[00:33] CarlFK, I thought I saw that in one of the messages before. Normally it is not too bad since that only means another clocksource gets selected. [08:23] anyone know if we added/removed support for "intex rtl8139D" nic [13:43] smb_tp , das mir reboot klappt :D danke vielmals :D [13:43] -r +t [13:44] es ist reboot=aw .. oder bessergesagt ich hab nur bw hw und aw getestet, und aw ging, da hab ik nicht noch die andern kombinationen weiter probiert [13:52] CSX_Lappy: In English? [13:54] it looks like german, but i dont recognize "bw hw und aw" [13:55] i own a Toshiba Satellite Pro 6050 , wich i don't got to reboot .. only switching off an on helps ... smb_tp said jesterday i should try the kernel boot option reboot with some parameters... now i just thanked him , cause my latop now reboots again :) [13:55] it was reboot=aw :D [13:56] but i olny tried bw hw and aw ... so i dont know if the others work, i was happy that aw worked and did not want to test further [13:57] a is acpi, which is becoming the default soon [13:57] :) [13:57] sounds great :) [13:57] now i need to get cpu-throtheling to work [13:58] hope it was correct wirtten *g* [13:58] CSX_Lappy: It's a celeron? The chip doesn't support voltage scaling. [13:59] yeah it is... but the speedstep or how it is called worked under xp , so i don't know why not with linux/ubuntu [14:01] there must be a "trick" to get it to work :) [14:01] No, it doesn't [14:01] It's doing CPU throttling. That doesn't save any power. [14:01] There's no speedstep support on the Celeron [14:02] hmmmmn [14:02] BenC: ping the xen64 pvops is working fine so far [14:02] mjg59: whats throttling do? [14:03] so why i could set it under Win xp ? [14:03] CarlFK: Asserts the stop line of the processor on a certain percentage of clock cycles [14:03] The CPU's still clocked at full speed, but it spends a proportion of time doing nothing [14:04] So at 50% throttling, it takes twice as long to do anything. And it's still taking something like 75% as much power. === asac_ is now known as asac [14:04] mjg59: sounds like that would save power, just not as much [14:04] hmmn i thougt i have read 500MHz insted the full 1200MHz in toshiba power-saver (a windows tool wich comes with the toshiba-drivers ) [14:04] CarlFK: No, not overall [14:05] CarlFK: If the cpu is idle, it saves no power. If it's active, it's now going to take you twice as long to finish the task and you're going to end up using more power to get that task finished [14:05] for real? [14:05] well, as i understand it, transistors use power when changing states, primarily [14:05] The only reason it's implemented is to reduce the immediate heat output [14:06] You trade off long-term power consumption for short-term heat reduction [14:06] hmmn.. so why should they make such a feature, when it's useless at the end [14:06] mjg59: that sounds like a violation of physics [14:06] CarlFK: most studies say that race-to-idle saves most power. Meaning "finish it as fast as you can and then sleep". Frequency scaling helps for long running processes e.g. video player constantly taking 40% cpu. [14:06] pwnguin: How? [14:06] pwnguin: You'll generate more heat in the long run, but it'll be spread out over a longer period of time [14:06] So your machine doesn't melt [14:07] you mean it's only to reduce heat, now to save ppower for later use? [14:07] s/now/not [14:08] I can see it making the battery last longer in time, which would seem good, which humans like. [14:09] i need it only for idle, to have a longer battery-live when i need the power :) [14:10] i think mjg's point is that when you're at idle, the reduced clock cycle isn't helping much [14:10] i guess the question is: how idel is the box when you aren't asking it to do anything? [14:10] CarlFK: Throttling might inprove battery life a bit only if the processor doesn't support frequency(voltage) scaling. If it does, turn off throttling and use frequency scaling. [14:10] with full lcd brightness at 1200 MHz my battery lasts 45 - 50 minutes... reducing brightness makes it lasts 2 - 2,5 hours [14:11] CSX_Lappy: a reduced heat will probably help battery longevity. batteries hate heat [14:11] which is unfortunate, because laptops make a lot of that [14:12] jeah, it gets a bit warm... enough to help out on cold nights in my bet ^^ [14:12] -t+d [14:14] mjg59 do you think i don't need any reducing of the cpu speed? [14:20] CarlFK: It can make the battery last slightly longer at the expense of you not getting as much done [14:21] CSX_Lappy: I think it depends on what you are doing. [14:21] CSX_Lappy: With a Celeron? No. [14:21] CSX_Lappy: The non-Celeron chips have voltage scaling, where the voltage is reduced at the same time as the frequency. That actually saves power. [14:21] mjg59: right - humans are more likely to notice the battery life than how much they got done in one charge :) [14:22] you never hear people express battery life in terms of productivity (true, but yeah, im basically kidding) [14:23] one benefit of reduced heat is reduced need for fans [14:23] less heat should lead to a slower battery kill over time [14:23] pwnguin: No, you're never going to generate enough heat to have any significant impact on the battery [14:24] Charge cycles are the main determining factor [14:24] didn't lasts the battery longer when it's not using the fan all over the time ? [14:24] Fans consume very little power [14:24] ok [14:24] * pwnguin didnt realize celerons were so good at not making heat [14:24] was just an idea :D [14:25] I bet there are some cases where getting less done over time is OK, like watching youtube - guessing it might render fewer frames, so not look as nice, but a 5 min video is still done in 5 min [14:25] Yeah, that's the only corner case [14:26] rendering a web page that you click off of before it is done [14:26] that't the other one :) [14:27] Stuff that will consume as much CPU as possible in order to provide a slightl nicer user experience [14:35] like vista xD [18:38] is it possible to have the xen-paravirt options built into the kernel? The reason Im asking is that all the documentation I have seen has it built into the kernel [21:27] I figured this issue may be under kernel [21:28] right now I am in Puppy Linux but trying to fix my ubuntu- it does not want to shut down.. [21:28] puppy linux shuts down fine [21:28] can anyone help [21:29] I would post it in the help forum but last time I had an issue it took 3 days to get one response [21:31] Or at least can someone direct me to the correct room to ask the question in [21:36] smb_tp: yay - no pausing. tried with highres=off nohz=off first, [21:36] taking them off now... [21:37] CarlFK, Ok, good. This narrows things down a bit at least. [21:37] rats - meant to grab the dmesg.. [21:38] more rats - it pauses without highres=off nohz=off first - should I try with just one? [21:39] CarlFK, Yeah, maybe. I'd suspect nohz maybe you could try that [21:41] booting just to a root prompt with single is good enough, right? [21:42] CarlFK, Think this is enough. IIRC the pauses were before (at least some) [21:43] good - cuz with -3 X gets cranky and I have to deal with that, then reboot into -2 [21:52] smb_tp: if nohz=no still paused - want me to try with just highres=off, and do you want the dmesg from all? [21:54] CarlFK, Yes, try the other if that doesn't work. From any pause, the dmesg would be interesting if the sysrq stuff gets in. Most interesting would be alt+print+q for the timer list. If that does not work dmesg is not needed since it won't show much === smb_tp is now known as smb_tp_away