The Linux Kernel Power Issues Continues To Bite Users
Besides the many of complaints about this kernel power bug within the Phoronix Forums, Ubuntu Forums, Fedora Forum, in the Ubuntu bug report, and elsewhere, Tom's Hardware wrote a review of Ubuntu 11.04 this past week. What did they have to say when it comes to mobile?
Several weeks ago, Michael Larabel of Phoronix reported that Ubuntu 11.04 suffers serious battery life regressions from previous versions of the distribution. His findings were originally specific to the kernel, and he received a lot of feedback on Slashdot and other forums asking whether or not this would manifest in real-world usage.
We not only confirm his findings, but also demonstrate that this issue has very significant implications in actual use. It also appears that the Unity interface is not the root cause. In fact, Ubuntu 11.04 with Unity gets more than 20 minutes additional battery life than Ubuntu 11.04 Classic. The previous release, Ubuntu 10.10 Maverick Meerkat gets over two hours more battery life than Natty Narwhal, regardless of the user interface. This was probably the main deciding factor for Asus when it recently chose to ship Ubuntu 10.10 Maverick Meerkat as opposed to 11.04 Natty Narwhal on the new Eee netbooks
Mobile users are urged to seriously consider these results, and possibly even avoid the Natty Narwhal. On the other hand, Lucid has serious issues with hibernation, as does Maverick, making this a difficult choice for road warriors. I hate to say it, especially in an Ubuntu review, but the mobile edge goes to Windows for now.
Ouch! In the article's conclusion, Adam Overa also adds, "As far as performance is concerned, there are compelling reasons for certain users to upgrade. There are also compelling reasons for folks to avoid this release at all costs. Linux gamers should see substantial improvements, while mobile users suffer a dramatic loss in battery life."
It's just not the Linux 2.6.38 power bugs, but I also illustrated there are even older kernel power regressions still outstanding. As noted earlier this week, the Linux 2.6.39.1 / Linux 3.0 kernels do lower the power usage a bit, but the actual regressions are still un-addressed.
These problems have been hitting many users now in released distributions, yet the problem still isn't resolved. In fact, Canonical, Red Hat, or other ISV/IHVs haven't even inquired about the work I've been doing to narrow down the issue... Which I think I may have tracked down and will announce then in a couple of days. [So, yes, I'll just put out the findings when it's to my maximum ad-revenue advantage.] I also still haven't received any viable recommendations concerning USB-based AC power meters, which would have also sped up the efforts. It's taken me a while due to the work involved and also still needing to push out other Phoronix benchmarks and content, along with Phoronix Test Suite / OpenBenchmarking.org developments, and other work simultaneously.
Stay tuned. I've also been dropping bits of information in my Twitter feed.
33 Comments
