Public bug reported:
On a machine with one i7-4930K CPU (six cores; twelve threads), I run
apt-get install gmp-ecm
echo
417851657851322288874010837856502301110211653321414464007924033448506685217369529551512221034931058157509867474898886146419540545861923501516226812207
> N
for u in $(seq 1 12); do ecm -c 200 1e8 < N > e$u & done
Expected behaviour:
* good interactive response from machine
* load average around 12.0
* twelve jobs appear in top, each with about 100% of a thread
* mpstat -P ALL 1 displays a job running at 100% on each of the twelve CPUs
Actual behaviour:
* lousy interactive response from machine:
pumpkin@pumpkin:/home/nfsworld/aliquot/2360.1573/e$ time ls /
bin boot core dev etc home initrd.img initrd.img.old lib lib32 lib64
lost+found media mnt oldroot opt proc root run sbin srv sys tmp usr
var vmlinuz vmlinuz.old
real 0m53.205s
* load average is somewhere around seven
* somewhere between four and six jobs appear in top, with many using much less
than 100% of a thread
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
5362 pumpkin 20 0 513756 504388 2912 R 100.0 1.5 1:54.28 ecm
5363 pumpkin 20 0 12552 2380 2224 R 100.0 0.0 1:32.32 ecm
5366 pumpkin 20 0 513692 503932 2832 R 100.0 1.5 0:40.80 ecm
5367 pumpkin 20 0 513752 504416 2940 R 93.2 1.5 2:32.51 ecm
5368 pumpkin 20 0 12596 3152 2836 R 16.2 0.0 6:00.99 ecm
5370 pumpkin 20 0 12596 3116 2800 R 9.9 0.0 6:31.30 ecm
* mpstat -P ALL 1 indicates several idle CPUs
18:26:46 CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %gnice %idle
18:26:47 all 50.04 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 49.96
18:26:47 0 100.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00
18:26:47 1 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
18:26:47 2 100.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00
18:26:47 3 100.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00
18:26:47 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
18:26:47 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
18:26:47 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
18:26:47 7 100.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00
18:26:47 8 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
18:26:47 9 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
18:26:47 10 100.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00
18:26:47 11 100.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00
This makes the machine pretty much unusable as a compute workstation,
which is what I was using it for pre-upgrade. Please advise.
** Affects: ubuntu
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1516289
Title:
scheduling of multiple CPU-intense applications is wrong
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+bug/1516289/+subscriptions
--
ubuntu-bugs mailing list
ubuntu-bugs@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
