|
Bugzilla – Full Text Bug Listing |
| Summary: | udev rule is not written correctly | ||
|---|---|---|---|
| Product: | [openSUSE] SUSE LINUX 10.0 | Reporter: | Matthias Fruehauf <mfrueh> |
| Component: | Hotplug | Assignee: | Kay Sievers <kasievers> |
| Status: | RESOLVED FIXED | QA Contact: | Klaus Kämpf <kkaempf> |
| Severity: | Normal | ||
| Priority: | P5 - None | CC: | hare, snwint |
| Version: | RC 4 | ||
| Target Milestone: | --- | ||
| Hardware: | Other | ||
| OS: | All | ||
| Whiteboard: | |||
| Found By: | Other | Services Priority: | |
| Business Priority: | Blocker: | --- | |
| Marketing QA Status: | --- | IT Deployment: | --- |
| Attachments: | y2log | ||
Hmm, did we get this wrong in 10.0 or does PPC need a different rule ? IDE is connected to the macio bus, maybe thats the reason why it failed. worked ok on pegasos with VIA ide controller. I get the content for ENV{ID_PATH} from hwinfo, so please attach y2log files and
we should be able to see where the missing "*" vanished.
Created attachment 50028 [details]
y2log
y2log provided Hwinfo does indeed deliver the strange name in the installation environment the dev_names are: "dev_names":["/dev/hdc", "/dev/disk/by-path/-ide-0:0"]. Later in the installed system dev_names are probably more correct: "dev_names":["/dev/hdc", "/dev/disk/by-path/mac-io_ata-3-ide-0:0"] The names come from udev. Hmm, /sbin/path_id seems not to work correctly during installation. I have absolutely no idea how to debug this without physical access to such a box. cowberry.suse.de is in the inst-sys, pwd is 'root'. Maybe it uses some tools that's missing? Nice service Olaf, thanks! Correct it misses tools. path_id for whatever reason contains this: PATH=/lib/klibc/bin #/usr:/usr/sbin:/usr/bin:/sbin:/bin and sed fails. bug in installation-images, klibc-utils not fully provided. ??? What do you need them for? There are the real ones. as you can see above, the "real" ones are not used. Well, they should, then. What's the point if that strange PATH setting? to not use the fat one by accident. at one point I had hope to get rid of glibc, but today we add even more shit to initrd... klibc specific path removed |
After a fresh installation of RC4 on a PowerBook, the udev rules for cdrom looks like the following: # cdrom links generated by YaST2 # SUBSYSTEM=="block", ENV{ID_PATH}=="-ide-0:0", SYSFS{removable}=="1", SYMLINK+="cdrecorder cdrom dvd" According to zoz it should look like this one: # cdrom links generated by YaST2 # SUBSYSTEM=="block", ENV{ID_PATH}=="*-ide-0:0", SYSFS{removable}=="1", SYMLINK+="cdrecorder cdrom dvd" Difference is the star in the ID path. Any other info you need?