cd-rom drive with data media not showing the volume

David Zeuthen david at
Mon Dec 6 14:41:18 PST 2004

On Tue, 2004-12-07 at 10:57 +1300, leon breedt wrote:
> On Mon, 06 Dec 2004 10:50:26 -0500, David Zeuthen <david at> wrote:
> > If we don't poll the drive, e.g. storage.media_check_enabled is set to
> > FALSE, we only add volumes when something is mounted for the drive. That
> > is on purpose.
> media_check_enabled is set to FALSE for these devices on the
> non-working machine, however, /etc/hal/hald.conf has a default value
> of TRUE, and i can't find a .fdi policy that sets it to FALSE, so I'm
> not sure how this value gets set.
> Just to be sure, I removed
> /usr/share/hal/fdi/20freedesktop/ide-drives.fdi, but the same symptoms
> occurred.
> I also started up hald in verbose mode, and first thought I found the
> problem when I saw these messages:
> 10:54:45.451 [I] linux/block_class_device.c:1385: Bus type is ide!
> 10:54:45.451 [I] linux/block_class_device.c:1409: Working around
> broken ide-cs driver for /org/freedesktop/Hal/devices/ide_1_1
> 10:54:45.464 [I] linux/block_class_device.c:339: profile -1

OK, your IDE channels appears to be on these sysfs paths


which is strange; normally it would appear as a child of a PCI device,
e.g. on your second box it is at


and this is what triggers what is apparently a bug in hal. 

So, what we do today is to say "if the IDE device is not a child of a
PCI device, then it is definitely something we can't poll on". Instead,
the logic we apply is that we should poll on IDE, if, and only if, the
device is an optical drive. And we can check that.

Please try the attached patch (against the hal 0.4.x branch in CVS;
should apply to 0.4.2 as well) to see if that makes it better? 

I will go ahead and test that it works with other IDE devices such as
Zip drives and Memory card PCMCIA readers.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: hal-fix-cdrom-detection.patch
Type: text/x-patch
Size: 4079 bytes
Desc: not available
Url :
-------------- next part --------------
hal mailing list
hal at

More information about the Hal mailing list