Processes stuck in 'D' state (fwd)

Alan Cox (apataki@viconet.com)
Sun, 29 Mar 1998 18:02:00 -0500 (EST)


This message is in MIME format. The first part should be readable text,
while the remaining parts are likely unreadable without MIME-aware tools.
Send mail to mime@docserver.cac.washington.edu for more info.

---1463811839-226631905-890881444=:1209
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
Content-ID: <Pine.LNX.3.96.980329175704.26539C@comet.bluesky.net>

I have tried looking into this some more.
It seems that the 2.0.34-pre4 exhibits the same problem too.
I tried putting print statements in lock_super and unlock_super to see why
the process would get stuck in the wait queue for the superblock (and thus
the 'D' state for a long time).
Putting the printk statements in reduced the frequency of the problem but
I still got it to happen a few times (that leads me to believe that it's
some race condition somewhere). When I killed the httpsd process, it
locked the superblock but never unlocked it.
How could I go on from here? Is there a way to get a trace of calls to
see where the locking is taking place? (I saw the lock_super function
getting called from quite a lot of places).
Any other ideas would be welcome.

Andras Pataki
apataki@viconet.com

---------- Forwarded message ----------
Date: Wed, 25 Mar 1998 22:04:04 -0500 (EST)
From: apataki@viconet.com
To: linux-kernel@vger.rutgers.edu
Subject: Processes stuck in 'D' state

I have compiled and installed apache-ssl 1.2.5.
It runs ok, but when I try to kill the main process, it gets stuck in the
'D' state. ps -l shows something like:

FLAGS UID PID PPID PRI NI SIZE RSS WCHAN STA TTY TIME
COMMAND
100544 202 224 1 0 0 0 0 12978d DW ?
(httpsd)

System.map around 12978d looks like:

00129690 T get_fs_type
00129740 T __wait_on_super
001297d0 T sync_supers

so it seems it's in __wait_on_super.

After this, more and more processes get stuck in the 'D' state starting
with 'update' continuing with shells, etc.
I can't reboot nicely, have to hit the reset button ...
No oops or anything in syslog.
Regular apache 1.2.5 (without the ssl patches works fine).

This is completely repeatable on two different machines with different
hardware. Both have been running Linux for quite some time without
problems (2 and 4 years respectively), so I'm quite positive it's not a
hardware problem.

I tried kernels 2.0.33, 2.0.31 and 2.0.29, all give the same result.
The kernel is the same between the two machines, except different modules
are loaded (attached is the .config).

Hardware:

machine #1:
486/66, 32M RAM
3com 509 ethernet
2 IDE disks

machine #2:
P133, 64M RAM
3com 590 ethernet
Buslogic 958 UW SCSI with Seagate drive

software:
glibc 2.0.6
gcc-2.7.2.3

Any help would be appreciated. Let me know if there is any other useful
info I could post. I'm not really a good kernel hacker, but I could try
any suggestions to track down the problem.

Thanks

Andras Pataki
apataki@viconet.com

---1463811839-226631905-890881444=:1209
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII; NAME=".config"
Content-Transfer-Encoding: BASE64
Content-ID: <Pine.LNX.3.96.980325220404.1209B@comet.bluesky.net>
Content-Description:
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---1463811839-226631905-890881444=:1209--

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu