Bug 151756

Summary:	hal thinks /dev/card1 is an alsa sequencer device
Product:	[openSUSE] SUSE Linux 10.1	Reporter:	Ludwig Nussel <lnussel>
Component:	Basesystem	Assignee:	Danny Kukawka <dkukawka>
Status:	RESOLVED WORKSFORME	QA Contact:	E-mail List <qa-bugs>
Severity:	Normal
Priority:	P5 - None
Version:	Beta 3
Target Milestone:	---
Hardware:	Other
OS:	Other
Whiteboard:
Found By:	Other	Services Priority:
Business Priority:		Blocker:	---
Marketing QA Status:	---	IT Deployment:	---
Attachments:	/var/log/messages

Description Ludwig Nussel 2006-02-17 10:59:00 UTC

hal thinks /dev/card1 is an alsa sequencer device instead of /dev/snd/seq. Leads to resmgrd installing acls on /dev/card1. See lshal output: https://bugzilla.novell.com/attachment.cgi?id=68991&action=view

[0] root@slimer:~ # tree /sys/class/drm
/sys/class/drm
|-- card0
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:02.0
|   |-- dri_library_name
|   `-- uevent
|-- card1
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:02.1
|   |-- dri_library_name
|   `-- uevent
`-- version

4 directories, 7 files
[0] root@slimer:~ # l /dev/card* /dev/dri
crw-rw----  1 root root 226, 0 2006-02-17 11:29 /dev/card0
crw-rw----+ 1 root root 226, 1 2006-02-17 11:29 /dev/card1

/dev/dri:
insgesamt 0
drwxr-xr-x  2 root root      60 2006-02-17 11:29 ./
drwxr-xr-x 12 root root    8020 2006-02-17 11:29 ../
crw-rw----  1 root video 226, 0 2006-02-17 11:29 card0

Comment 1 Danny Al-Gaaf 2006-02-17 11:13:15 UTC

please attach output of "cd /sys/class/sound ; grep . * -R"

Comment 2 Danny Al-Gaaf 2006-02-17 11:17:38 UTC

Btw. the info about the alsa.device_file and linux.sysfs_path came both from udev. And the sysfspath is '/sys/class/sound/seq'. This is IMO a sequencer device.

Maybe something in udev or the kernel is wrong?

Comment 3 Ludwig Nussel 2006-02-17 11:27:48 UTC

The recursive grep is insane. There are symlinks forming loops ...

Here is the tree output:

.
|-- adsp
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- audio
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- controlC0
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- dsp
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- mixer
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- pcmC0D0c
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- pcmC0D0p
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- pcmC0D1c
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- pcmC0D2c
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- pcmC0D3c
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- pcmC0D4p
|   |-- dev
|   |-- device -> ../../../devices/pci0000:00/0000:00:1f.5
|   `-- uevent
|-- seq
|   |-- dev
|   `-- uevent
`-- timer
    |-- dev
    `-- uevent

24 directories, 26 files
# cat seq/dev 
116:10
# l /dev/snd/seq 
crw------- 1 joesix audio 116, 10 2006-02-17 11:29 /dev/snd/seq

udev knows about snd/seq but not about the dri devices:

[0] root@slimer:~ # udevinfo -q all -n snd/seq
P: /class/sound/seq
N: snd/seq
[0] root@slimer:~ # udevinfo -q all -n card1
no record for 'card1' in database
[3] root@slimer:~ # udevinfo -q all -n card0
no record for 'card0' in database
[3] root@slimer:~ # udevinfo -q all -n dri/card0
no record for 'dri/card0' in database

Comment 4 Danny Al-Gaaf 2006-02-17 11:44:57 UTC

I can't do anything here, because we get this infos in HAL from outside and this is IMO udev. Could you restart hal with --verbose --use-syslog and attach the hal output in /var/log/messages?

Comment 5 Ludwig Nussel 2006-02-17 12:51:59 UTC

Created attachment 69040 [details]
/var/log/messages

Comment 6 Danny Al-Gaaf 2006-02-17 13:19:50 UTC

If you restart hal is the output of lshal the same? Is this maybe fixed then? If not, I need a strace of the verbose hal (with -ff -s1024 -o /tmp/hal/strace)

Comment 7 Kay Sievers 2006-02-17 14:08:42 UTC

Please add the patch, that clears the event buffer for the udev event, that is in CVS. But that problem should not show up with hal coldplug, only if devices are created by hotplug events.

Comment 8 Danny Al-Gaaf 2006-02-20 14:45:58 UTC

added patch. 

@ludwig: need a comment to #6

Comment 9 Ludwig Nussel 2006-02-20 15:15:19 UTC

ah, sorry. I upgraded to beta4 meanwhile and now I can't reproduce it anymore. Neither after boot nor after rchal restart.