https://bugzilla.novell.com/show_bug.cgi?id=479855 User vkrevs@yahoo.com added comment https://bugzilla.novell.com/show_bug.cgi?id=479855#c2 --- Comment #2 from Vadim Krevs 2009-02-26 07:50:51 MST --- Created an attachment (id=275715) --> (https://bugzilla.novell.com/attachment.cgi?id=275715) dmesg output after echo t > /proc/sysrq-trigger -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=479855 User lpechacek@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=479855#c4 Libor Pecháček changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |vkrevs@yahoo.com --- Comment #4 from Libor Pecháček 2009-03-02 08:13:21 MST --- Thanks. Unfortunately I was a little bit too fast on requesting information from you and did not realize that you *can* kill the process with KILL signal. That means the process is not in D state, and sysrq-t will not show anything special. Overall the process list from comment #2 looks more or less OK, so I'd like to know more about your problem: 1) If you run tail(1) on other files in the system, can it be killed with Ctrl-C or kill -TERM? Please try e.g. 'tail -f /var/log/mail', 'tail -f /etc/passwd' or some others. 2) Please attach full output of 'ps sxa', after you tried to kill it with kill(1). 3) Is your system up-to-date? Does 'yast2 online_update' report any updates to install? 4) Which KDE version do you use? KDE3 or KDE4? Can this bug be reproduced in the text console? Switch to text mode console with Ctrl-Alt-F1, log in as root and run the tail process. Does it show the same behavior as in KDE? TIA for answering the above questions. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=479855 User lpechacek@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=479855#c6 Libor Pecháček changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WORKSFORME --- Comment #6 from Libor Pecháček 2009-03-05 04:35:37 MST --- Thanks for response, in case it reappears, reopen this bug and provide information requested in comment #4, points 1 and 2. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=479855 User vkrevs@yahoo.com added comment https://bugzilla.novell.com/show_bug.cgi?id=479855#c8 Vadim Krevs changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|WORKSFORME | --- Comment #8 from Vadim Krevs 2009-03-20 04:15:16 MST --- It's back. stal-dev-lx1:~ # cat /proc/version Linux version 2.6.27.20-1-debug (geeko@buildhost) (gcc version 4.3.2 [gcc-4_3-branch revision 141291] (SUSE Linux) ) #1 SMP 2009-03-18 23:06:47 +0100 Answers to questions from comment #4: 1. "tail -f" cannot be killed with Ctrl-C, kill, kill -HUP or kill -TERM for any file - /var/log/messages, /etc/password or any other file. 2. Attached in #7 3. Latest KDE 4.2.1 for x86_64. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=479855 User vkrevs@yahoo.com added comment https://bugzilla.novell.com/show_bug.cgi?id=479855#c10 --- Comment #10 from Vadim Krevs 2009-03-20 04:21:06 MST --- Interestingly, once I attach to the tail process via gdb and issue kill, it gets killed. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=479855 User lpechacek@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=479855#c12 Libor Pecháček changed: What |Removed |Added ---------------------------------------------------------------------------- Component|Kernel |KDE4 Workspace AssignedTo|bnc-team-screening@forge.pr |kde-maintainers@suse.de |ovo.novell.com | Summary|tail -f /var/log/messages |processess spawned from |can only be killed with |konsole have signals |kill -9 <PID> |blocked Severity|Normal |Major --- Comment #12 from Libor Pecháček 2009-03-21 15:16:40 MST --- I bet on a KDE bug here. Attachment 280871 (ps sxa) shows that the processes' sigmask is set to block the signals. Moreover there are large number of zombie processes (nepomukserver, kmix, konsole, kopete, ...) which probably means that their parent is stuck. Attachment 275715 (sysrq-t of another instance of this issue) shows that many KDE processes were reparented to init, what means that some crucial process died. The problem could also be introduced by the Oracle server running on the host, but I don't think there is a connection. Vadim, please provide the output of ps axouid,pid,ppid,pending,blocked,ignored,caught,stat,tty,time,command so that we are able to see the process hierarchy along with the signal mask states. The process should be killable by some of the non-blocked signals like ILL, ABRT and few others. Changing component to KDE - oS11.1 + KDE4.2 from Factory. KDE maintainers, please have a look. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=479855 User llunak@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=479855#c14 Lubos Lunak changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |NEEDINFO Info Provider| |vkrevs@yahoo.com --- Comment #14 from Lubos Lunak 2009-03-23 09:17:37 MST --- https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-180/+bug/3... ? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=479855 User lpechacek@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=479855#c16 Libor Pecháček changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Component|KDE4 Workspace |Kernel Resolution| |INVALID AssignedTo|kde-maintainers@suse.de |bnc-team-screening@forge.pr | |ovo.novell.com --- Comment #16 from Libor Pecháček 2009-03-24 11:32:19 MST --- Hmm, I prematurely ruled out possible kernel problems. Thanks Lubos for help. The point is that in the kernel space there is no memory protection. The binary only drivers are very often buggy and write to memory locations which they should not touch. Once you randomly write something somewhere in the kernel space, strange things might happen. Closing as INVALID. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.