Note: This is a public test instance of Red Hat Bugzilla. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback at bugzilla.redhat.com.
Bug 750103 - SELinux is preventing /sbin/rpc.statd from 'write' accesses on the file rpc.statd.pid.
Summary: SELinux is preventing /sbin/rpc.statd from 'write' accesses on the file rpc.s...
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: None
Product: Fedora
Classification: Fedora
Component: selinux-policy
Version: 16
Hardware: x86_64
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Miroslav Grepl
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard: abrt_hash:1e4c1fc0a81a01c2f7f2e3072df...
: 750162 (view as bug list)
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2011-10-31 00:17 UTC by Charles R. Anderson
Modified: 2011-11-23 15:01 UTC (History)
4 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2011-11-23 15:01:02 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
script of NFS mount attempt (16.11 KB, text/plain)
2011-10-31 10:36 UTC, Charles R. Anderson
no flags Details

Description Charles R. Anderson 2011-10-31 00:17:18 UTC
libreport version: 2.0.6
executable:     /usr/bin/python
hashmarkername: setroubleshoot
kernel:         3.1.0-5.fc16.x86_64
reason:         SELinux is preventing /sbin/rpc.statd from 'write' accesses on the file rpc.statd.pid.
time:           Sun Oct 30 20:16:32 2011

description:
:SELinux is preven(removed)ing /sbin/rpc.s(removed)a(removed)d from 'wri(removed)e' accesses on (removed)he file rpc.s(removed)a(removed)d.pid.
:
:*****  Plugin ca(removed)chall (100. confidence) sugges(removed)s  ***************************
:
:If you believe (removed)ha(removed) rpc.s(removed)a(removed)d should be allowed wri(removed)e access on (removed)he rpc.s(removed)a(removed)d.pid file by defaul(removed).
:Then you should repor(removed) (removed)his as a bug.
:You can genera(removed)e a local policy module (removed)o allow (removed)his access.
:Do
:allow (removed)his access for now by execu(removed)ing:
:# grep rpc.s(removed)a(removed)d /var/log/audi(removed)/audi(removed).log | audi(removed)2allow -M mypol
:# semodule -i mypol.pp
:
:Addi(removed)ional Informa(removed)ion:
:Source Con(removed)ex(removed)                sys(removed)em_u:sys(removed)em_r:rpcd_(removed):s0
:Targe(removed) Con(removed)ex(removed)                unconfined_u:objec(removed)_r:var_run_(removed):s0
:Targe(removed) Objec(removed)s                rpc.s(removed)a(removed)d.pid [ file ]
:Source                        rpc.s(removed)a(removed)d
:Source Pa(removed)h                   /sbin/rpc.s(removed)a(removed)d
:Por(removed)                          <Unknown>
:Hos(removed)                          (removed)
:Source RPM Packages           nfs-u(removed)ils-1.2.5-1.fc16
:Targe(removed) RPM Packages           
:Policy RPM                    selinux-policy-3.10.0-46.fc16
:Selinux Enabled               True
:Policy Type                   (removed)arge(removed)ed
:Enforcing Mode                Enforcing
:Hos(removed) Name                     (removed)
:Pla(removed)form                      Linux (removed) 3.1.0-5.fc16.x86_64 #1 SMP Thu Oc(removed) 27
:                              03:46:50 UTC 2011 x86_64 x86_64
:Aler(removed) Coun(removed)                   2
:Firs(removed) Seen                    Sun 30 Oc(removed) 2011 08:14:58 PM EDT
:Las(removed) Seen                     Sun 30 Oc(removed) 2011 08:15:41 PM EDT
:Local ID                      b979e6f2-3dbb-4bc2-a848-e7c70b0fec0e
:
:Raw Audi(removed) Messages
:(removed)ype=AVC msg=audi(removed)(1320020141.260:91): avc:  denied  { wri(removed)e } for  pid=2284 comm="rpc.s(removed)a(removed)d" name="rpc.s(removed)a(removed)d.pid" dev=(removed)mpfs ino=27781 scon(removed)ex(removed)=sys(removed)em_u:sys(removed)em_r:rpcd_(removed):s0 (removed)con(removed)ex(removed)=unconfined_u:objec(removed)_r:var_run_(removed):s0 (removed)class=file
:
:
:(removed)ype=SYSCALL msg=audi(removed)(1320020141.260:91): arch=x86_64 syscall=open success=no exi(removed)=EACCES a0=7fc2f7f3b284 a1=241 a2=1b6 a3=7ffff6b11550 i(removed)ems=0 ppid=2283 pid=2284 auid=4294967295 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 (removed)(removed)y=(none) ses=4294967295 comm=rpc.s(removed)a(removed)d exe=/sbin/rpc.s(removed)a(removed)d subj=sys(removed)em_u:sys(removed)em_r:rpcd_(removed):s0 key=(null)
:
:Hash: rpc.s(removed)a(removed)d,rpcd_(removed),var_run_(removed),file,wri(removed)e
:
:audi(removed)2allow
:
:#============= rpcd_(removed) ==============
:allow rpcd_(removed) var_run_(removed):file wri(removed)e;
:
:audi(removed)2allow -R
:
:#============= rpcd_(removed) ==============
:allow rpcd_(removed) var_run_(removed):file wri(removed)e;
:

Comment 1 Miroslav Grepl 2011-10-31 08:43:39 UTC
Have you ever started rpcd directly?

# restorecon -R -v /var/run/rpc*

Should fix your issue. If I am wrong, please reopen the bug. Thank you.

Comment 2 Charles R. Anderson 2011-10-31 10:08:53 UTC
*** Bug 750162 has been marked as a duplicate of this bug. ***

Comment 3 Charles R. Anderson 2011-10-31 10:10:17 UTC
I need some advice then.  How are you supposed to mount an NFS filesystem in Fedora 16?  It failed to mount until I performed this sequence of events:

1. Install from Fedora 16 Final RC2 Live Desktop x86_64
2. yum install nfs-utils
3. systemctl start rpcbind.service
4. systemctl start nfs-lock.service (fails)
5. setenforce 0
6. mount -t nfs ...
7. Report these AVCs (write & unlink on rpc.statd.pid)

Comment 4 Charles R. Anderson 2011-10-31 10:36:36 UTC
Created attachment 530943 [details]
script of NFS mount attempt

Full script of session with attempt to mount an NFS filesystem right after a fresh install of Fedora 16 Final RC2 Live Desktop x86_64.  If the procedure I followed was incorrect in such a way as to cause a mislabeled file to be created, then the process is fragile and improvement as well as explanation for the proper procedure would be appreciated.
Thanks.

Comment 5 Miroslav Grepl 2011-10-31 11:45:55 UTC
How about

# yum update --enablerepo=updates testing

and

# systemctl start nfs-lock.service


I am able to start this service.


Note You need to log in before you can comment on or make changes to this bug.