Saturday arrives and WIN comes PING MuyLinux , and knowing how many of you know I like the section, it gives me the feeling of living my day Groundhog particular (xD). Everything is because not miss anything (or at least possible). Come to the agenda …
GNOME-Shell
reason to love Ubuntu 4.12 LTS (could use to list in reverse, but Believe it or not, I’m not Kaffir 
No related posts.
work done by Michael Larabel Phoronix once again, where he analyzed the performance of two dozen graphics cards from AMD to study especially the behavior of different drivers that exist, and special focus on Open Source drivers AMD’s own.
Article first
huge number of pages devoted to a brief description of all cards have been included in the comparison, ranging from models more ancient to art cards and especially powerful, such as AMD Radeon HD 6950.
The findings make clear that with all this choice available cards are for everyone , but mostly affect the issue of drivers.
Arguably the open source driver for Radeon Linux driver is also the best Open Source dedicated graphics card and at least has the support of AMD – not as NVIDIA that does not support Open Source, because the community depends on reverse engineering methods to provide a driver with mixed results. If the driver Open Source works on your system and meets all your requirements, great, but otherwise we still have Catalyst drivers .
That seems to make clear the situation is also true for NVIDIA cards, and they also seem to Phoronix takes out an article. Today proprietary drivers are still the best solution for users who want to take full advantage of these components. Open Source drivers are making progress, but they still have some twists to match the proprietary solutions.
This topic is long and looks set to continue over what we would like to laptop users with the Linux kernel 2.6.38 or higher. At Phoronix still investigating the problem and now requires assistance to affected users.
late April raised the alarm later that was something to take into account portable Phoronix few days ago discovered the problem and WIN at the same link you have the solution , although seems not work for everyone. That is the reason that Phoronix have been put to work.
In an article published yesterday in the linked site is meant what is said at this point, there are still problems with power management and battery life in laptops compared to Windows or earlier versions of GNU / Linux and Ubuntu 10.04.
To investigate the problem thoroughly requested the cooperation of affected users , which should provide the information your team in this thread in the Phoronix Forums. At least the data are requested netbook or notebook model and its CPU / GPU , although the ideal would be to use the Phoronix Test Suite (available for most distributions) and share the results of “phoronix- test-suite system-info “or better yet to” test-suite-phoronix benchmark battery-power-usage “.
see what is all this, but it is really important a situation to solve , so a little baffled from the highest levels of kernel developers still have not dealt with as they should.
have long been talking about the problems of the 2.6.38 kernel, which brought many improvements but also included a “regression” that caused the battery life of portable shorten markedly in specific cases.
Phoronix
-you explain himself confirmed in the comments, using the 2.6.38 kernel does not always cause these problems, but if you have a laptop with Ubuntu 11.04, Fedora 15 or any of new distributions of the market that integrate that Linux kernel, it is quite possible that you have noticed how your battery’s autonomy is reduced considerably.
Fortunately, it appears that there is already a very efficient solution of the problem until the kernel developers for universal solutions. By using the benchmarks of the famous Phoronix Test Suite has been able to uncover the cause of the problem that caused the decrease in battery life.
According to Phoronix and its creator, Michael Larabel, the great cause of the problem of energy consumption in the 2.6.38 kernel is the change that took place in the so-called ASPM (Active- Power State Management ) for PCI Express. Specifically, the change is documented in the commit 2f671e2dbff6eb5ef4e2600adbec550c13b8fe72 coded under the title “PCI: disable ASPM, the BIOS asks us.”
Phoronix graphs make clear that the improvement in battery life is remarkable.
feature Active-State Power Management should save energy by setting a power saving state in PCI Express links are not being used. The disadvantage is that ASPM may increase the latency of the devices, but the energy savings more than offsets the latency. Although it can run on desktops, the feature is specially designed for laptops, but if the BIOS indicates that does not support ASPM, that behavior is modified .
And here we arrive at the cause of the problem: misconfigured BIOS , since many laptop manufacturers support ASPM but not configured correctly in the Fixed ACPI Description Table call, which is that “autoconfigure “the BIOS during startup.
What is the solution? Simple: add a line to our boot to force the use of ASPM
pcie_aspm = force
This enables the feature
whatever it says the BIOS, and the evidence shows that this solution energy savings are significant that increase battery life so sensitive. Congratulations to Phoronix for this great work is sure to brighten the lives of many users. By the way, you can thank them with a donation via PayPal are sure to come in handy.
testing on Linux for comparing systems was a complex task, but doing so comparing Linux to Mac OS X or Windows was almost impossible to consistently. Phoronix Test Suite just solved both deficiencies with some features that are still growing release after release, and it shows the notice of Phoronix Test Suite 3.2 “Grimstad ‘.
This is the latest edition of a benchmarking suite that allows evaluate a large number of parameters of our Linux distribution that comes with some interesting improvements.
Although the official announcement did not specify what changes have been too, we can see these improvements in the version history available at the official development page, which indicates that PTS have the following improvements 3.2 :
Support for new sensors
Enhanced
Thu Since version 3.0 which appeared in February this suite is directly connected to two parallel projects, and OpenBenchmarking.org Phoromatic, allowing verify the results of the tests that have been published Others and make comprehensive studies on the results that are different configurations of hardware and software.
Unfortunately one of the most desirable features of the suite is still not available. This is the graphical user interface that does exist for Phoronix Test Suite 2.x development through a python, but that is not available for PTS 3.x, forcing us to use the suite from terminal , something that obviously “scared” to many new users.
Hopefully soon address this deficiency of which is the undisputed test suite in the Open Source segment.