Evo N410c (BIOS F.0E A) APM screen blanking problem under Linux

[Hugo Haas]

[ This is a repost of something that I have posted in the Evo notebook forum since it is Linux-specific. I haven't figured out cross-posting. ]

Hi.

My experience with Linux on the Evo N410c is that power management isn't perfect.

ACPI doesn't work (fan not detected), APM is flaky, etc.

I have documented my experience at:

http://larve.net/people/hugo/2002/12/evo410

I have updated the BIOS to F.0E A. It fixed some of my problems, but another one which is pretty annoying appeared.

When running on batteries, after a few minutes the screen goes blank in order to save power. I am running 2.4.20.

When a hit a key to bring the screen display back, I can see things but then the keyboard locks up. The system still runs but the keyboard cannot be used anymore.

With the AC adapter plugged in, the screen, when blanked, isn't turned off, and this problem doesn't occur.

My questions are:
- has anybody seen this problem too?
- is there a way to prevent screen blanking? setterm and xset didn't do anything; it seems to be a BIOS thing.
- is there a workaround?

Regards,

Hugo

[Bruce Copeland]

I'll try to answer as many of your questions as possible. I don't have an evo N410, but I do have a relatively recent notebook with similar hardware and ACPI. Almost everything you describe has been seen previously by someone else (not necessarily on an evo). But the situation is a bit more complex than that. The areas where you are having difficulty are affected by specific notebook hardware, bios, kernel version, kernel configuration, ACPI, APM, graphics hardware, and probably Linux distribution.

Some people DO in fact run both ACPI and APM. However, most recommendations for ACPI say to turn off APM in the kernel config. Second, most people who are serious about trying to run ACPI apply one of the appropriate ACPI kernel patches from the ACPI project at SourceForge. Third, ACPI does not have a universal set of properties. There is code support for a lot of ACPI features in newer kernels (especially with the ACPI patch), but ACPI availability, features, and /proc/acpi directory, filenames, and file information ALL depend on the bios for a particular computer. Suspending introduces further variability. There is another kernel patch for software suspend available from SourceForge. Some people use it; some people don't. It works for some people; for others not at all.

I suggest you try building a new 2.4.20 kernel WITH the ACPI patch and with APM DISABLED!. See how that works for you, remembering that you might need to go in and tweak some ACPI scripts or code to adjust /proc/acpi directory and filenames, etc. If you get something that is an improvement over your current kernel, THEN you might want to go on and experiment with software suspend.

Good luck! You can find some further guidance about the 2.4.20 kernel patches, kernel configuration, ACPI, and related issues at my ze4100 site:

http://www.cybersym.com/pages/linux-ze4100.html

Recognize, however, that what I do is on different hardware.

Bruce

[Georg Luebeck · Accepted solution]

The keyboard lockup after the display blank also occurs on the N610c. There is a fix for this using the mouse_uf diver.

Check my posting in response to the query by Stephan Merz 02/07/03 in the Linux forum

http://forums.itrc.hp.com/cm/QuestionAnswer/1,,0xd42e31ec5e34d711abdc0090277a778c,00.html

Hope that helps.

Georg

[Hugo Haas]

Thanks Georg. The mouse_uf, even if it is an ugly hack, does fix the problem!

In the meantime, since I couldn't find the solution to the APM problem until now, I had started investigating ACPI a bit further as Bruce suggested. Making it work isn't trivial, but I did manage to get it functional, though not perfectly:

http://larve.net/people/hugo/2003/03/acpi-n410c.html

I think that I am going to switch back to APM though since it does work better now, but I thought I would post my ACPI notes here.

Regards,

Hugo

[Kirill_7]

I am currenty battling with a RedHat Power Management problem that I might be able to fix using the 'mouse_uf' fix that you mentioned here. Unfortunately this utility is no longer available on the web, so it would be a great help for me if you can send it to me if you still have a copy.

Many thanks

Kirill