Date: Sat, 17 Mar 2012 17:30:46 +0100 From: Hans Petter Selasky <hselasky@c2i.net> To: freebsd-usb@freebsd.org, holm@freibergnet.de Subject: Re: USB UHCI Problems on 8-Stable Message-ID: <201203171730.46127.hselasky@c2i.net> In-Reply-To: <20120317085853.GA53895@beast.freibergnet.de> References: <20120317085853.GA53895@beast.freibergnet.de>
next in thread | previous in thread | raw e-mail | index | archive | help
On Saturday 17 March 2012 09:58:53 Holm Tiffe wrote: > >Hi, > > > >The BULK transfer stops by a stall. I suspect the UHCI is receiving more > >data than it can, and is hardware programmed to stop it looks like. I > >can't change that behaviour. It is not a bug in the FreeBSD USB driver I > >think. > > > >Try this patch to /usr/ports/devel/openocd/ > > > >make extract patch > > > >ee ./work/openocd-0.5.0/src/jtag/drivers/jlink.c > > Uhh, may I use Vi? :-) > > >Change: > > > >result = jlink_usb_read(jlink, in_length); > > > >Into: > > > >/* must read at least one packet at a time! */ > >result = jlink_usb_read(jlink, in_length + ((-in_length) & 63)); > > > >Then re-compile and try again! > > > >--HPS > > Ok, I'll try that patch(~ sunday eavening, I'm out of town currntly), > but it adresses the problem with openocd. ( which is good anyways). > I is still a bug in FreeBSD's driver that there is now way to > wake up the controller again and I think that should be changed. > (please remember, most Kerenels are running with compiled in drivers, so > thie singe solution is a reboot to reactivate the controller. > I dont know that much about USB, but I sure that FreeBSDs behavior in this > case isn't that what we really want. I think usbconfig -d ugen1.1 reset > schould be made working.. I will think about it. Currently a suspend and resume will do that, though it might be an idea to allow that runtime using usbconfig. --HPS
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201203171730.46127.hselasky>