Re: mbuff in 2.2.14 rtlinux-2.2

[Ray <rehenry-at-up.net>]

Sunday mbuff hung in again.  I combined the advice of  Fred and Will and
came up with the following report of the problem.  After the EMC closed, I
ran:

[root-at-localhost ray]# lsmod
Module                  Size  Used by
mbuff                   5356   1

[root-at-localhost ray]# rmmod mbuff
rmmod: mbuff: Device or resource busy
[root-at-localhost ray]#

So mbuff is definitely hanging up on something.

[root-at-localhost ray]# ps ax
  PID TTY      STAT   TIME COMMAND
    1 ?        S      0:03 init [5]
    2 ?        SW     0:00 [kflushd]
    3 ?        SW     0:00 [kupdate]
    4 ?        SW     0:00 [kpiod]
    5 ?        SW     0:02 [kswapd]
  221 ?        SW     0:00 [portmap]
  284 ?        S      0:00 syslogd -m 0
  294 ?        S      0:00 klogd
  309 ?        S      0:00 /usr/sbin/atd
  324 ?        S      0:00 crond
  339 ?        SW     0:00 [inetd]
  354 ?        SW     0:00 [lpd]
  370 ?        SW     0:00 [lockd]
  371 ?        SW     0:00 [rpciod]
  381 ?        SW     0:00 [rpc.statd]
  459 ?        S      0:00 /usr/lib/postfix/master
  465 ?        S      0:00 qmgr -l -t fifo -u
  481 ?        S      0:00 gpm -t ps/2
  496 ?        S      0:00 httpd
  536 ?        S      0:00 /usr/X11R6/bin/fontfs
  553 ?        S      0:45 xfs -port -1 -daemon
  586 tty1     SW     0:00 [mingetty]
...
  592 ?        S      0:00 kdm -nodaemon
 6954 ?        S      0:00 kppp
 7080 ?        S     17:45 /etc/X11/X -auth /etc/X11/xdm/authdir/A:0-NaDPH3
 7081 ?        S      0:00 -:0
 7094 ?        S      0:22 kwm
 7133 ?        S      0:05 /usr/bin/kswarm.kss -delay 15 -install -corners iiii 7144 ?        Z      0:00 [kwmsound <defunct>]
 7145 ?        S      0:17 kpanel
 7146 ?        S      0:39 kfm
 7147 ?        S      0:01 krootwm
 7148 ?        S      0:02 kbgndwm
10265 ?        S      0:00 httpd
...
15034 ?        S      0:00 kdesud
15184 ?        S      0:00 pickup -l -t fifo
15445 ?        S      0:03 kmail -caption Mail Client -icon kmail.xpm
15446 ?        S      0:00 konsole -icon konsole.xpm -miniicon konsole.xpmi -cap15447 pts/0    S      0:00 /bin/bash
15460 pts/0    S      0:00 su
15461 pts/0    S      0:00 bash
15476 pts/0    R      0:00 ps ax

I didn't find any leftover EMC parts there so I looked for time consumers.

[root-at-localhost ray]#top

57 processes: 55 sleeping, 1 running, 1 zombie, 0 stopped
CPU states:  0.9% user,  0.5% system,  0.0% nice,  8.3% idle
Mem:    61184K av,   59548K used,    1636K free,   56232K shrd,    2564K buff
Swap:  128484K av,    6380K used,  122104K free                   21728K cached


  PID USER     PRI  NI  SIZE  RSS SHARE STAT  LIB %CPU %MEM   TIME COMMAND
15490 root      19   0  1084 1084   888 R       0  7.3  1.7   0:00 top
    1 root       0   0   112   68    48 S       0  0.0  0.1   0:03 init
    2 root       0   0     0    0     0 SW      0  0.0  0.0   0:00 kflushd
    3 root       0   0     0    0     0 SW      0  0.0  0.0   0:00 kupdate
    4 root       0   0     0    0     0 SW      0  0.0  0.0   0:00 kpiod
    5 root       0   0     0    0     0 SW      0  0.0  0.0   0:02 kswapd
  221 bin        0   0    80    0     0 SW      0  0.0  0.0   0:00 portmap
  284 root       0   0   384  368   292 S       0  0.0  0.6   0:00 syslogd
  294 root       0   0   524  168   136 S       0  0.0  0.2   0:00 klogd
  309 daemon     0   0   140  108    64 S       0  0.0  0.1   0:00 atd
  324 root      10   0   172  112    72 S       0  0.0  0.1   0:00 crond
  339 root       0   0    68    0     0 SW      0  0.0  0.0   0:00 inetd
  354 root       0   0    80    0     0 SW      0  0.0  0.0   0:00 lpd
  370 root       0   0     0    0     0 SW      0  0.0  0.0   0:00 lockd

This seemed to meet Will's criterion so I looked at:

[root-at-localhost ray]# ipcs

------ Shared Memory Segments --------
key       shmid     owner     perms     bytes     nattch    status
0x00000000 0         root      600       1056768   11        dest
0x00000000 1         nobody    600       46084     11        dest

------ Semaphore Arrays --------
key       semid     owner     perms     nsems     status

------ Message Queues --------
key       msqid     owner     perms     used-bytes  messages

Now it seems to me that these two shared memory segments should be removed
so I tried. (after a few syntax problems)

[root-at-localhost ray]# ipcrm shm 1
resource deleted
[root-at-localhost ray]# ipcrm shm 0
resource deleted
[root-at-localhost ray]# ipcs

------ Shared Memory Segments --------
key       shmid     owner     perms     bytes     nattch    status
0x00000000 0         root      600       1056768   11        dest
0x00000000 1         nobody    600       46084     11        dest

------ Semaphore Arrays --------
key       semid     owner     perms     nsems     status

------ Message Queues --------
key       msqid     owner     perms     used-bytes  messages

But they did not go away.  And mbuff was still listed a busy when I tried
to rmmod it.

I don't quite know what to make of all this but at least we know a few
things it is not.  I tried running emc a couple of times after all this
and these shmem segments hung around.  The display did not slow down like
it sometimes does but I had to reboot to clean up the shared memory.

After reboot the ipcs report was empty.  After several runs, it remained
empty and mbuff would go away.

Ray