Skip to contentSkip to conversation

crash with 2.5.74 while loading policy

DISCUSSION Colin Walters 3 participants 4 messages Jul 15, 2003 SELinux security module

lore
Share & tools
lore
#0 lore raw

Hi,

Linux oopsed while loading a policy today on my SELinux 2.5.74 machine. It doesn't offhand look directly related to SELinux, but I was curious if anyone here had seen it before.

Here's the oops output:

Jul 15 01:53:42 meta kernel: security:  3 users, 5 roles, 392 types
Jul 15 01:53:42 meta kernel: security:  30 classes, 29820 rules
Jul 15 01:53:44 meta kernel: Bad page state at free_hot_cold_page
Jul 15 01:53:44 meta kernel: flags:0x01000014 mapping:00000000 mapped:1 count:0
Jul 15 01:53:44 meta kernel: Backtrace:
Jul 15 01:53:44 meta kernel: Call Trace: [bad_page+93/144]  [free_hot_cold_page+112/304]  [__pagevec_free+28/48]  [release_pages+357/416]  [__pte_chain_free+129/144]  [free_pages_and_swap_cache+106/160]  [zap_pte_range+466/560]  [zap_pmd_range+75/112]  [unmap_page_range+75/128]  [unmap_vmas+254/752]  [exit_mmap+236/576]  [mmput+111/208]  [do_exit+326/1152]  [do_group_exit+66/224]  [syscall_call+7/11]
Jul 15 01:53:44 meta kernel: Trying to fix it up, but a reboot is needed
Jul 15 01:54:34 meta kernel:
Jul 15 01:54:34 meta kernel: avc:  denied  { use } for  pid=3864 exe=/usr/bin/apt-get path=/dev/pts/0 dev= ino=2 scontext=system_u:system_r:apt_t tcontext=root:staff_r:newrole_t tclass=fd
Jul 15 01:54:39 meta kernel: Bad page state at prep_new_page
Jul 15 01:54:39 meta kernel: flags:0x01000004 mapping:00000000 mapped:1 count:0
Jul 15 01:54:39 meta kernel: Backtrace:
Jul 15 01:54:39 meta kernel: Call Trace: [bad_page+93/144]  [prep_new_page+64/96]  [buffered_rmqueue+238/416]  [__alloc_pages+145/784]  [do_anonymous_page+135/752]  [handle_mm_fault+262/432]  [do_page_fault+604/1145]  [do_brk+313/544]  [sys_brk+222/272]  [do_page_fault+0/1145]  [error_code+45/56]
Jul 15 01:54:39 meta kernel: Trying to fix it up, but a reboot is needed
Jul 15 01:54:39 meta kernel: Unable to handle kernel paging request at virtual address 00010004
Jul 15 01:54:39 meta kernel:  printing eip:
Jul 15 01:54:39 meta kernel: c014d58d
Jul 15 01:54:39 meta kernel: *pde = 00000000
Jul 15 01:54:39 meta kernel: Oops: 0000 [#1]
Jul 15 01:54:39 meta kernel: CPU:    0
Jul 15 01:54:39 meta kernel: EIP:    0060:[page_add_rmap+125/320]    Not tainted
Jul 15 01:54:39 meta kernel: EFLAGS: 00010246
Jul 15 01:54:39 meta kernel: eax: 01008004   ebx: c10ca288   ecx: 00010000   edx: c1000000
Jul 15 01:54:39 meta kernel: esi: c5def8c0   edi: c5ea2e14   ebp: c10ca288   esp: c51edeb4
Jul 15 01:54:39 meta kernel: ds: 007b   es: 007b   ss: 0068
Jul 15 01:54:39 meta kernel: Process dpkg (pid: 3879, threadinfo=c51ec000 task=c4ff1920)
Jul 15 01:54:39 meta kernel: Stack: c5ea2e14 00000000 050dd067 c0148b45 00000020 c5def8c0 00000207 cd8a9084
Jul 15 01:54:39 meta kernel:        ceb6f9c0 08785274 c01491f6 ceb6f9c0 cb4976e0 c5ea2e14 cd8a9084 00000001
Jul 15 01:54:39 meta kernel:        00000e14 00000286 ceb6f9c0 ceb6f9e0 cb4976e0 c4ff1920 c011aaec ceb6f9c0
Jul 15 01:54:39 meta kernel: Call Trace: [do_anonymous_page+501/752]  [handle_mm_fault+262/432]  [do_page_fault+604/1145]
 [do_brk+313/544]  [sys_brk+222/272]  [do_page_fault+0/1145]  [error_code+45/56]
Jul 15 01:54:39 meta kernel: Code: 8b 41 04 85 c0 75 53 8b 01 89 c2 83 e2 07 89 3c 91 48 89 01
Jul 15 01:54:39 meta kernel:  <6>note: dpkg[3879] exited with preempt_count 2

--
This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to [email protected] with the words "unsubscribe selinux" without quotes as the message.

#1 lore raw
Quoted 5 lines · Colin Walters

Hi,

Linux oopsed while loading a policy today on my SELinux 2.5.74 machine. It doesn't offhand look directly related to SELinux, but I was curious if anyone here had seen it before.

I've not seen anything like this with 2.5. And according to your report, it seems that dpkg was running when the oops occurred.

- James

Signature · 8 lines
James Morris
<[email protected]>


--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.
#2 lore raw
Quoted 2 lines · James Morris

I've not seen anything like this with 2.5. And according to your report, it seems that dpkg was running when the oops occurred.

That's true, but I think dpkg is not at fault; besides, any kind of crash is a kernel bug. The only thing I can think of that's even slightly unusual about this machine is that it's SMP (dual PII-350).

--
This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to [email protected] with the words "unsubscribe selinux" without quotes as the message.

#3 lore raw
In reply to Colin Walters
Quoted 3 lines · Colin Walters

Linux oopsed while loading a policy today on my SELinux 2.5.74 machine. It doesn't offhand look directly related to SELinux, but I was curious if anyone here had seen it before.

I haven't seen it, but that doesn't mean it isn't our fault ;) You might want to try with 2.6.0-test1-selinux1, get it from
http://www.nsa.gov/selinux/lk/2.6.0-test1-selinux1.patch.gz. The other patch files in the lk directory are simply the individual logical change sets, i.e. one to address a problem with setxattr and one to add the selinux module.

Signature · 8 lines
Stephen Smalley <[email protected]>
National Security Agency


--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.