On Sun, Apr 5, 2009 at 3:54 PM, Francesco Gringoli <francesco.gring...@ing.unibs.it> wrote: > > On Apr 5, 2009, at 8:58 PM, Michael Buesch wrote: > >> On Sunday 05 April 2009 20:01:22 Francesco Gringoli wrote: >>> On Mar 30, 2009, at 11:35 PM, Francesco Gringoli wrote: >>> >>>> >>>> On Mar 28, 2009, at 12:41 AM, Michael Buesch wrote: >>>> >>>>> > Hi Michael, > >> >> I think this really is completely getting overengineered by now. >> >> I'm not going to apply such a patch unless you tell me why it's >> needed. >> Having such an incredible mechanism for an absolute corner case that >> happens >> once in a billion frames but doesn't harm anybody is not really >> acceptable. > No problem :-) I simply sent the patch I'm using in my test > environment where I get this behavior for the 0.1% of the received > frames when FCSFAIL is set. Note that here we collect traces for > experiments with 802.11 protocol, so we need this kind of patches. > > I understand that very few of us are doing such kind of experiments > and users are not, I simply sent a comment about these devices. It may > improve knowledge about them. > > Cheers, > -FG >
While I more or less agree with Michael on this, I can also see that these changes may be useful in some environments. Thus, rather than always adding unneeded complexity to the driver, it might be better if changes like these could be placed under a general network define kernel option. Then users wishing to have strict frame checks when FCSFAIL is set have the option to do so. Regards, David Ellingsworth _______________________________________________ Bcm43xx-dev mailing list Bcm43xx-dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
