The Leading Cause Of The Recent Linux Kernel Power Problems

[phoronix]

Phoronix: The Leading Cause Of The Recent Linux Kernel Power Problems

"Mobile users are urged to seriously consider these results, and possibly even avoid the Natty Narwhal...I hate to say it, especially in an Ubuntu review, but the mobile edge goes to Windows for now...There are also compelling reasons for folks to avoid [Ubuntu 11.04] at all costs. Linux gamers should see substantial improvements, while mobile users suffer a dramatic loss in battery life," were among the critical comments that Tom's Hardware had in their Ubuntu 11.04 review as they were referencing the power regressions I discovered nearly two months ago within the mainline Linux kernel. As I mentioned on Sunday, the Phoronix Test Suite stack and I have now nailed this major power regression in the Linux 2.6.38 kernel that is affecting a significant number of mobile Linux users. Here is what is happening and a way that you should be able to workaround the serious regression should it affect your computer system(s).

http://www.phoronix.com/vr.php?view=16181

[d4ddi0]

Curse you bad BIOSEN!!!

Thanks, Michael!

Oh, and let me take this opportunity to reiterate how much I pine for coreboot support...

[Garp]

Well done, and thank you.

So stupid BIOS inconsistencies cause problems yet again. This is absurd. There has to be a better, more standardised way to do things.

[Xake]

There has to be a better, more standardised way to do things.

Coreboot!:-D

The problem is that there are lots and lots of stadards, but hardware vedors ignores them as long as it "works with current windows". Guess why many MoBo-manufacturers has to update BIOSes when new versions of windows comes out? Because that version needs new things in the BIOS or because the old BIOS did not follow standards in some routines, which newer version of Windows did need to be done right?

Remember a blog-post from a Linux kernel dev that tried to get PM working on a PCI-E card, but never got it to wake up after he told the chips to go down into sleep mode. Did turn out IIRC that the chip in question did support that kind of PM, however the hardware manufacturer for the card did not wire the "wake up" pin to the bus, since "Windows did not support this kind of power management anyway"...

[Nobu]

There are...the issue is that apparently the people who write BIOS's ignore them when it's convenient.

(coreboot ftw! Just wish my board was supported .)

[ChrisXY]

After reading the article I disabled adblock on phoronix.com.
But sorry, I won't disable flashblock. My battery life and the fan speed of my laptop is still more important.
Ironically in your article about minimizing power consumption you also have one of the most power draining advertisement techniques known to humanity...
Please, make sure you have only text or image advertisements. They just make the site slow to load (which is bad enaugh for slow internet connections) but keep away from draining your power on uselessness.

When is BIOS finally going to die? Over the years I have had so many problems with buggy BIOSses it's not funny anymore.
I have two different P4-era boards that can set AHCI mode for s-ata but they can't boot from a disk in AHCI mode. It simply doesn't load the bootloader. So I had to run in IDE compatibility mode for years. In retrospect I could have put /boot on a USB disk but that's not the point... Booting from USB sticks often didn't work at all on many boards that had an option for that in the BIOS for a very long time and it's until now that it finally seems to work on most.
But my eeePC 1000h's BIOS doesn't seem to remember to boot from a USB disk. When rebooting after the BIOS screen I simply get a black screen with nothing on it until I go into the BIOS and set it again to boot from USB disk.
I remember on some mainboard entering the BIOS and then boot messages scroll over the opened BIOS screen.

This are some that come to my mind immediately but there are probably many more.

Please, let it die and replace it with something capable of today's requirements and something that is working.

[smitty3268]

Wasn't there just an article here

about how someone reversed engineered how Windows sent the shutdown signal and put the new logic into the 3.0 kernel?

Someone needs to do the same here, because obviously the MB manufacturers are just hacking things up until they get it to work with Windows.

[BlackStar]

Great work finding the cause of the issue, thanks!

I'm testing pcie_aspm=force on my dm1z laptop.

[kokoko3k]

On my system i've:

Gozer ~ # dmesg|grep -i aspm
[    0.216012] pci 0000:01:00.0: disabling ASPM on pre-1.1 PCIe device.  You can enable it with 'pcie_aspm=force'
[    0.216360] pci 0000:03:00.0: disabling ASPM on pre-1.1 PCIe device.  You can enable it with 'pcie_aspm=force'
[    0.216630] pci 0000:02:00.0: disabling ASPM on pre-1.1 PCIe device.  You can enable it with 'pcie_aspm=force'
[    0.217299] Unable to assume _OSC PCIe control. Disabling ASPM

Should i still try "pcie_aspm=force"?
Mine is a desktop system, but if i could save some power i would be happy and i have no way to measure it.

[ssam]

about how someone reversed engineered how Windows sent the shutdown signal and put the new logic into the 3.0 kernel?

Someone needs to do the same here, because obviously the MB manufacturers are just hacking things up until they get it to work with Windows.

http://mjg59.livejournal.com/137313.html

or we need the manufactures to consider linux when they make the hardware. think this already happens at the server end of hardware. i think you need as many people asking "and does it work with linux" to their supplier, so that it gets pushed up the supply chain.