https://bugzilla.novell.com/show_bug.cgi?id=807377 https://bugzilla.novell.com/show_bug.cgi?id=807377#c1 Paolo Roascio <paolo.roascio@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |paolo.roascio@gmail.com --- Comment #1 from Paolo Roascio <paolo.roascio@gmail.com> 2013-04-28 20:26:18 UTC --- Hello, sorry for my bad english, I had exactly the same problem in openSUSE 12.3 with two Sony OPTIARC AD-7261S: only /dev/sr0 was linked in /dev/disk/by-id. Despite this, kde was able to "use" both disks (if units are empty, information center reports only one device - sr1, but if i load two dvd, the "only one" device reports both disks - like a multitray unit). But worse, k3b only saw sr1... Investigating a bit, i discovered that half problem is due to udisks and half to udev: - undev reports, for my units, same ID_MODEL and ID_SERIAL, then, in udev rules, the latter overvrite the former in symlinking (this is the reason why in /dev/disk/by-id is present only sr1. - udisks2 exports node data according to its order of preference: WWN, SERIAL, PATH. in case of optical units (at least mine), the WWN data is skipped to the SERIAL one, but, being identical, one unit is "skipped" and here too dbus based services can see only one device. Googling around, i found this bug report: https://bugzilla.redhat.com/show_bug.cgi?id=887979 (related to a particular family of SSD units whith identical WWN, but different ID_SERIAL) which pointed me in the right direction. Then i attach a patch which solves both problems: the solution for the SSD is explained in the bugreport and its attached patch (WWN data is expanded with SERIAL data), my (ugly) hack - i'm not a programmer - is: if WWN step is skipped, then process tries the SERIAL step, this is matched only if ID_SERIAL != ID_MODEL, otherwise process tries the PATH step. I don't know if LG devices expose same ID_NAME and ID_SERIAL, but the patch covers both cases. Maybe this is a workaround? Acording to udevadm info -n /dev/sr[0,1], nodes that make them unique devices are - in my case - ID_PATH and MINOR, then, using %n should not be difficult to write a rule to differentiate symlinks in /dev/disk/by-id -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.