View previous topic :: View next topic |
Author |
Message |
mbar Veteran
Joined: 19 Jan 2005 Posts: 1990 Location: Poland
|
Posted: Thu Sep 19, 2013 8:02 am Post subject: Haswell, Intel P State and fiery inferno |
|
|
I have that infamous Sony Vaio Pro 13 ultrabook with i5-4200U Haswell CPU and I'm struggling with power consumption / performance.
To cut long story short, I can't make the gentoo 3.11.1 kernel (intel_pstate.c patched with Haswell ID) or even 3.12-rc1 to behave. It is not downclocking the CPU in idle when Intel P state control driver is enabled. Quite the contrary, this is the situation in idle:
Code: | uptime
09:55:06 up 15 min, 3 users, load average: 0,00, 0,03, 0,05 |
And see i7z:
Code: | Cpu speed from cpuinfo 2294.00Mhz
cpuinfo might be wrong if cpufreq is enabled. To guess correctly try estimating via tsc
Linux's inbuilt cpu_khz code emulated now
True Frequency (without accounting Turbo) 2294 MHz
CPU Multiplier 23x || Bus clock frequency (BCLK) 99.74 MHz
Socket [0] - [physical cores=2, logical cores=4, max online cores ever=2]
TURBO ENABLED on 2 Cores, Hyper Threading ON
Max Frequency without considering Turbo 2393.74 MHz (99.74 x [24])
Max TURBO Multiplier (if Enabled) with 1/2/3/4 Cores is 26x/23x/23x/23x
Real Current Frequency 2296.89 MHz [99.74 x 23.03] (Max of below)
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % Temp
Core 1 [0]: 2292.35 (22.98x) 100 1 0 0 55
Core 2 [1]: 2296.89 (23.03x) 1 100 0 0 55
C0 = Processor running without halting
C1 = Processor running with halts (States >C0 are power saver)
C3 = Cores running with PLL turned off and core cache turned off
C6 = Everything in C3 + core state saved to last level cache
Above values in table are in percentage over the last 1 sec
[core-id] refers to core-id number in /proc/cpuinfo
'Garbage Values' message printed when garbage values are read
Ctrl+C to exit
|
One core is running without halting! And yes, I have full dyntics, and all power management options known to mankind enabled, etc.
Situation changes somewhat when emerging, it will downclock when all cores are used due to power envelope.
As you can see, in idle, Intel P state driver keeps this CPU constantly in turbo mode.
If I switch to plain ACPI P-state and ondemand governor, it underclocks to 800 MHz, but then I have no turbo modes available. Only 800 - 1600 MHz range, but during emerge it kinda sticks to 800 MHz, which is pointless.
Any help is greatly appreciated.
EDIT: ACPI ondemand driver:
Code: | Cpu speed from cpuinfo 2295.00Mhz
cpuinfo might be wrong if cpufreq is enabled. To guess correctly try estimating via tsc
Linux's inbuilt cpu_khz code emulated now
True Frequency (without accounting Turbo) 2295 MHz
CPU Multiplier 23x || Bus clock frequency (BCLK) 99.78 MHz
Socket [0] - [physical cores=2, logical cores=4, max online cores ever=2]
TURBO ENABLED on 2 Cores, Hyper Threading ON
Max Frequency without considering Turbo 2394.78 MHz (99.78 x [24])
Max TURBO Multiplier (if Enabled) with 1/2/3/4 Cores is 26x/23x/23x/23x
Real Current Frequency 798.86 MHz [99.78 x 8.01] (Max of below)
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % Temp
Core 1 [0]: 798.26 (8.00x) 100 65.2 0 0 47
Core 2 [1]: 798.86 (8.01x) 1 98.9 1 0 46
C0 = Processor running without halting
C1 = Processor running with halts (States >C0 are power saver)
C3 = Cores running with PLL turned off and core cache turned off
C6 = Everything in C3 + core state saved to last level cache
Above values in table are in percentage over the last 1 sec
[core-id] refers to core-id number in /proc/cpuinfo
'Garbage Values' message printed when garbage values are read
Ctrl+C to exit
|
|
|
Back to top |
|
|
PaulBredbury Watchman
Joined: 14 Jul 2005 Posts: 7310
|
Posted: Thu Sep 19, 2013 9:16 am Post subject: Re: Haswell, Intel P State and fiery inferno |
|
|
mbar wrote: | all power management options |
Also undervolt the CPU. |
|
Back to top |
|
|
mbar Veteran
Joined: 19 Jan 2005 Posts: 1990 Location: Poland
|
Posted: Thu Sep 19, 2013 9:22 am Post subject: |
|
|
Will look into that.
Also please scratch that under ACPI driver turbo is not available. It is. This is during emerge libreoffice, about 30 minutes into it:
Code: | Cpu speed from cpuinfo 2294.00Mhz
cpuinfo might be wrong if cpufreq is enabled. To guess correctly try estimating via tsc
Linux's inbuilt cpu_khz code emulated now
True Frequency (without accounting Turbo) 2294 MHz
CPU Multiplier 23x || Bus clock frequency (BCLK) 99.74 MHz
Socket [0] - [physical cores=2, logical cores=4, max online cores ever=2]
TURBO ENABLED on 2 Cores, Hyper Threading ON
Max Frequency without considering Turbo 2393.74 MHz (99.74 x [24])
Max TURBO Multiplier (if Enabled) with 1/2/3/4 Cores is 26x/23x/23x/23x
Real Current Frequency 1921.87 MHz [99.74 x 19.27] (Max of below)
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % Temp
Core 1 [0]: 1921.77 (19.27x) 99.7 16.5 0 0 64
Core 2 [1]: 1921.87 (19.27x) 99.3 16.8 0 0 66
C0 = Processor running without halting
C1 = Processor running with halts (States >C0 are power saver)
C3 = Cores running with PLL turned off and core cache turned off
C6 = Everything in C3 + core state saved to last level cache
Above values in table are in percentage over the last 1 sec
[core-id] refers to core-id number in /proc/cpuinfo
'Garbage Values' message printed when garbage values are read
Ctrl+C to exit
|
Not too bad, seems I'll settle with ACPI P state and ondemand for now. |
|
Back to top |
|
|
mbar Veteran
Joined: 19 Jan 2005 Posts: 1990 Location: Poland
|
Posted: Thu Sep 26, 2013 10:12 am Post subject: |
|
|
All those troubles were caused by full dynticks. After I reverted to dyntics idle, everything is OK even with intel_pstate:
Code: | Cpu speed from cpuinfo 2294.00Mhz
cpuinfo might be wrong if cpufreq is enabled. To guess correctly try estimating via tsc
Linux's inbuilt cpu_khz code emulated now
True Frequency (without accounting Turbo) 2294 MHz
CPU Multiplier 23x || Bus clock frequency (BCLK) 99.74 MHz
Socket [0] - [physical cores=2, logical cores=4, max online cores ever=2]
TURBO ENABLED on 2 Cores, Hyper Threading ON
Max Frequency without considering Turbo 2393.74 MHz (99.74 x [24])
Max TURBO Multiplier (if Enabled) with 1/2/3/4 Cores is 26x/23x/23x/23x
Real Current Frequency 1631.04 MHz [99.74 x 16.35] (Max of below)
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % Temp
Core 1 [0]: 1595.34 (16.00x) 1 98.8 1 0 44
Core 2 [1]: 1631.04 (16.35x) 4.97 96.5 0 0 46
C0 = Processor running without halting
C1 = Processor running with halts (States >C0 are power saver)
C3 = Cores running with PLL turned off and core cache turned off
C6 = Everything in C3 + core state saved to last level cache
Above values in table are in percentage over the last 1 sec
[core-id] refers to core-id number in /proc/cpuinfo
'Garbage Values' message printed when garbage values are read
Ctrl+C to exit
|
Yeah, the frequency is still 1600 Mhz, but both cores are halted nearly 100% of the time. |
|
Back to top |
|
|
shad0w_GR n00b
Joined: 06 Apr 2011 Posts: 16
|
Posted: Sat Nov 16, 2013 10:11 am Post subject: |
|
|
Hi, i have 4570, do you need acpi_cpufreq in order pstates to throttle cpu? |
|
Back to top |
|
|
Gusar Advocate
Joined: 09 Apr 2005 Posts: 2665 Location: Slovenia
|
Posted: Sat Nov 16, 2013 11:47 am Post subject: |
|
|
The intel_pstate driver is a replacement for acpi_cpufreq, you load one or the other. It contains internal governors, so whatever acpi governors you choose in the kernel config is irrelevant. And it doesn't actually do much with p-states because switching them constantly is an outdated concept on modern Intel CPUs. In deeper C-states, the frequency is zero anyway! And Haswell will spend most of its time in the C7 state. For more info, read this: https://plus.google.com/114657443111661859546/posts/dLn9T4ehywL
One more thing - if you want to use i7z, you must use i7z-git from here: https://github.com/ajaiantilal/i7z. Any other version is outdated and will not show info correctly on Haswell. For example, the i7z output in mbar's posts is bogus. This is how it should look like (note the additional column for C7): Code: | Cpu speed from cpuinfo 2893.00Mhz
cpuinfo might be wrong if cpufreq is enabled. To guess correctly try estimating via tsc
Linux's inbuilt cpu_khz code emulated now
True Frequency (without accounting Turbo) 2892 MHz
CPU Multiplier 29x || Bus clock frequency (BCLK) 99.72 MHz
Socket [0] - [physical cores=4, logical cores=4, max online cores ever=4]
TURBO ENABLED on 4 Cores, Hyper Threading OFF
Max Frequency without considering Turbo 2991.72 MHz (99.72 x [30])
Max TURBO Multiplier (if Enabled) with 1/2/3/4 Cores is 36x/35x/33x/32x
Real Current Frequency 3534.81 MHz [99.72 x 35.45] (Max of below)
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % C7 % Temp VCore
Core 1 [0]: 3534.81 (35.45x) 1.27 6.54 0 0 91.9 30 0.9780
Core 2 [1]: 3527.95 (35.38x) 1 0.211 0 0 99.2 27 1.0220
Core 3 [2]: 3531.66 (35.41x) 1 0.0421 0 0 99.6 26 1.0245
Core 4 [3]: 3517.45 (35.27x) 1 4.36 0 0 95.1 30 1.0198
C0 = Processor running without halting
C1 = Processor running with halts (States >C0 are power saver modes with cores idling)
C3 = Cores running with PLL turned off and core cache turned off
C6, C7 = Everything in C3 + core state saved to last level cache, C7 is deeper than C6
Above values in table are in percentage over the last 1 sec
[core-id] refers to core-id number in /proc/cpuinfo
'Garbage Values' message printed when garbage values are read
Ctrl+C to exit
|
You could also use turbostat, you'll find it in the kernel source, /usr/src/linux/tools/power/x86/turbostat to be exact. Example output: Code: | cor CPU %c0 GHz TSC SMI %c1 %c3 %c6 %c7 CTMP PTMP %pc2 %pc3 %pc6 %pc7 Pkg_W Cor_W GFX_W
0.40 3.51 2.89 0 2.69 0.12 0.10 96.69 33 34 92.74 0.00 0.00 0.00 8.01 0.34 0.01
0 0 0.84 3.52 2.89 0 3.78 0.39 0.20 94.79 33 34 92.74 0.00 0.00 0.00 8.01 0.34 0.01
1 1 0.75 3.51 2.89 0 5.47 0.09 0.20 93.49 33
2 2 0.02 3.35 2.89 0 1.49 0.00 0.00 98.49 29
3 3 0.01 3.45 2.89 0 0.02 0.00 0.00 99.98 32
|
This one also shows Package C states and power consumption. Note how the CPUs are at the max *turbo* freq I have. But that doesn't matter, because they're actually at this freq only in the C0 state. |
|
Back to top |
|
|
shad0w_GR n00b
Joined: 06 Apr 2011 Posts: 16
|
Posted: Sat Nov 16, 2013 11:57 am Post subject: |
|
|
So having CONFIG_X86_ACPI_CPUFREQ=m isn't bad, as much as CONFIG_X86_INTEL_PSTATE=y is enabled. Right?
I am using cpupower, and cpupower indeed shows that most of the time is in C7 state, but frequency is pretty high. Is cpu throttle in this cpus unnecessary?
Code: | |Nehalem || SandyBridge || Mperf || Idle_Stats
CPU | C3 | C6 | PC3 | PC6 || C7 | PC2 | PC7 || C0 | Cx | Freq || POLL | C1 | C2
0| 0.00| 0.00| 0.00| 0.00|| 86.16| 0.00| 0.00|| 6.90| 93.10| 1197|| 0.00| 5.84| 87.21
1| 0.00| 0.00| 0.00| 0.00|| 81.42| 0.00| 0.00|| 8.97| 91.03| 1366|| 0.00| 8.65| 82.30
2| 0.00| 0.00| 0.00| 0.00|| 92.42| 0.00| 0.00|| 4.46| 95.54| 1489|| 0.00| 2.49| 93.06
3| 0.00| 0.00| 0.00| 0.00|| 88.48| 0.00| 0.00|| 5.31| 94.69| 1482|| 0.00| 5.48| 89.17
|
Using turbostat
Code: | cor CPU %c0 GHz TSC SMI %c1 %c3 %c6 %c7 CTMP PTMP %pc2 %pc3 %pc6 %pc7 Pkg_W Cor_W GFX_W
8.93 1.83 3.20 0 9.78 0.00 0.00 81.29 29 30 0.00 0.00 0.00 0.00 7.78 1.58 0.12
0 0 10.72 1.62 3.20 0 13.58 0.00 0.00 75.69 29 30 0.00 0.00 0.00 0.00 7.78 1.58 0.12
1 1 7.27 1.65 3.20 0 8.77 0.00 0.00 83.96 26
2 2 10.62 1.94 3.20 0 8.83 0.00 0.00 80.55 27
3 3 7.11 2.17 3.20 0 7.94 0.00 0.00 84.94 26
|
But in lsmod acpi_cpufreq is loaded. Without it, lowest frequency is never reached. I can understand that in C7 states, frequency doesn't matter, but why?
Thanks |
|
Back to top |
|
|
Gusar Advocate
Joined: 09 Apr 2005 Posts: 2665 Location: Slovenia
|
Posted: Sat Nov 16, 2013 12:30 pm Post subject: |
|
|
shad0w_GR wrote: | frequency doesn't matter, but why? |
Did you real the link I gave? It explains everything in detail: The freq is *zero* when the CPU is in C3, C6 or C7 states. And at the lowest when it's in the C1 state. It's only at the number you see when it's in the C0 state. |
|
Back to top |
|
|
shad0w_GR n00b
Joined: 06 Apr 2011 Posts: 16
|
Posted: Sat Nov 16, 2013 12:44 pm Post subject: |
|
|
Gusar wrote: | shad0w_GR wrote: | frequency doesn't matter, but why? |
Did you real the link I gave? It explains everything in detail: The freq is *zero* when the CPU is in C3, C6 or C7 states. And at the lowest when it's in the C1 state. It's only at the number you see when it's in the C0 state. |
Oh i see. But turbo(using turbostat) doesn't seem to work. Based on frequency Ghz column in turbostat reports.
More from i7z:
Code: | Cpu speed from cpuinfo 3199.00Mhz
cpuinfo might be wrong if cpufreq is enabled. To guess correctly try estimating via tsc
Linux's inbuilt cpu_khz code emulated now
True Frequency (without accounting Turbo) 3198 MHz
CPU Multiplier 32x || Bus clock frequency (BCLK) 99.94 MHz
Socket [0] - [physical cores=4, logical cores=4, max online cores ever=4]
TURBO ENABLED on 4 Cores, Hyper Threading OFF
Max Frequency without considering Turbo 3297.94 MHz (99.94 x [33])
Max TURBO Multiplier (if Enabled) with 1/2/3/4 Cores is 36x/36x/35x/34x
Real Current Frequency 3199.69 MHz [99.94 x 32.02] (Max of below)
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % C7 % Temp VCore
Core 1 [0]: 3199.08 (32.01x) 85.6 6.69 0 0 7.76 40 0.9509
Core 2 [1]: 3199.59 (32.02x) 45.7 15.6 0 0 38.8 40 0.9480
Core 3 [2]: 3199.69 (32.02x) 21.6 20 0 0 58.4 39 0.9484
Core 4 [3]: 3199.25 (32.01x) 44.1 10.4 0 0 45.5 35 0.9510
C0 = Processor running without halting
C1 = Processor running with halts (States >C0 are power saver modes with cores idling)
C3 = Cores running with PLL turned off and core cache turned off
C6, C7 = Everything in C3 + core state saved to last level cache, C7 is deeper than C6
Above values in table are in percentage over the last 1 sec
[core-id] refers to core-id number in /proc/cpuinfo
'Garbage Values' message printed when garbage values are read
Ctrl+C to exit
|
|
|
Back to top |
|
|
toralf Developer
Joined: 01 Feb 2004 Posts: 3922 Location: Hamburg
|
Posted: Sat Nov 16, 2013 3:58 pm Post subject: |
|
|
shad0w_GR wrote: | So having CONFIG_X86_ACPI_CPUFREQ=m isn't bad, as much as CONFIG_X86_INTEL_PSTATE=y is enabled. Right? | Yes, and You can switch off pstate at the kernel command line via intel_pstate=disabke IIRC. SO compiling ondemand governor as module and you always have a working fall back. |
|
Back to top |
|
|
shad0w_GR n00b
Joined: 06 Apr 2011 Posts: 16
|
Posted: Sun Nov 17, 2013 12:20 am Post subject: |
|
|
Everything works with acpi_cpufreq module loaded.
Turbo isn't used if acpi_cpufreq is missing. |
|
Back to top |
|
|
Gusar Advocate
Joined: 09 Apr 2005 Posts: 2665 Location: Slovenia
|
Posted: Sun Nov 17, 2013 12:45 am Post subject: |
|
|
What exactly does acpi_cpufreq missing mean - that there's no driver handling the CPU or that intel_pstate is running? Turbo works here with intel_pstate on my i5-4570S.
Oh, I just remembered something: Which kernel are you using? If it's not 3.12, you need a patch for intel_pstate to work on haswell, did you apply that patch? |
|
Back to top |
|
|
shad0w_GR n00b
Joined: 06 Apr 2011 Posts: 16
|
Posted: Sun Nov 17, 2013 9:24 am Post subject: |
|
|
Gusar wrote: | What exactly does acpi_cpufreq missing mean - that there's no driver handling the CPU or that intel_pstate is running? Turbo works here with intel_pstate on my i5-4570S.
Oh, I just remembered something: Which kernel are you using? If it's not 3.12, you need a patch for intel_pstate to work on haswell, did you apply that patch? |
I mean that if i have blacklist acpi_cpufreq, turbo doesn't work.
I use vanilla 3.11.6. Ok ill give a shot to 3.12. Thanks |
|
Back to top |
|
|
shad0w_GR n00b
Joined: 06 Apr 2011 Posts: 16
|
Posted: Sun Nov 17, 2013 2:10 pm Post subject: |
|
|
Gusar wrote: | What exactly does acpi_cpufreq missing mean - that there's no driver handling the CPU or that intel_pstate is running? Turbo works here with intel_pstate on my i5-4570S.
Oh, I just remembered something: Which kernel are you using? If it's not 3.12, you need a patch for intel_pstate to work on haswell, did you apply that patch? |
Thanks, using 3.12 turbo works without loading acpi_cpufreq. |
|
Back to top |
|
|
albright Advocate
Joined: 16 Nov 2003 Posts: 2588 Location: Near Toronto
|
Posted: Sat Jan 11, 2014 11:12 pm Post subject: |
|
|
I have a thinkpad T440s, i5-4300u and i7z does not seem to work
properly. I have the git cloned version but the output is
like this:
Code: | Cpu speed from cpuinfo 2494.00Mhz
cpuinfo might be wrong if cpufreq is enabled. To guess correctly try estimating via tsc
Linux's inbuilt cpu_khz code emulated now
True Frequency (without accounting Turbo) 2494 MHz
CPU Multiplier 25x || Bus clock frequency (BCLK) 99.76 MHz
Socket [0] - [physical cores=2, logical cores=4, max online cores ever=2]
TURBO ENABLED on 2 Cores, Hyper Threading ON
Max Frequency without considering Turbo 2593.76 MHz (99.76 x [26])
Max TURBO Multiplier (if Enabled) with 1/2/3/4 Cores is 29x/26x/26x/26x
Real Current Frequency 2456.61 MHz [99.76 x 24.63] (Max of below)
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % Temp VCore
Core 1 [0]: 2456.61 (24.63x) 1 98.7 1 0 36 0.8752
Core 2 [2]: 2202.06 (22.07x) 1 100 0 0 35 0.8755
|
Am I right to assume that i7z is not yet ready to handle the i5-4300u? _________________ .... there is nothing - absolutely nothing - half so much worth
doing as simply messing about with Linux ...
(apologies to Kenneth Graeme) |
|
Back to top |
|
|
Gusar Advocate
Joined: 09 Apr 2005 Posts: 2665 Location: Slovenia
|
Posted: Sat Jan 11, 2014 11:23 pm Post subject: |
|
|
albright wrote: | Am I right to assume that i7z is not yet ready to handle the i5-4300u? |
It's possible. Observe its output when you quit it (or start it with --nogui), then it'll tell you if it doesn't recognize the CPU.
In the meantime, try turbostat. And make sure you're running a 3.12 kernel. |
|
Back to top |
|
|
albright Advocate
Joined: 16 Nov 2003 Posts: 2588 Location: Near Toronto
|
Posted: Mon Jan 13, 2014 12:24 am Post subject: |
|
|
Here's some output from turbostat (no c7):
Code: | ~/Docs/insts/linux-v3.1-rc6-1332-g18e42dc $ sudo ./turbostat
Password:
cr CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6
0.25 2.88 2.49 99.72 0.02 0.01 0.00 0.00
0 0 0.12 2.88 2.49 99.85 0.03 0.01 0.00 0.00
0 1 0.85 2.89 2.49 99.12 0.03 0.01 0.00 0.00
1 2 0.01 2.79 2.49 99.97 0.01 0.01 0.00 0.00
1 3 0.03 2.79 2.49 99.96 0.01 0.01 0.00 0.00
cr CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6
0.25 2.81 2.49 99.71 0.03 0.01 0.00 0.00
0 0 0.11 2.80 2.49 99.82 0.05 0.02 0.00 0.00
0 1 0.87 2.82 2.49 99.06 0.05 0.02 0.00 0.00
1 2 0.01 2.80 2.49 99.99 0.00 0.00 0.00 0.00
1 3 0.03 2.70 2.49 99.97 0.00 0.00 0.00 0.00
cr CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6
0.25 2.88 2.49 99.74 0.01 0.00 0.00 0.00
0 0 0.12 2.88 2.49 99.88 0.01 0.00 0.00 0.00
0 1 0.85 2.89 2.49 99.15 0.01 0.00 0.00 0.00
1 2 0.01 2.82 2.49 99.98 0.01 0.00 0.00 0.00
1 3 0.03 2.75 2.49 99.96 0.01 0.00 0.00 0.00
cr CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6
0.27 2.89 2.49 99.70 0.02 0.01 0.00 0.00
0 0 0.17 2.89 2.49 99.77 0.03 0.02 0.00 0.00
0 1 0.87 2.89 2.49 99.08 0.03 0.02 0.00 0.00
1 2 0.01 2.86 2.49 99.99 0.00 0.00 0.00 0.00
1 3 0.03 2.77 2.49 99.97 0.00 0.00 0.00 0.00
cr CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6
0.25 2.88 2.49 99.72 0.02 0.01 0.00 0.00
0 0 0.51 2.89 2.49 99.44 0.04 0.01 0.00 0.00
0 1 0.45 2.89 2.49 99.50 0.04 0.01 0.00 0.00
1 2 0.01 2.83 2.49 99.98 0.01 0.00 0.00 0.00
1 3 0.03 2.76 2.49 99.96 0.01 0.00 0.00 0.00
cr CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6
0.28 2.89 2.49 99.70 0.01 0.00 0.00 0.00
0 0 0.88 2.89 2.49 99.10 0.02 0.00 0.00 0.00
0 1 0.22 2.89 2.49 99.75 0.02 0.00 0.00 0.00
1 2 0.01 2.82 2.49 99.99 0.00 0.00 0.00 0.00
1 3 0.03 2.79 2.49 99.97 0.00 0.00 0.00 0.00
|
_________________ .... there is nothing - absolutely nothing - half so much worth
doing as simply messing about with Linux ...
(apologies to Kenneth Graeme) |
|
Back to top |
|
|
|