Can we have a patch and a use flag for enable smart status in gnome-disk-utility?
how do you think we can do that if it doesn't work out of the box ?
(In reply to comment #0) > Can we have a patch and a use flag for enable smart status in > gnome-disk-utility? > What patch and what configure option is needed? On my system, I am able to see smart status from palimpsest: 1. Launch it 2. Go to my harddisk 3. Click on "SMART data" -> A window is opened showing the information
My hd support smart and smart is active but palimpsest doesn't work. I have found this searching on google https://bugs.launchpad.net/ubuntu/+source/devicekit-disks/+bug/553282
What sys-fs/udisks versions do you have? Please try with latest one also
sys-fs/udisks-1.0.1-r1 I will try to unmask 1.0.2 version.
Same situation with sys-fs/udisks-1.0.2
This is probably the same race conditions I have seen. This is partly fixed in latter udev. To confirm, try run the following in a terminal as root: grep dbus /etc/init.d/udev-postmount && killall udisks-daemon && echo OK! If you are left with an "OK!", then try run palimsest again and check your SMART status. If this does not return "OK!" the try re-emegre udev, reboot and run that command again. I have had the same behaviour because of two races: 1. udev-postmount (i.e. udevadm trigger --type=failed) being run before dbus is running (creating problems with apps like udisks that needs dbus). This has been fixed in udev. 2. login before udev-postmount has been launched. This I am currently hit by. gnome-desktop or gvfs seems to start udisks-daemon before udev-postmount is done, making udisks-daemon use old versions of /udev/.udev/data* which has not been updated with SMART status. The workaround is "killall udisks-daemon", the long term workaround/fix is probably make xdm depend on udev-postmount. The real fix I guess would be to make udisks-daemon re-read the info from /dev/.udev/data if it has been modified.
I found another problem if you have a seperate /usr (needs to enable /dev/.udev/udev.log): 1303304475.565758 [2638] util_run_program: '/lib64/udev/udisks-probe-ata-smart' (stderr) '/lib64/udev/udisks-probe-ata-smart: error while loading shared libraries: libatasmart.so.4: cannot open shared object file: No such file or directory' For some reason this is not picked up by "udevadm trigger --attr-match=failed" later in udev-postmount, I do not really know why or if it even should. This is currently the only way for me to get gnome-disk-utilities to recognize SMART support (apart from installing libatasmart into /lib): udevadm trigger && killall udisks-daemon This may crash Xorg, and eat other babies, but afterwards you can see your smart status in palimsest without trouble.
(In reply to comment #8) > I found another problem if you have a seperate /usr > > udevadm trigger && killall udisks-daemon I've run into this as well with a separate user (although it did work previously but stopped recently). Safer workaround I found posted that works for me (doesn't crash anything) is: # echo change > /sys/block/sdx/uevent # udisks --ata-smart-refresh /dev/sdx Substituting sdx for the proper device names (sda, sdb, etc.).
(In reply to comment #9) > I've run into this as well with a separate user Should read: I've run into this as well with a separate /usr
Separate /usr or /var is broken configuration without initramfs that mounts them before init. They are not supported.
So if the only problems here are caused by invalid partition scheme, propably duplicate of bug 364235.
(In reply to comment #11) > Separate /usr or /var is broken configuration without initramfs that mounts > them before init. They are not supported. Been running Gentoo for ages now and this has never been a problem in the past. Is this something new? Where is it documented? Thanks.
(In reply to comment #13) > (In reply to comment #11) > > Separate /usr or /var is broken configuration without initramfs that mounts > > them before init. They are not supported. > > Been running Gentoo for ages now and this has never been a problem in the past. > Is this something new? Where is it documented? Thanks. udev stopped supporting them, documented nowhere else than in bug 364235 yet, far as I know
Also http://archives.gentoo.org/gentoo-dev/ has several threads about it.
Closing as per Comment #11.