starts to sound boring... but thanks... again ;) -- Rafał _______________________________________________ Bcm43xx-dev mailing list Bcm43xx-dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/bcm43xx-dev