Alan Stern wrote: > On Mon, 2 Oct 2006, Mike Panetta wrote: > > > I have written a patch based on the companion patch that will allow me > > to disable arbitrary ports on my EHCI root hubs for a product we are > > working on (The system has a very naughty built on compact flash card > > reader that cannot be disabled in bios) and I want to make sure this is > > the best way to do it. > > This is very unclear. Why do you want to disable the ports? Why not > simply fail to enable them in the first place? Mainly because I do not know how to do that? Is there a way to tell the EHCI driver to not enable certain root hub ports? > > If you don't want to use the built-in CF card reader, why not unplug (or > cut!) the cable connecting it to the EHCI controller?
Because its hardwired on the circuit board. I suppose we could reverse engineer the motherboard (its a VIA mini itx board) to do this, but doing it in software is a much preferred solution as we can RMA the boards back to VIA if something goes wrong with them. > > > Also, it seems to have one simple issue. If > > there is something plugged into the port, and I disable it, khubd never > > seems to notice the event, so it will only see the port is disabled > > after I try to access the device. Is there any way to fake a port > > disconnect event so that I can get it to rescan the ports? > > Wouldn't that defeat the whole purpose? If you create a disconnect event > then khubd will rescan the port and enable it all over again. My patch makes the port connect events go unnoticed if the port is 'disabled'. > > Or put it another why: If you do want the port to be enabled so you can > access it, then why disable it in the first place? Because we are using 2 different motherboards and one does not have the bad hardware issues. We are trying to release a product and not have to worry about there being 2 different kernel versions to support (one with the ports enabled and one with the ports disabled in the driver). If there is another way to enable/disable ports in software that I do not know about then that should work fine. > > > The same > > applies to me re-enabling a port that already has something pluged > into it. > > (Redundant: It's impossible to enable a port that doesn't have something > plugged into it.) True. That was a braino :P > > Why do you want to re-enable a port that has something plugged into it? > Why hasn't the port been enabled all along? Or if you disabled it for > some reason, why have you changed your mind? See above. > > > I guess basically this patch seems to work, except I need to figure out > > a way to fake port disconnect/connect events. Is this possible? The > > reason I would need this is all of our USB devices are pretty much > > hardwired and always exist, so there is only one plug event at boot time. > > What are you really trying to accomplish? What's the reason for going > through all this rigmarole? See above :) > > Alan Stern > Thanks, Mike Panetta ------------------------------------------------------------------------- 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 _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
