RESEARCHUT

The power consumption on the Linux kernel 3.10 is pretty bad.

On kernel 3.10, with the follwing config, the PowerTop results are:

#

# Timers subsystem

#

CONFIG_TICK_ONESHOT=y

CONFIG_NO_HZ_COMMON=y

# CONFIG_HZ_PERIODIC is not set

CONFIG_NO_HZ_IDLE=y

# CONFIG_NO_HZ_FULL is not set

CONFIG_NO_HZ=y

CONFIG_HIGH_RES_TIMERS=y

PowerTOP v2.0 Overview Idle stats Frequency stats Device stats Tunables

The battery reports a discharge rate of 28.0 W

The estimated remaining time is 23 minutes

Summary: 1785.5 wakeups/second, 0.0 GPU ops/second, 0.0 VFS ops/sec and 22.1% CPU use

As someone who learned computers on his own, I always acknowledged the value that the Free Software movement has brought. The accessibility of these topics, which are only supposed to be part of text books and schools, is available for anyone and everyone who has the resource and passion to do it.

But this past week, 2 things made me pretty impressed with the maturity and quality of work that we do.

[rrs@learner ~] tar -axvf researchut-1.2.0.tar.bz2 && ls

It is still 2012 in this part of the world and the world is still intact. Since nothing major happened, I thought of spending the new gifted time to add a long pending item to apt-offline. As shown in the screen shots, apt- offline’s GUI now has support to detect and display the downloaded offline bug reports.

This is part of the just released, version 1.3.

I recently switched to a ThinkPad W530 laptop. It is a fairly recent machine with the following hardware config:

• Intel(R) Core(TM) i7-3720QM CPU @ 2.60GHz
• 8 GiB RAM
• nVIDIA Optimus
• Samsung SSD Drive

On the software front, I decided to take my chances. Hence:

Hello World. This is the follow-up to the last [2](http://blog/report-bugs- with-apport) updates on the state of apport in Debian.

A lot has changed since the last update on Apport. Currently, in Experimental, we have version 2.6.1-2. With this version, and going forward, there will be no hacks to make it work for Debian. Thanks to Martin Pitt, with his assistance, Apport now has a very basic crashdb in place for Debian. The Debian crashdb provides Apport the interface to interact with the Debian BTS.

This is a follow up to the previous post regarding Reporting bugs with Apport

Apport, version 2.2.3-2, has been pushed to experimental. There were some nice feedback that led to some more changes that I will talk here.

**Opt out: **As a developer, if you see the volume of reports to be annoying, you have the option to opt out of apport reports. To do this, you should specify the XBS-Apport: No field in your package ’s control file. When your application crashes, apport first checks for that field in your package’s description and acts based on what you chose.

So yet another bug reporting tool. :-)

When I prepared Apport for Debian, I wasn’t sure how it will look going forward. If you look at the current one lying in experimental, it just detects your crashes and pops up in your systray. It doesn’t have the mechanism to interact with BTS.

So, like the title of this post says, this is the first worthy feature for Apport in the context of Debian.

Laptop Mode Tools, version 1.61, has been released and will land up soon for Debian. This is the version that would be targetting Wheezy.
This release includes many bug fixes and should make power savings much better on your machines.

This is mainly a bug fix release. Some parallel module execution approach has been used which could show runtime improvements.

Changelog:

1.61 - Thu May 17 17:44:26 IST 2012
* Handle devices with persistent device naming. This fixes the issues where
you don’t have a disk referenced by a block name, the commit= value was
completely skipped
* Fix issue where hdparm skips SSDs for power management
* Add parallel execution for the modules. In theory this should speeden up the
execution. See git commit log comments for details
* Add support for non-deafult customized settings
* calculate design_capacity_warning on machines/arches where it is not readily
available

Now these are the kind of moves that needs to happen more often. Whether this will cause a negative impact on the overall market, and the further invention of drugs (including patent control), but the affordability of the medicatoin to an average citizen is a great move.

The typical Chemotherapy can be on an average of 22 times. When summed up with the dosage (somewhere around 250 mg IIRC), the cost comes to approx: (22/4) * 15000 = 82k, which now, will be affordable at 27k.