Active Topics
-
Camera phone competition March 2023 "Endless" (35)
to General by Maemish - 4 hrs, 15 mins ago -
TLS 1.2 for N9 is ready (28)
to MeeGo / Harmattan by smartblu9 - 5 hrs, 29 mins ago -
Maemo repository (3)
to Community by torsen - 20 hrs, 58 mins ago -
Error after installing Leste on sdcard (11)
to Maemo 7 / Leste by Maemish - 1 day, 18 hrs ago -
Introducing ubiboot N9 (multiboot OS loader) (1,393)
to Alternatives by smatkovi - 3 days, 1 hr ago - more...
calltrace |
2010-04-07
, 19:45
|
|
Posts: 25 |
Thanked: 2 times |
Joined on Jan 2010
@ Prague
|
#1601
|
Originally Posted by Rushmore
this is because (I think and test which i have done confirm that) playing an mp3 in media player provide low cpu usage ,so system underclock it at 125MHz ,and those usage percent are related to this frequenci (test : i set lowest freq to 500MHz and cpu usage was at ~22% @ mp3 vbr v3 (~190kbps))
|
|
2010-04-07
, 19:46
|
|
Posts: 1,224 |
Thanked: 1,763 times |
Joined on Jul 2007
|
#1602
|
Originally Posted by Rushmore
Something seems to not make sense with Conky. Anyone notice that regardless of the clock speed, the percent of cpu usage stays fairly "constant"?
Example playing a 160KB MP3 with built-in media player:
Clock fluctuates from 125mhz on up to 800mhz, but the cpu usage percent ranges from 50% to 53%. I would expect both values (clock and usage) to be relative and not such a wide margin of fluctuation with the clock and usage. If one is constant, the other should be. I could see if there is lag for one value to the next, but that should be an issue of displacement of values, rather than one staying always "constant" (the usage).
BTW, most of the time, cpu is at 125mhz, but does fluctuate for several seconds up from 500 to 800, but usage does not deviate from 50% to 53% range.
The reason is that the kernel does not do exact accounting of CPU time - it only samples once for every slice (probably 10ms on N900, I did not check this), and accounts all the slice to the process, even if it uses only a part of the slice and then yields.
The way to get more accurate is to run another process that simply counts at a lower priority, and see how slower it becomes when an mp3 player is running. In this method mp3 players usually are over-reported, and use a lot less CPU than what seems.
I won't be surprised if pulse-audio also uses a lot less than what is always reported.
__________________
My repository
"N900 community support for the MeeGo-Harmattan" Is the new "Mer is Fremantle for N810".
No more Nokia devices for me.
My repository
"N900 community support for the MeeGo-Harmattan" Is the new "Mer is Fremantle for N810".
No more Nokia devices for me.
| The Following User Says Thank You to Matan For This Useful Post: | ||
|
|
2010-04-07
, 20:03
|
|
Posts: 126 |
Thanked: 327 times |
Joined on Nov 2009
@ Finland
|
#1603
|
Originally Posted by titan
No Im not. Theres no real way to test it so I cant tell. On my site I've put a note to not to use the DSP overclocked ones unless for some reason you really have to.
@Matan and Lehto:
are you sure you're overclocking the DSP?
did also modify the tables in drivers/dsp/bridge/rmgr/drv_interface.c ?
People should rather use your kernels, I'll continue on experimenting but as I dont have a wide knowledge of how all the kernel architechture works mine will be just shitty hacks.
 |
|
2010-04-07
, 20:09
|
|
Posts: 547 |
Thanked: 461 times |
Joined on Aug 2009
@ UK
|
#1604
|
Lehto any chance of a 125 - 650 kernel please? fiasco style so i can flash on device, thanx
|
|
2010-04-07
, 20:14
|
|
Posts: 126 |
Thanked: 327 times |
Joined on Nov 2009
@ Finland
|
#1605
|
Originally Posted by biggzy
I have to say that im incompetent and dont have a clue how to make a fiasco image.
Lehto any chance of a 125 - 650 kernel please? fiasco style so i can flash on device, thanx
|
|
2010-04-07
, 20:21
|
|
Posts: 946 |
Thanked: 1,650 times |
Joined on Oct 2009
@ Germany
|
#1606
|
Originally Posted by Lehto
http://talk.maemo.org/showpost.php?p...&postcount=861
I have to say that im incompetent and dont have a clue how to make a fiasco image.
 |
|
2010-04-07
, 20:22
|
|
Posts: 1,559 |
Thanked: 1,786 times |
Joined on Oct 2009
@ Boston
|
#1607
|
Originally Posted by egoshin
I though that was what i said.
Flandry, this N900 works on highest CPU voltage with 600MHz anyway. There is no way to increase it even more for higher frequencies (at least I don't know).
The top OPP is considered above spec and that's what's used for 600 MHz. That's what i was referring to...I definitely don't see any value in tweaking the voltages even higher, if the way to do it becomes known. Without any way to improve the cooling that's just trouble waiting to happen.
__________________
Unofficial PR1.3/Meego 1.1 FAQ
Unofficial PR1.3/Meego 1.1 FAQ
***
Classic example of arbitrary Nokia decision making. Couldn't just fallback to the no brainer of tagging with lat/lon if network isn't accessible, could you Nokia?
| The Following User Says Thank You to Flandry For This Useful Post: | ||
|
|
2010-04-07
, 20:22
|
|
Posts: 547 |
Thanked: 461 times |
Joined on Aug 2009
@ UK
|
#1608
|
Originally Posted by Lehto
ok Lehto no probs, keep up this great work.
I have to say that im incompetent and dont have a clue how to make a fiasco image.
|
|
2010-04-07
, 20:23
|
|
Posts: 946 |
Thanked: 1,650 times |
Joined on Oct 2009
@ Germany
|
#1609
|
Originally Posted by Lehto
was just asking because I also missed the files in the DSP bridge driver.
No Im not. Theres no real way to test it so I cant tell. On my site I've put a note to not to use the DSP overclocked ones unless for some reason you really have to.
It may be possible that the DSP is not overclocked over 430MHz unless you change that driver too.
 |
|
2010-04-07
, 20:25
|
|
Posts: 290 |
Thanked: 165 times |
Joined on Sep 2009
|
#1610
|
Originally Posted by prana
its located in /var/log/ (syslog file)
Where is the log file located? I want to see those messages. I am using the same OC kernel.
But you have to have syslogd or sysklogd installed.
 |
«
Previous Thread
|
Next Thread
»
|
All times are GMT. The time now is 14:43.