Prima che con il tempo mi dimentichi tutto, scrivo questo piccolissimo how-to sulla mia configurazione per lo scaling della cpu.

La configurazione prevede questo: quando accendo il portatilino, se l’alimentazione è presente viene settata la politica di performance, altrimenti powersave.

Inoltre durante l’uso del pc, se viene scollegato o ricollegato l’alimentatore, viene automaticamente settata la relativa politica.

1. yaourt -S acpid pm-utils

2. [~] grep DAEMONS /etc/rc.conf
   DAEMONS=(acpid !cpufreq ... )

   notate che ho disabilitato il demone cpufreq perchè quando uso l’alimentatore preferisco usare la politica performance, ma se usate la ondemand potete abilitarlo.

3. [~] cat /etc/conf.d/cpufreq
   #configuration for cpufreq control
   
   # valid governors:
   #  ondemand, performance, powersave,
   #  conservative, userspace
   governor="ondemand"
   min_freq="800MHz"
   max_freq="1.60GHz"

4. Modificate la sezione ac_adapter) del file /etc/acpi/handler.sh come segue:

   ac_adapter)
           case "$2" in
               ACAD)
                   case "$4" in
                       00000000)
                           echo "conservative" >/sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
                           echo "conservative" >/sys/devices/system/cpu/cpu1/cpufreq/scaling_governor
                           echo -n $minspeed >$setspeed
                           #/etc/laptop-mode/laptop-mode start
                       ;;
                       00000001)
                           echo "performance" >/sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
                           echo "performance" >/sys/devices/system/cpu/cpu1/cpufreq/scaling_governor
                           echo -n $maxspeed >$setspeed
                           #/etc/laptop-mode/laptop-mode stop
                       ;;
                   esac
                   ;;
               *)  logger "ACPI action undefined: $2" ;;
           esac
           ;;
   

   Comunque per sapere se quei codici e stringhe che ci sono nei vari case sono corretti, potete usare acpi_listen da shell, provando a scollegare e a ricollegare il cavo d’alimentazione.

5. Infine questo script controlla lo stato della batteria al boot e setta la politica corrispondente, fate in modo che venga eseguito al boot ad esempio facendolo eseguire dall’rc.local:

   [~] cat /home/marinz/battery_governor.sh
   #!/bin/bash
   #
   # battery_governor.sh
   # marinz at email dot it
   
   ac_adapter="ACAD"         # check it with acpi_listen
   governor="powersave"
   
   # don't edit:
   #
   state=`sed "s/.*off-line.*/off-line/" /proc/acpi/ac_adapter/$ac_adapter/state`
   if [ "$state" = "off-line" ]; then
     cpufreq-set --governor $governor
   fi
   

Un buen artículo que no solo explica la diferencia entre Suspender e Hibernar, sino que además describe de forma sencilla los distintos modos de la especificación ACPI.

Suspender e Hibernar

我终于发现啦，原来有些东东系不能编译到内核滴。因为需要在休眠前卸载以避免冲突。

将以下以及intel显卡驱动编译成模块后，休眠的问题差不多解决了。

• Intel PRO/Wireless 3945ABG Network Connection
  iwl3945
• Generic IEEE 802.11 Networking Stack
  mac80211
• Marvell Yukon 88E8039 PCI-E Fast Ethernet Controller
  sky2
• USB Video Class
  uvcvideo
• IEEE 1394 (FireWire) support
  firewire-core firewire-sbp2 firewire-ohci
• TI Flash Media PCI74xx/PCI76xx host adapter support
  tifm_7xx1

Our client is looking for a senior level systems engineer to join their team.  The ideal individual would be a self-starter that has no qualms about working on high performing and highly scalable systems.

Job Responsibilities

• Work on defining test harnesses to qualify Scalable and High-Performance systems.
• Work with product management and support to design, develop and document (and execute) test plans and test cases
• Provide leadership in guiding the company to achieve the above goals i.e. drive and gather requirements and highlight bottlenecks. Follow up on all the open issues and driving towards their closure.
• Execute benchmark tests (CPU, Memory, Storage and Network) and qualify our client’s product against well-known and established industry standards.

Required Skills

• A minimum of 10+ year’s cumulative experience developing or testing system level software on Linux/UNIX and Windows environments.
• Strong experience in standard benchmarks (applications) to qualify systems (database, servers, platforms, etc).
• Good scripting skills using any of the well known scripting languages e.g. Perl, Expect, Python and Bash.
• Good knowledge in Operating System concepts.
• Strong experience with systems level software.
• Experience in testing clustering software will be a plus
• Working experience with Xen, KVM, VMware and Microsoft Hyper-V is desirable.
• Knowledge of x86 CPU architecture, memory management, BIOS, ACPI, etc. a plus
• Bachelor’s or Master’s degree in Computer Science or Engineering.
• Excellent communication, documentation and organizational/time-management skills.

If you are interested in this position, please send your resume to kathy@mindsource.com.

Not all computer cases are equally well designed. That, I’m sure, you know. The thing you may not know, is just how badly designed certain cases are. Not only are they hard to service, they can be detrimental to your hardware! I recently had to change the hard drive from my backup box and ended up changing the whole kit altogether because the old machine did not recognize the new 500GB hard drive. So I reused my old Compaq Presario 6400nx computer, which does recognize the 500GB hard drive, but makes it run very hot. I mean, very hot.

The new drive, although almost idling except for the nightly backup scripts and whatnots, got to 50°C. Even if 50°C is within the manufacturer’s expected operating parameters (10 to 55°C, or similar), that’s about 20°C more than my main workstation’s hard drive that runs around 30–35°C. That got me worried because it is well known that high drive temperature shortens its life considerably. So I decided to solve the problem using tie-wraps—no, not à la McGyver.

Opened, the casing looks like this:

The Compaq Presario 6400NX, opened, on its side

So it looks pretty much like an ordinary casing with slots for externally accessible drives (floppy and CD/DVD), the motherboard and lots of cables. The guys probably thought that the airflow, which is meant to bring cool air into the computer and take the heat away through the rear fans (the casing’s and the power supply’s) would be a nice airflow like this one,

Supposed Air Flow

with a strong stream covering hard drives, RAM (that you can’t see on the picture, it’s somewhere under the IDE cables), and CPU. Maybe even a cute small eddy on the side to ventilate that corner. Well, with all the cabling and the ill-designed face-plate, the airflow is pretty much like this:

The real airflow, likely

Which leaves the drive bay (shown in bright red in the next picture) and the drives it contains out of the airflow and free to heat up as much as they want. As I said, the drive got up to 50°C while the room temperature is about 20°C, which, after discussing with friends, is apparently not that surprising, even not worrying. Nevertheless, I find this dangerously hot, especially that my other hard drive (same brand, but different capacity) runs at 30°C in the same conditions—although obviously in a better case.

Guess where are the hard drives? (Shown in bright red!)

So I dug a 80mm fan from my “left-over” drawer (the one we all have, the one with a 1996 logitech mouse, a 5¼″ drive, 42 USB cables for devices we don’t even own anymore, and the like), two black tie-wraps, and frankensteined the fan on the underside of the hard drive bay, resulting in the new airflow (I suppose):

Airflow with an extra fan

*
* *

The new airflow mustn’t be all that bad. The CPU remains at a constant 40-ish°C, which it always have been, and the hard drive now runs at 30–35°C, just like the other one. Of course the extra fan adds a little noise but that’s not bad, all things considered.

You should also take time to assess the thermal health of your machine(s). On Linux, ACPI and board-specific sensor support will help you look at the temperature of the chip-set, motherboard and CPU; tools like the smartmontools package will tell you how hot your drives are. If your OS cannot report those or your missing drivers, very often the BIOS has a hardware monitor screen (or similar) that reports voltages and temperatures.

If you are in doubt regarding the actual temperatures of the components (some report ludicrous temperature as their sensors are very roughly calibrated) get an IR touchless thermometer (they’re not that expensive anymore) and point it at various components to validate the temperatures reported by the OS. Most of my hardware seems to report correct temperatures, except for my GPU on the video card (it reports 60+°C when the heat sink just above the chip is barely 40°C).

If you have a less than cromulent casing, like the one on the Compaq computer I have, make sure that none of your computer’s components runs excessively hot. Make sure that the fans are clear from dust and other debris, that will also help quite a lot (to remove dust, I blow the casing, at a safe distance, with ~50 PSI air blown with a general purpose garage-style air nozzle once in a while… gets the job done!).

A happy computer is a cool computer.

Further readings

The Wikibook Minimizing Hard Disk Drive Failure and Data Loss contains, amongst other things, a link to a paper by guys from Google that document drive failure rate in relation to drive temperature, and they show that the sweet spot is somewhere around 35-45°C.

i wrote a simple plugin to display the percentage of battery on fbpanel (i am using openbox + debian)

put this code inside ~/.fbpanel/default

Plugin {
type=genmon
config {
Command = echo $(acpi - i | grep "harging" | cut -d, -f2)
PollingTime = 2
TextSize = small
TextColor = red
}
}

this is how it look like

Dieser Artikel veranschaulicht, wie man mit einfachen Bordmitteln Informationen über den Stromverbrauch seines Laptop-Akkus herausfinden kann.

Stromverbrauch bei Laptops

Das Verzeichnis /proc liefert eine Unmenge an Informationen über das laufende System. Die Informationen über den Akku des Notebooks findet man so (ausgehend von einem Akku mit der Bezeichnung BAT0):

# cd /proc/acpi/battery/BAT0
# ls
alarm  info  state

Werfen wir doch mal einen Blick in info, um Aufschluss über den momentanen Zustand des Akkus meines HP Pavillion zu bekommen:

# cat state
present:                 yes
design capacity:         6000 mAh
last full capacity:      2848 mAh
battery technology:      rechargeable
design voltage:          14800 mV
design capacity warning: 147 mAh
design capacity low:     89 mAh
capacity granularity 1:  10 mAh
capacity granularity 2:  25 mAh
model number:            Primary
serial number:
battery type:            LION
OEM info:                Hewlett-Packard

Laut Hersteller sollte die Batterie eine Kapazität von 6000 mAh haben (design capacity). Meine hat bei der letzten Vollladung gerade mal müde 2848 mAh zustande gebracht (last full capacity). Zeit für eine neue Batterie…

Das erste Warnsignal (“Achtung! Batterie wird schwach!”) wird bei 147 mAh angezeigt (design capacity warning). Bei 89 mAh sollte das Herunterfahren des Systems veranlasst werden (design capacity low).

Wir befinden uns noch immer im Verzeichnis /proc/acpi/battery/BAT0 und sehen uns nun die Datei state an.

# cat state
present:                 yes
capacity state:          ok
charging state:          charged
present rate: 2243 mA
remaining capacity:      2848 mAh
present voltage:         12411 mV

Das Notebook ist gerade mit dem Netzteil verbunden, die verbleibende Kapazität entspricht der (weiter oben ausgelesenen) vollen Kapazität von 2848 mAh (remaining capacity). Bei einer Entladung von 2243 mA (present rate) hält der Akku gerade mal eine gute Stunde im vollgeladenen Zustand.

Wer also wissen will, wie viel zusätzlicher Strom z.B. durch das Ansehen einer DVD oder den download der jüngsten Ubuntu-Distribution verbraucht wird, der kann dies ganz einfach durch einen Blick in das /proc Verzeichnis herausfinden. Alles was man tun muss, ist die Differenz von present rate vor und während der Ausführung einer Anwendung zu vergleichen.

I am running Mythbuntu 8.10 on my Asus A8V-VM SE with a Athlon 64 3200+ and faced severe problems. The most annoying one was the problem, that mythtv periodically wasn’t able to detect the tuner cards, although they were detected by the kernel and were configured correctly.

I tried several tips like moving the daemon call of mythtv-backend in /etc/rc5.d to a later position, from S25mythtv-backend to S99mythtv-backend – this didn’t really help. Until I finally got a setup, which works now in most cases or let’s say, nearly with all bootups.

So to help you preventing one or another headache, this is my setup wich works now really well:

In the Asus BIOS, you have to disable:
APIC
ACPI 2.0 support

and in Ubuntu, you have to add

noapic


in /boot/grub/menu.lst, so that it looks like this:


kernel /boot/vmlinuz-2.6.27-11-generic root=UUID=xyz ro locale=de_DE quiet noapic


The next article I am planning will describe how to setup mythwelcome and acpi wakeup with Mythbuntu 8.10 on the crappy Asus A8V-VM SE – stay tuned.

Update:
I recently posted exactly here another socalled dirty tip, which should solve the problem – but as I have faced the same problem today again, that mythtv sometimes just refuses to detect the cards, I deleted it. But I think (and hope) I found the real deal today, the problem, why mythtv sometimes just doesn’t detect the cards, although the kernel detects it correctly.

The solution is posted in another post here:
Link

I changed my Sylvania G (original, non-Meso) netbook to Windows XP/Linux dual-boot to test some software I’m working on, and discovered that while Windows XP certainly does boot and run in general on the G, some kind of system timer or timing loop is severely out of whack!  I wanted to use my little G as a portable gaming machine from the Windows XP install, and to my horror, ZSNES couldn’t decide what speed it wanted to run!  Now, I’ve never had a single issue with ZSNES on any computer I’ve ever tried it on, even preferring the Windows port of it over the Linux native one, and not once has a problem existed with ZSNES that I couldn’t find an easy fix for, until now.

I’ve been researching the matter and gathering evidence, and I may have a potential answer to the problem.  Linux requires activation of the VIA C7 Enhanced PowerSaver module e_powersaver to clock the VIA C7 CPU properly between 400 and 1200 MHz; apparently the default speed of the CPU is only 600 MHz instead of 1200 MHz, because Linux installs without e_powersaver and Windows XP report a ~600 MHz processor where a 1.2 GHz one exists.  Here’s the extremely weird part, though: if I check the System control panel shortly after bootup and read the clock speed, sometimes it registers a clock speed of 198 MHz (about 200 MHz) which isn’t even one of the ACPI P-states for the VIA C7-M 1.2 processor.

I’ve unlocked the Windows HAL options (I’ll post how to do that at another time) and switched between ACPI Multiprocessor PC (the default for the image I used) and ACPI Uniprocessor PC and MPS Uniprocessor PC, all of which use the local APIC for IRQ routing but the MPS variant of which doesn’t theoretically touch ACPI.  Nothing seems to have helped.  I have two working theories as to what’s going on here, and how it might be fixed:

1. A calibration loop in Windows a la BogoMIPS in Linux is being screwed up by the VIA C7, or
2. The VIA C7’s PowerSaver feature is ignored or incorrectly used by Windows (via generic ACPI P-states) and it’s throwing off some kind of timer that ZSNES relies on for proper emulation of the 65816 CPU and SPC audio processor.

So far, I haven’t found a solution to this problem, and Sylvania’s site is extremely unhelpful, with only Windows drivers and a new version of gOS, but no BIOS updates or further information.  I’m looking into the technical stuff on the VIA c7 now, and it looks like the solution (assuming Windows isn’t doing something sinister) lies in clever manipulation of the C7 model-specific registers (MSRs) that control the processor’s power state.  If ZSNES is mis-calibrating some kind of tight internal timing loop because of some kind of CPU clocking issue, then tweaking the MSRs may be the solution to the problem.  Unfortunately, I’m no Windows developer, so I’m not certain how I should approach the problem.  I don’t think it’s isolated to ZSNES either, but I don’t recall what I saw that justifies that belief.  In any case, I’m working on it.  It’s just one of many pesky projects I’m hitting my head against at the moment.  We’re still working on that remote access software package; in fact, someone found our site and called us, and I had to sort of turn her away.  It’s all a bit behind schedule, and there’s not really much I can do to make things proceed any more quickly.  Stay tuned…

The Animation Council of the Philippines, Inc. (ACPI), in cooperation with the UNESCO National Commission of the Philippines (UNACOM), proudly announces the first nationwide animated flash fiction scriptwriting contest as part of its active campaign in producing original Filipino content in animation that’s grounded on the rich cultural heritage of the country.

With the immense popularity of animated films and cartoons among the young generation, not to mention adults, animation has become an effective tool in raising awareness and educating the public about Philippine traditions, history and culture. However, there is a dearth of Philippine culture-centered animation. Foreign productions have been dominating the movie houses and taking over the boob tube.

As the world progresses with developments in information communication technology, the world of creative writing keeps up with the fast-paced lifestyle of people and joins the bandwagon of instant creations through instant storytelling. A new trend of short story writing has been introduced to the literary world known as “flash fiction.” It is defined as a very short story that ranges from 100 to 1000 words, and something that can be written or read in one sitting.

What better way to use this emerging genre of flash fiction in transforming and developing it into a script for a short animated feature project that centers on and promotes cultural diversity, as well as encourages the development of original Filipino content in animation.

The competition is open to Filipino citizens of all ages. The entry must be good for a 5-minute story (average of 8 to 10 pages) which can be funded for a possible production in the future. All stories must explore one or a combination of the following ideas: a) celebration of Filipino freedom from colonization, b) recognition of the Filipino’s cultural diversity, c) promotion of the Philippine cultural heritage, its traditions and way of life.

Entrants may submit a maximum of three entries for the contest, but each entry shall contain different themes/ideas as specified for this contest.  Entries may be in English or Filipino. However, entrants writing in Filipino must be ready to submit an English translation of the work when requested or when the need arises.

Adaptations of existing material (e.g., flash fiction by other authors) are also eligible. In this case, the entrant must submit along with his or her entry the written consent of the author of the existing work to have his or her work be adapted for contest entry.

Entries may be delivered personally to the secretariat or sent in by mail or courier but must be received on or before the deadline. Envelope(s) must be properly labeled as “Flash fiction scriptwriting contest,” and addressed to the ACPI secretariat: 9^th floor, The Palisades Condominium, 107 Perea St., Legaspi Village, Makati City. The deadline for submission of entries is 12 noon of 15 May 2009.

The winners of the 1^st animated flash fiction scriptwriting contest will be announced and awarded on 15 June 2009. The prizes are: 1st – PhP 10,000.00; 2nd – PhP 7,000.00; and 3rd – PhP 5,000.00. All winners will be posted on the ACPI website, and shall receive an official email notification from the ACPI.

The Animation Council of the Philippines, Inc. shall have at its disposal, the option to have all the winning entries produced as a five-minuter animated feature later on as part of its vision and mission in producing and developing animated features of original Filipino content.

For inquiries on the submission of entries and contest rules, please call the ACPI secretariat at 817-2727 local 108, or email animationcouncil@gmail.com, or log on to www.animationcouncil.org.

Un compañero de trabajo me contacto para pedirme que le solucione un problema que le estaba sucediendo en un servidor virtualizado. La máquina tenía instalado un CentOS 5, corriendo en MS Virtual Server. Cuando verificabas la hora del servidor, tardaba 3 o 4 segundos en actualizarse el segundero. Al principio no lo creía, pero lo verifiqué ejecutando:

$ watch -n1 'date'

El problema se estaba generando por un conflicto de los controladores acpi del kernel y producto de MS. La solución, fue deshabilitar el acpi mediante parámetros pasados al kernel.

Editamos el archivo de configuración de grub

$ vi /boot/grub/menu.lst

y buscamos la/s linea/s que comienza por kernel /vmlinuz-2.6.18-8.1.15.el5. Al final, agregamos los parámetros clock=pit noapic nolapic acpi=off y listo.
Debería quedar algo similar a esto:

title CentOS (2.6.18-8.1.15.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-8.1.15.el5 ro root=/dev/VolGroup00/LogVol00 rhgb quiet clock=pit noapic nolapic acpi=off
initrd /initrd-2.6.18-8.1.15.el5.img
title CentOS (2.6.18-8.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-8.el5 ro root=/dev/VolGroup00/LogVol00 rhgb quiet clock=pit noapic nolapic acpi=off
initrd /initrd-2.6.18-8.el5.img


Lo único, hay que reiniciar (se podía hacer con setpci, modificando interrupciones, pero dependía del chipset y no era permanente como esto).

Por un momento, creí que estaba adentro de un capítulo de Lost…

To determine power consumption of a laptop:

watch -n1 'cat /proc/acpi/battery/BAT0/*'

taken from here.

Em quase todas as distribuições, é possível especificar opções adicionais durante o boot. Estas opções são repassadas diretamente ao kernel, alterando seu comportamento, ou desativando recursos que podem fazer com que o sistema trave durante o boot, ou que determinados não seja reconhecidos.

Você pode combinar várias opções, caso necessário, combinando de uma vez várias opções para solucionar problemas comuns (acpi=off, noapic, nolapic, etc.) de maneira a tentar fazer a máquina terminar o boot e depois ir testando uma a uma para descobrir qual elas que realmente resolveu o problema.

Para saber quais são esses parametros e opções, leia a dica no site Guia do Hardware.

http://www.gdhpress.com.br/blog/opcoes-de-boot/