1 |
Peter Humphrey <prh@××××××××××.uk> posted |
2 |
200706230916.07711.prh@××××××××××.uk, excerpted below, on Sat, 23 Jun |
3 |
2007 09:16:07 +0100: |
4 |
|
5 |
> Here's what top showed then. Look at the /nice/ values on lines 3 and 4, |
6 |
> and compare those with the %CPU and Processor fields of processes 5279 |
7 |
> and 5280 (sorry about the line wraps). This has me deeply puzzled: |
8 |
|
9 |
Fixed the line wraps and removed a bit of extraneous information. =8^) |
10 |
|
11 |
> top - 09:04:59 up 23 min, 5 users, load average: 3.60, 4.79, 3.91 |
12 |
> Tasks: 124 total, 2 running, 122 sleeping, 0 stopped, 0 zombie |
13 |
|
14 |
> Cpu0: 0.3%us, 0.3%sy, 0.0%ni, 99.3%id, [zeroes] |
15 |
> Cpu1: 0.0%us, 0.3%sy, 99.7%ni, 0.0%id, [zeroes] |
16 |
|
17 |
> PID USER PR NI S %CPU %MEM TIME+ P COMMAND |
18 |
> 5279 prh 34 19 S 50 1.0 6:53.97 1 setiathome-5.12 |
19 |
> 5280 prh 34 19 S 50 1.0 6:54.08 0 setiathome-5.12 |
20 |
|
21 |
|
22 |
> I don't think this is a scheduling problem; it goes deeper, so that the |
23 |
> kernel doesn't have a consistent picture of which processor is which. |
24 |
|
25 |
Critical question here, is that in SMP Irix or SMP Solaris mode? (See |
26 |
the top manpage if you don't know what I mean.) Asked another way, is |
27 |
that displaying percent of total CPU time (both CPUs) or percent of |
28 |
total divided by number of CPUs (so percent of one CPU)? |
29 |
|
30 |
If it's Irix mode (percent total CPU time), then it's reporting full |
31 |
usage of both CPUs, one on each. The CPU0 line would then be the one |
32 |
screwed up, since it's reporting idle when it clearly has to be in use. |
33 |
|
34 |
If it's Solaris mode (percent of a single CPU's time, so total of all |
35 |
percentages should be 200% if you have two CPUs), then the CPUs |
36 |
lines would seem to be correct, both processes would appear to be |
37 |
running on CPU1, maxing it out, and the P column of the 5280 line |
38 |
would have to be screwed up. (That's assuming you let the figures |
39 |
stabilize after the last schedtool call you made.) |
40 |
|
41 |
In either case, I'm not sure where your bug is, but you are correct, |
42 |
the problem appears to be way deeper than scheduling. I'd guess it's |
43 |
ultimately a kernel bug, possibly due to a hardware bug, possibly not, |
44 |
but you might wish to file it on top initially, just to see if they've |
45 |
seen similar and can tell you what's going on. Unless you want to |
46 |
double-check patching status yourself, you might as well file the bug |
47 |
with Gentoo first, in case it's a Gentoo bug. They'll probably end |
48 |
up closing it "upstream", but at least then when you file it upstream, |
49 |
you can say you've cleared it with Gentoo first. |
50 |
|
51 |
As for top, note that there's a trick you can use with it. You'll |
52 |
likely want to trim the memory columns etc as I did for your bug |
53 |
report, but you may not want to mess up your regular config to do |
54 |
so. Not a problem! =8^) Create a symlink to top called something |
55 |
else (say topbug). Then run it using the symlink, and you can change |
56 |
and save your setttings, and it'll save them in a different rc file |
57 |
(topbugrc using my example). That way, you can run it with the bug |
58 |
report settings when you want to, without messing up your regular |
59 |
config. |
60 |
|
61 |
Of course, don't forget to mention in your bug report whether you were |
62 |
in Solaris or Irix SMP mode, because as I explained, it /does/ make a |
63 |
difference. |
64 |
|
65 |
Let me know how this goes, post the bug number when you file it or |
66 |
whatever, as I'd like to follow it too. You definitely have a |
67 |
strange one here, and I'd /love/ to see what the real experts have |
68 |
to say about it! You are absolutely correct, it doesn't seem to |
69 |
make any sense at all! |
70 |
|
71 |
Good luck. That's one /strange/ problem you have going there! |
72 |
No /wonder/ you were expressing frustration earlier! |
73 |
|
74 |
-- |
75 |
Duncan - List replies preferred. No HTML msgs. |
76 |
"Every nonfree program has a lord, a master -- |
77 |
and if you use the program, he is your master." Richard Stallman |
78 |
|
79 |
-- |
80 |
gentoo-amd64@g.o mailing list |