cyclictest count value seems incorrect...

cyclictest count value seems incorrect...

[Gautam Thaker]

I ran the following 3 hour test using cyclictest. The count for T:0
that I saw when I returned to my screen in the morning seemed
puzzling. Because I use the '-M' option it display updates when there
is a new maximum on any CPU. Does the output as shown mean on CPU 0
the maximum value of 26 happened on count 168704 and as it never
increased after that the C: xxx is simply not updated?

I wish even with -M option at end of the display would do a final
update. Because of -M the only things that will update on the last
screen update would be count values, but this would be useful.


node-0> date; sudo ./cyclictest  -d  0 -p 90  -S -D 3h  -m -M; date
Tue Jun 21 21:37:22 MDT 2022
# /dev/cpu_dma_latency set to 0us
policy: fifo: loadavg: 0.00 0.03 0.04 2/166 5380
policy: fifo: loadavg: 0.16 0.03 0.01 2/170 5702
T: 0 ( 5368) P:90 I:1000 C: 168704 Min:      3 Act:    5 Avg:    5 Max:      26
T: 0 ( 5368) P:90 I:1000 C:9742264 Min:      3 Act:    6 Avg:    5 Max:      61
T: 1 ( 5369) P:90 I:1000 C:9742261 Min:      3 Act:  132 Avg:    5 Max:     132
T: 2 ( 5370) P:90 I:1000 C:9742257 Min:      3 Act:   13 Avg:    5 Max:      67
T: 3 ( 5371) P:90 I:1000 C:9742254 Min:      3 Act:    6 Avg:    5 Max:      91
T: 4 ( 5372) P:90 I:1000 C:9742245 Min:      3 Act:    6 Avg:    5 Max:      59
T: 5 ( 5373) P:90 I:1000 C:9742247 Min:      3 Act:   10 Avg:    5 Max:      41
T: 6 ( 5374) P:90 I:1000 C:9742244 Min:      3 Act:    7 Avg:    5 Max:      67
T: 7 ( 5375) P:90 I:1000 C:9742241 Min:      3 Act:    7 Avg:    5 Max:      29
Wed Jun 22 00:37:22 MDT 2022

Re: cyclictest count value seems incorrect...

[John Kacur]

On Wed, 22 Jun 2022, Gautam Thaker wrote:

Please always include the version in bug reports

> I ran the following 3 hour test using cyclictest. The count for T:0
> that I saw when I returned to my screen in the morning seemed
> puzzling. Because I use the '-M' option it display updates when there
> is a new maximum on any CPU. Does the output as shown mean on CPU 0
> the maximum value of 26 happened on count 168704 and as it never
> increased after that the C: xxx is simply not updated?

It looks to me that for some reason processor 0 was printed twice and the 
second one is correct.

> 
> I wish even with -M option at end of the display would do a final
> update. Because of -M the only things that will update on the last
> screen update would be count values, but this would be useful.


> 
> 
> node-0> date; sudo ./cyclictest  -d  0 -p 90  -S -D 3h  -m -M; date
> Tue Jun 21 21:37:22 MDT 2022
> # /dev/cpu_dma_latency set to 0us
> policy: fifo: loadavg: 0.00 0.03 0.04 2/166 5380
> policy: fifo: loadavg: 0.16 0.03 0.01 2/170 5702
> T: 0 ( 5368) P:90 I:1000 C: 168704 Min:      3 Act:    5 Avg:    5 Max:      26
> T: 0 ( 5368) P:90 I:1000 C:9742264 Min:      3 Act:    6 Avg:    5 Max:      61
> T: 1 ( 5369) P:90 I:1000 C:9742261 Min:      3 Act:  132 Avg:    5 Max:     132
> T: 2 ( 5370) P:90 I:1000 C:9742257 Min:      3 Act:   13 Avg:    5 Max:      67
> T: 3 ( 5371) P:90 I:1000 C:9742254 Min:      3 Act:    6 Avg:    5 Max:      91
> T: 4 ( 5372) P:90 I:1000 C:9742245 Min:      3 Act:    6 Avg:    5 Max:      59
> T: 5 ( 5373) P:90 I:1000 C:9742247 Min:      3 Act:   10 Avg:    5 Max:      41
> T: 6 ( 5374) P:90 I:1000 C:9742244 Min:      3 Act:    7 Avg:    5 Max:      67
> T: 7 ( 5375) P:90 I:1000 C:9742241 Min:      3 Act:    7 Avg:    5 Max:      29
> Wed Jun 22 00:37:22 MDT 2022
>