Re: utime accounting regression since 4.6 (was: Re: [PACTH v2 0/3] Implement /proc/<pid>/totmaps)

[Michal Hocko]

On Tue 23-08-16 17:46:11, Rik van Riel wrote:
[...]
> Does the workload consistently get zero user
> time?

root@test1:~# for i in $(seq 20); do /usr/bin/time -v awk '/^Rss/{rss+=$2} /^Pss/{pss+=$2} END {printf "rss:%d pss:%d\n", rss, pss}' /proc/3028/smaps; done 2>&1 | grep "User time" | sort | uniq -c
     20         User time (seconds): 0.00

sounds like pretty much consistent bad luck to me. I've attached the
single awk strace run, maybe this will help you understand the timing...

[...]

> Do I need to do anything special to reproduce
> your bug, besides running the max mmap program
> and the awk script?

That's how I noticed this. Nothing special really need. I am running
that in a kvm virt. machine but the bare metal gives me the same
results.

-- 
Michal Hocko
SUSE Labs

Attachment: awk.strace.gz
Description: application/gzip