From mboxrd@z Thu Jan 1 00:00:00 1970 From: Greg KH Date: Mon, 08 Nov 2004 22:12:31 +0000 Subject: Re: udev-043 Message-Id: <20041108221230.GA17867@kroah.com> List-Id: References: <200411070449.21228.warpy@gmx.de> In-Reply-To: <200411070449.21228.warpy@gmx.de> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-hotplug@vger.kernel.org On Mon, Nov 08, 2004 at 11:02:30PM +0100, Kay Sievers wrote: > On Mon, 2004-11-08 at 13:18 -0800, Greg KH wrote: > > On Mon, Nov 08, 2004 at 10:00:54PM +0100, Kay Sievers wrote: > > > On Sun, 2004-11-07 at 04:49 +0100, Michael Geithe wrote: > > > > Hi, > > > > > > > > only one udev-message found in my logs ;-) > > > > > > > > wait_for_sysfs[10013]: either wait_for_sysfs (udev 043) needs an update to > > > > handle the device '/class/i2c-dev/i2c-2' properly (no bus device link) or the > > > > sysfs-support of your device's driver needs to be fixed, please report to > > > > > > > > > > > > > > class: /sys/class/i2c-dev/i2c-2/ > > > device: /sys/devices/platform/i2c-2 > > > bus: /sys/bus/i2c/devices/2-0290 > > > > Where do you get this "bus:" info from? > > I extracted it from Michael's post: > > > > ls -l /sys/bus/i2c/devices > ... > > 7. Nov 03:53 2-0290 -> ../../../devices/platform/i2c-2/2-0290 > > > > > > Hmm, looks like the "device" link should point to: > > > > > > /sys/devices/platform/i2c-2/2-0290 > > > > > > instead, to match the bus link. > > > > No, the i2c-2 class device points to the i2c-2 adapter device. Now, > > that adapter device can "find" a number of individual i2c chip devices > > that i2c chip drivers can then bind to. Those chip devices are the > > "2-0290" and such devices. > > > > Make more sense? Yeah, it's a mess, sorry, but the i2c code isn't the > > nicest either :) > > I know nothing about i2c. :) > > Hmm, so libsysfs should look for the "bus" at the parent of > the /sys/devices/ device to get the "bus" of _this_ device? Ah, so the "problem" is that there is no bus assigned to this device, right? Yeah, I'll agree with that, that's the way the kernel code works for this device, and for others (like scsi-host "devices" right?) Hm, I think no bus is correct for this case, as that's what really is happening. > Or should there be one /sys/bus/devices link for every "chip" created by > the kernel? There should be one in the /sys/bus/i2c/devices directory for every "i2c chip", but there is nothing in /sys/bus/* for any i2c adapter devices (think of them as the same way scsi host devices are.) Does that help? thanks, greg k-h ------------------------------------------------------- This SF.Net email is sponsored by: Sybase ASE Linux Express Edition - download now for FREE LinuxWorld Reader's Choice Award Winner for best database on Linux. http://ads.osdn.com/?ad_idU88&alloc_id065&op=click _______________________________________________ Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net Linux-hotplug-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel