Skip to contentSkip to conversation

ksoftirqd accounting broken in 2.6.27-rc9

DISCUSSION Andi Kleen 2 participants 3 messages Oct 8, 2008

lore
Share & tools
lore

Andi Kleen

author
#0 lore raw

On one test box running 2.6.27rc9 for a few hours I got:

  PID TTY      STAT   TIME COMMAND
    1 ?        Ss     0:01 init [3]  
    2 ?        S<     0:00 [kthreadd]
    3 ?        S<     0:00 [migration/0]
    4 ?        S<     0:00 [ksoftirqd/0]
    5 ?        S<     0:00 [watchdog/0]
    6 ?        S<     0:00 [migration/1]
    7 ?        S<     0:00 [ksoftirqd/1]
    8 ?        S<     0:00 [watchdog/1]
    9 ?        S<     0:00 [migration/2]
   10 ?        S<   21133795:38 [ksoftirqd/2]

...
19 ? S< 21133795:38 [ksoftirqd/5]
...
25 ? S< 21133795:38 [ksoftirqd/7]

The other kernel threads seem to have correct accounting.

-Andi

Signature · 1 line
[email protected]
#1 lore raw
Quoted 13 lines · Andi Kleen

On one test box running 2.6.27rc9 for a few hours I got:

  PID TTY      STAT   TIME COMMAND
    1 ?        Ss     0:01 init [3]  
    2 ?        S<     0:00 [kthreadd]
    3 ?        S<     0:00 [migration/0]
    4 ?        S<     0:00 [ksoftirqd/0]
    5 ?        S<     0:00 [watchdog/0]
    6 ?        S<     0:00 [migration/1]
    7 ?        S<     0:00 [ksoftirqd/1]
    8 ?        S<     0:00 [watchdog/1]
    9 ?        S<     0:00 [migration/2]
   10 ?        S<   21133795:38 [ksoftirqd/2]
Quoted 6 lines

...
19 ? S< 21133795:38 [ksoftirqd/5]
...
25 ? S< 21133795:38 [ksoftirqd/7]

The other kernel threads seem to have correct accounting.

It looks like http://bugzilla.kernel.org/show_bug.cgi?id=11209, doesn't it?

Rafael

author
#2 lore raw
Quoted 19 lines · Rafael J. Wysocki
Quoted 18 lines · Andi Kleen

On one test box running 2.6.27rc9 for a few hours I got:

  PID TTY      STAT   TIME COMMAND
    1 ?        Ss     0:01 init [3]  
    2 ?        S<     0:00 [kthreadd]
    3 ?        S<     0:00 [migration/0]
    4 ?        S<     0:00 [ksoftirqd/0]
    5 ?        S<     0:00 [watchdog/0]
    6 ?        S<     0:00 [migration/1]
    7 ?        S<     0:00 [ksoftirqd/1]
    8 ?        S<     0:00 [watchdog/1]
    9 ?        S<     0:00 [migration/2]
   10 ?        S<   21133795:38 [ksoftirqd/2]

...
19 ? S< 21133795:38 [ksoftirqd/5]
...
25 ? S< 21133795:38 [ksoftirqd/7]

Quoted 3 lines
Quoted 1 line

The other kernel threads seem to have correct accounting.

It looks like http://bugzilla.kernel.org/show_bug.cgi?id=11209, doesn't it?

Yes, looks like a dup of that one. Thanks.

-Andi