RE: [PATCH 0/4] sched: Various reweight_entity() fixes

["Doug Smythies" <dsmythies@telus.net>]

On 2026.02.11 08:22 Peter Zijlstra wrote:
> On Wed, Feb 11, 2026 at 11:48:33AM +0100, Peter Zijlstra wrote:
>
>> Now let me go check on that latency thing from Doug (although hopefully
>> that was the same boo-boo).
>
> Only ~3 hours so far, I'll let it run overnight.
>
> So on my ADL with P-cores only (unplug the E things):
>
> $ NR=`grep -c processor /proc/cpuinfo`; for ((i=0; i<$NR; i++)) do yes > /dev/null & done
> $ turbostat --quiet --Summary --show Busy%,Bzy_MHz,IRQ,PkgWatt,PkgTmp,TSC_MHz,Time_Of_Day_Seconds,usec --interval 1 --out
/dev/shm/turbo.log
>
> $ gawk '/^usec/ {next} { if (T) { d=$2-T; bucket[int(d*1000)]++; } T=$2 } END { for (i in bucket) { printf "%0.3f: %d\n", i/1000,
bucket[i] }}' < /dev/shm/turbo.log
> 1.000: 4943
> 1.001: 5000
> 1.002: 21
> 1.003: 10
> 1.004: 1
> 1.005: 1
>
> Which seems to suggest things are looking better again.

Same here, 6 hours and 20 minutes so far, and I'll leave it overnight:
By the way, thank you for the cool parsing and histogram command line.

$ gawk '/^usec/ {next} { if (T) { d=$2-T; bucket[int(d*1000)]++; } T=$2 } END { for (i in bucket) { printf "%0.3f: %d\n", i/1000,
bucket[i] }}' < /dev/shm/turbo.log
1.000: 12381
1.001: 10388
1.002: 32
1.003: 20
1.004: 9
1.005: 2
1.006: 2

System:
Processor: Intel(R) Core(TM) i5-10600K CPU @ 4.10GHz, 6 cores 12 CPUs.
CPU frequency scaling driver: intel_pstate.
CPU frequency scaling governor: powersave.
HWP: enabled.
idle governor teo.