ACPI resume broken on 2.6.28 (82801 Centrino laptop)?

Joseph K.

Not that suspend and resume have always worked perfectly, but they have been for me since 2.6.27. After upgrading to 2.6.28, it appears to suspend fine but resume results in a frozen system (requiring a power off). Anyone else noticed this?

The laptop is an HP nx8220, which has the Intel 82801 Centrino chipset and an ATI Radeon card. Just curious to know if anyone else is having issues (or I just screwed up somewhere).

Thanks, cheers.

szczerb · Posted: Mon Feb 02, 2009 10:47 am Post subject:

Did you try without X running? A lot of suspend/resume issues are caused by graphic drivers (especially the nvidia blob).

(I have an Lenovo 300 N200 with ICH8 (82801[H/HEM/HBM]), GM965, 3945ABG, BCM5906M and both suspend and hibernation work just great and with fbsplash on gentoo-sources-2.6.28-r1)

Joseph K. · Posted: Tue Feb 03, 2009 6:29 am Post subject:

No, I haven't tried without X running, but that would generally defeat the purpose for my liking, and as I said, it works perfectly in 2.6.27.

Just for the record, this is an older chipset, ICH6 from memory.

szczerb · Posted: Tue Feb 03, 2009 7:27 am Post subject:

Please think for a minute here. I am asking you to try X to determine if the problem is about your X video driver. I'm not telling you to quit X each time you want to suspend. That would indeed, be quite idiotic in it's purpose.

165177 · Apprentice Joined: 27 Apr 2007 Posts: 156

I don' think, this issue relates to a broken video driver. My laptop shows the same failure with 2.6.28, while working perfectly with 2.6.27 and 2.6.26 (despite the nvidia binary). Looks like a kernel regression in some patch on the way to 2.6.28.

I'd recommend to stick with 2.6.27 and try the next revision of 2.6.28 or just wait for 2.6.29.

szczerb · Posted: Tue Feb 03, 2009 9:51 am Post subject:

On the other hand, I don't have any problems like this and you guys, probably both using binary blobs for Xorg, have problems with 2.6.28.

Joseph K. · Posted: Tue Feb 03, 2009 12:33 pm Post subject:

Well, just to settle the matter, I tried 2.6.28 without X, and with the video driver unloaded, and still the same problem. So let's move past the binary blob argument unless there is something that actually points to it.

lunar, what hardware do you have? I will open a report on bugs.gentoo.org and it would be good if you could add your case.

165177 · Apprentice Joined: 27 Apr 2007 Posts: 156

Joseph K. · Posted: Wed Feb 04, 2009 4:03 am Post subject:

Thanks, it's bug #257487. I assume that you have used lspci before? That would be sufficient.

Joseph K. · Posted: Thu Feb 12, 2009 11:51 am Post subject:

I've just recently noticed that, after running for a few days and going through half a dozen suspend-resume cycles, my laptop fails to suspend on 2.6.27-r8.

This is obviously different to the bug experienced in 2.6.28* and in fact the machine returns to a normal state after a short delay (but it never suspends). Am going to try with 2.6.27-r7, which is the last version that I recall having no problems at all with.