Description of problem: When switching to the strict selinux policy setroubleshootd is no longer functional Version-Release number of selected component (if applicable): selinux-policy-2.4.6-137.el5 How reproducible: 100% Steps to Reproduce: 1. install strict policy 2. relable filesystem 3. try and launch setroubleshootd Actual results: Cannot launch: avc: denied { connectto } for comm="sealert" egid=0 euid=0 exe="/usr/bin/python" exit=0 fsgid=0 fsuid=0 gid=0 items=0 path="/var/run/setroubleshoot/setroubleshoot_server" pid=3616 scontext=root:sysadm_r:sysadm_t:s0-s0:c0.c1023 sgid=0 subj=root:sysadm_r:sysadm_t:s0-s0:c0.c1023 suid=0 tclass=unix_stream_socket tcontext=system_u:system_r:setroubleshootd_t:s0 tty=(none) uid=0 Expected results: launching of setroubleshootd Additional info:
Are you logging in as sysadm_r:sysadm_t via XWindows? This is not considered safe in strict policy.
That would explain that.
Hmmm... It won't let me mark it as "not a bug"
------- Comment From pc.com 2010-12-21 11:30 EDT------- 1.Server architecture(s) (please list all effected) (x86/POWER6/Z/etc.): only tried on POWER7 2.Server type (9117-MMA/HS20/s390/etc.): POWER7 blade 8406-71Y 3.General component (desktop/kernel/base OS/dev tools/etc.): anaconda 4.Other components involved (ixgbe/java/emulex/etc.): none 5.Does the server have the latest GA firmware? no (7.12 vs 7.15) 6.Has the problem been shown to occur on more than one system? no 7.Is a tested patch available? no 8.What is the latest official Red Hat build on which this bug has been seen? RHEL6 GA ------- Comment From 2011-01-03 03:15 EDT------- Redhat, Any updates? Thanks Muni ------- Comment From 2011-01-04 00:32 EDT------- Redhat, Below are the steps mostly due to which Anaconda exception occurs: Install RHEL6 GA in graphics mode Select Custom partitioning Try to delete/add/modify the existing partitions. Exception occurs. -Muni ------- Comment From pc.com 2011-02-07 16:50 EDT------- (In reply to comment #16) > What > scenarios are you using when you see errors adding or modifying partitions? It's been almost two months, so if the logs don't include the information you seek, I doubt my memory will be able to serve. :-) I usually allow the installer to create it's default layout, then I tweak it. I don't use LVM as a general rule, preferring raw ext3 (or ext4) partitions. I seem to recall in this case establishing a partition layout, then attempting to undo and redo it differently. I don't recall much more than that. The logs show the anaconda crash, correct? Can you work backwards from that? Why did it crash, and how could that condition occur? (Just trying to help.) ------- Comment From 2011-02-21 03:08 EDT------- Redhat, Any updates on this issue? Thanks Muni ------- Comment From 2011-03-07 03:33 EDT------- Redhat, Did you get a chance to look into this issue? Any updates? Thanks Muni
Created attachment 488540 [details] anaconda's crash information ------- Comment on attachment From pc.com 2010-12-09 11:10 EDT------- ---Problem Description--- anaconda crashed during installation while working on adding/deleting/editing partitions. Contact Information = Paul Clarke <pc.com> ---uname output--- ? Machine Type = POWER7 blade 8406-71Y ---Debugger--- A debugger is not configured ---Steps to Reproduce--- unsure, but I'll attach the debug file that anaconda provided post-crash ---Anaconda Component Data--- Userspace tool common name: anaconda The userspace tool has the following bit modes: ? Userspace rpm: anaconda Userspace tool obtained from project website: na *Additional Instructions for Paul Clarke <pc.com>: -Post a private note with access information to the machine that the bug is occuring on. -Attach ltrace and strace of userspace application.