GNU/Linux S5R3 App 4.38 available for Beta test

Ed1934158
Ed1934158
Joined: 10 Nov 04
Posts: 62
Credit: 14,481,483
RAC: 0

RE: Check whether Linux is

Message 79890 in response to message 79889

Quote:
Check whether Linux is running your processor at the full 1.70GHz - it may have dropped down to a power-save mode if there is little high-priority load on the CPU.


How can I check this? Using top I can see that E@H uses 99%.
cat /proc/cpuinfo returns:
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 13
model name : Intel(R) Pentium(R) M processor 1.70GHz
stepping : 6
cpu MHz : 1694.636
cache size : 2048 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr mce cx8 sep mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss tm pbe up est tm2
bogomips : 3391.33

Michael Karlinsky
Michael Karlinsky
Joined: 22 Jan 05
Posts: 888
Credit: 23,502,182
RAC: 0

RE: cpu MHz :

Message 79891 in response to message 79890

Quote:

cpu MHz : 1694.636

It does run at 1.7 GHz.

Michael

Richard Haselgrove
Richard Haselgrove
Joined: 10 Dec 05
Posts: 2,143
Credit: 2,920,749,972
RAC: 960,395

RE: RE: Check whether

Message 79892 in response to message 79890

Quote:
Quote:
Check whether Linux is running your processor at the full 1.70GHz - it may have dropped down to a power-save mode if there is little high-priority load on the CPU.

How can I check this? Using top I can see that E@H uses 99%.
cat /proc/cpuinfo returns:
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 13
model name : Intel(R) Pentium(R) M processor 1.70GHz
stepping : 6
cpu MHz : 1694.636
cache size : 2048 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr mce cx8 sep mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss tm pbe up est tm2
bogomips : 3391.33


Have a look at this post in the 'KUBUNTU v8.04-BETA --- excruciating slowdown' thread.

As Michael says, your CPU was running at 1.7GHz at the moment you ran cpuinfo, but that isn't to say that it always runs at that speed - that probably depends on your cpufreq-selector setting.

It would be good if you could keep us posted on the outcome of your investigations - we never heard the outcome from the originator of that other thread.

Ed1934158
Ed1934158
Joined: 10 Nov 04
Posts: 62
Credit: 14,481,483
RAC: 0

I run: sudo cpufreq-selector

I run:
sudo cpufreq-selector -g performance
output:
No cpufreq support
I also don't use Beagle. Could it be that Debian kernel in some way is responsable for this?

Output in Messages:
Mon 21 Apr 2008 01:46:05 AM CEST||Starting BOINC client version 5.4.11 for i686-pc-linux-gnu
Mon 21 Apr 2008 01:46:05 AM CEST||libcurl/7.15.5 OpenSSL/0.9.8c zlib/1.2.3 libidn/0.6.5
Mon 21 Apr 2008 01:46:05 AM CEST||Data directory: /var/lib/boinc-client
Mon 21 Apr 2008 01:46:05 AM CEST|Einstein@Home|Found app_info.xml; using anonymous platform
Mon 21 Apr 2008 01:46:05 AM CEST||Processor: 1 GenuineIntel Intel(R) Pentium(R) M processor 1.70GHz
Mon 21 Apr 2008 01:46:05 AM CEST||Memory: 504.25 MB physical, 3.91 GB virtual
Mon 21 Apr 2008 01:46:05 AM CEST||Disk: 17.30 GB total, 13.20 GB free
Mon 21 Apr 2008 01:46:05 AM CEST|Einstein@Home|URL: http://einstein.phys.uwm.edu/; Computer ID: 1240463; location: home; project prefs: default
Mon 21 Apr 2008 01:46:05 AM CEST||General prefs: from Einstein@Home (last modified 2008-03-28 00:06:15)
Mon 21 Apr 2008 01:46:05 AM CEST||General prefs: no separate prefs for home; using your defaults
Mon 21 Apr 2008 01:46:05 AM CEST||Reading preferences override file
Mon 21 Apr 2008 01:46:05 AM CEST||Local control only allowed
Mon 21 Apr 2008 01:46:05 AM CEST||Listening on port 31416
Mon 21 Apr 2008 01:46:05 AM CEST|Einstein@Home|Resuming task h1_0930.30_S5R3__256_S5R3b_1 using einstein_S5R3 version 438
Mon 21 Apr 2008 01:48:36 AM CEST|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
Mon 21 Apr 2008 01:48:36 AM CEST|Einstein@Home|Reason: Requested by user
Mon 21 Apr 2008 01:48:36 AM CEST|Einstein@Home|(not requesting new work or reporting completed tasks)
Mon 21 Apr 2008 01:48:41 AM CEST|Einstein@Home|Scheduler request succeeded
Mon 21 Apr 2008 01:48:41 AM CEST|Einstein@Home|Message from server: Resent lost result h1_0826.75_S5R3__128_S5R3b_3
Mon 21 Apr 2008 01:48:41 AM CEST|Einstein@Home|Message from server: Resent lost result h1_1009.75_S5R3__546_S5R3b_1
Mon 21 Apr 2008 01:48:43 AM CEST|Einstein@Home|Started download of file skygrid_0830Hz_S5R3.dat
Mon 21 Apr 2008 01:48:43 AM CEST|Einstein@Home|Started download of file h1_0826.75_S5R3
Mon 21 Apr 2008 01:48:54 AM CEST|Einstein@Home|Finished download of file skygrid_0830Hz_S5R3.dat
Mon 21 Apr 2008 01:48:54 AM CEST|Einstein@Home|Throughput 118742 bytes/sec
...

Ed1934158
Ed1934158
Joined: 10 Nov 04
Posts: 62
Credit: 14,481,483
RAC: 0

I hope I have found the

I hope I have found the solution. ACPI process is what I think is responsible for the problem. It controls cpu cycle usage:

ed@killers2:/proc/acpi/processor/CPU0$ ls
info  limit  power  throttling
ed@killers2:/proc/acpi/processor/CPU0$ cat info
processor id:            0
acpi id:                 0
bus mastering control:   yes
power management:        yes
throttling control:      yes
limit interface:         yes
ed@killers2:/proc/acpi/processor/CPU0$ cat limit
active limit:            P0:T2
user limit:              P0:T0
thermal limit:           P0:T2
ed@killers2:/proc/acpi/processor/CPU0$ cat power
active state:            C2
max_cstate:              C8
bus master activity:     00000007
states:
    C1:                  type[C1] promotion[C2] demotion[--] latency[000] usage[00000010] duration[00000000000000000000]
   *C2:                  type[C2] promotion[C3] demotion[C1] latency[001] usage[00002700] duration[00000000000024999924]
    C3:                  type[C3] promotion[C4] demotion[C2] latency[085] usage[00000000] duration[00000000000000000000]
    C4:                  type[C3] promotion[--] demotion[C3] latency[185] usage[00000000] duration[00000000000000000000]
ed@killers2:/proc/acpi/processor/CPU0$ cat throttling
state count:             8
active state:            T2
states:
    T0:                  00%
    T1:                  12%
   *T2:                  25%
    T3:                  37%
    T4:                  50%
    T5:                  62%
    T6:                  75%
    T7:                  87%
ed@killers2:/proc/acpi/processor/CPU0$


I'll report the change tomorrow.

Bikeman (Heinz-Bernd Eggenstein)
Bikeman (Heinz-...
Moderator
Joined: 28 Aug 06
Posts: 3,522
Credit: 693,602,198
RAC: 107,333

So, if it's really true that

So, if it's really true that this is an issue of "nice" processes not bringing the CPU to full clockspeed, I wonder what

nice -19 cat /proc/cpuinfo

would display ? Maybe worth a try.

CU

Bikeman

Ed1934158
Ed1934158
Joined: 10 Nov 04
Posts: 62
Credit: 14,481,483
RAC: 0

RE: So, if it's really true

Message 79896 in response to message 79895

Quote:

So, if it's really true that this is an issue of "nice" processes not bringing the CPU to full clockspeed, I wonder what

nice -19 cat /proc/cpuinfo

would display ? Maybe worth a try.

CU

Bikeman


The same, 1694Mhz. I'm quite convinced that the problem was with this line:

/proc/acpi/processor/CPU0$ cat throttling
state count: 8
active state: T2
states:
T0: 00%
T1: 12%
*T2: 25%
T3: 37%
T4: 50%
T5: 62%
T6: 75%
T7: 87%


I've uninstalled few packages that control cpu speed:
-cpufreqd
-cpudyn
-cpufrequtils
And still the same it varies from T2 to T3. I feel helpless. In BIOS everything is set to maximum. I've asked the question on Debian and Ubuntu forums, maybe someone there knows the solution. I'll inform about results.

Ed1934158
Ed1934158
Joined: 10 Nov 04
Posts: 62
Credit: 14,481,483
RAC: 0

Problem solved. One guy

Problem solved. One guy helped me to turn off acpi in kernel, my next workunit will be over in about 36000 seconds, which is OK.
Here are the instructions:
Edit your /boot/grub/menu.lst and add acpi=off

kernel /boot/vmlinuz-2.6.20-16-generic root=UUID=4f146ca4-8f7f-47c4-a72e-e19e101288cd ro quiet splash acpi=off

Or /usr/bin/cpufreq-selector --help for arguments

The first worked for me. The second one should work for most also.

Btw, one way to find out that you're an addict is thinking about changing OS that works perfectly for you because of E@H. In few months I'll get two Q6600, I'm afraid to think what will happen then. :)

Paul D. Buck
Paul D. Buck
Joined: 17 Jan 05
Posts: 754
Credit: 5,385,205
RAC: 0

RE: Btw, one way to find

Message 79898 in response to message 79897

Quote:
Btw, one way to find out that you're an addict is thinking about changing OS that works perfectly for you because of E@H. In few months I'll get two Q6600, I'm afraid to think what will happen then. :)


You will lie awake at night thinking about how much more you could do with three ...

Happens to me all the time ...

Heck I just got a dual 3.2 GHz Mac Pro and I am already thinking about how to get more done with upgrading other computers ... Haven't even paid the bill for the Pro yet (due the 8th, check in the mail)...

And no, you don't want to know how much I spent ...

Annika
Annika
Joined: 8 Aug 06
Posts: 720
Credit: 494,410
RAC: 0

Sounds familiar to me... I

Sounds familiar to me... I consider getting one of those nice quadcores (Intel I think) in September or October to replace my single core Athlon, and believe me I'm already wondering how much difference that will make in terms of RAC, even if I'm not yet 100% sure I'll actually upgrade...

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.