Red Hat Bugzilla – Bug 207206
need to sync before hibernate?
Last modified: 2007-11-30 17:11:43 EST
For some reason hibernate stopped working for my Sony Vaio VGN-TR3A laptop between FC5 and FC6test3. It goes most of the way down but then just sits there without actually shutting down. I have to forcibly poweroff to do anything at this point. The problem I'm reporting here though is that when I reboot, I have so far every time got reports of orphaned inodes. This leads me to wonder -- is hibernate on FC6 not running a sync to flush buffers to disk before swapping out the contents of RAM? (I would like to report the hibernate problem separately, but this bug is not about that, it is about the sync issue. I know there are lots of issues with different hardware having problems with hibernate. What is the protocol for debugging and reporting these?) Thanks!
Can you attach the output of dmesg ? I wonder if your orphan inode problem is related to us not doing the right thing with HPA on resume on some boxes.
Unfortunately I can't boot into FC6 right now because of bug 207208. Will try to fix the booting issue soon so I can provide the dmesg output.
Sorry, that last comment was probably confusing... FC6test3 trashed my grub, and I ended up re-installing XP Pro. I now need to get Fedora up and running again. Will try installing FC6 final to see if there's even an issue there.
Created attachment 140083 [details] Output of dmesg
So I now have a fresh install of FC6 final on the same laptop, and it hibernates and suspends fine, and resumes from hibernate fine, but it doesn't resume from suspend properly. It looks like the video card is not getting initialized correctly. I will file this as a separate bug; however the problem of not hibernating completely seems fixed in FC6 final, so I can't test whether the orphaned inodes problem remains. Is the dmesg output helpful?
To get video back after suspend to ram, you probably need to run a userspace utility. Do you know what chipset you have?
i855GM. The video resume from suspend worked fine in FC6test3, not sure why I would now need a userspace utility. Is there one I should be using though? Bug 213593 shows screenshots of the problem. The screen gradually turns white, and has a weird uneven grey border around it. I've seen this on multiple chipsets before (at least nVidia as well as Intel), so I think it's actually an xorg thing.
Sorry for the glacially slow reply. 915resolution is probably the tool you're after.
Closing due to lack of further information.