SELinux-2.6-PPC
First of all, congrats to the NSA guys for getting SELinux merged into the linus tree. Thanks for all your hard work. I've got a SELinux kernel running on my PPC now, so we'll see how it goes on a new arch.
I've got a quick question about the new userland, are all of the future releases planned to have the split up libselinux, checkpolicy, policycoreutils, etc packages? I just want to be sure; I don't want to create packages for these, and then find that future releases get remerged into a new selinux-small :)
Linux version 2.6.0-test2-bk3 (root@beta) (gcc version 3.2.3 20030422 (Gentoo Linux 1.4 3.2.3-r2, propolice)
Found a Grand Central mac-io controller, rev: 2, mapped at 0xfdf00000 PowerMac motherboard: Unknown PowerSurge
Cache coherency enabled for bandit/PSX
Found Bandit PCI host bridge at 0xf2000000. Firmware bus number: 0->0 On node 0 totalpages: 65536
DMA zone: 65536 pages, LIFO batch:16
Normal zone: 0 pages, LIFO batch:1
HighMem zone: 0 pages, LIFO batch:1
Building zonelist for node : 0
Kernel command line: root=/dev/sda9 ro l2cr=0xab000000 console=tty0 console=ttyS0,38400
l2cr set to ab000000
System has 32 possible interrupts
PID hash table entries: 2048 (order 11: 16384 bytes)
GMT Delta read from XPRAM: -360 minutes, DST: offvia_calibrate_decr: ticks per jiffy = 131246 (787481 ticks)
Console: colour dummy device 80x25
Calibrating delay loop... 1048.57 BogoMIPS
Memory: 255808k available (1812k kernel code, 740k data, 276k init, 0k highmem)
Security Scaffold v1.0.0 initialized
SELinux: Initializing. SELinux: Starting in permissive mode
There is already a security framework initialized, register_security failed.
Failure registering capabilities with the kernelselinux_register_security: Registering secondary module capability Capability LSM initialized
Signature · 12 lines
Chris PeBenito <[email protected]> Developer, SELinux Hardened Gentoo Linux Public Key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE6AF9243 Key fingerprint = B0E6 877A 883F A57A 8E6A CB00 BC8E E42D E6AF 9243 -- 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.
Quoted 2 lines · Chris PeBenito
First of all, congrats to the NSA guys for getting SELinux merged into the linus tree. Thanks for all your hard work.
Thanks also to James Morris for his help.
Quoted 2 lines
I've got a SELinux
kernel running on my PPC now, so we'll see how it goes on a new arch.
I would expect it to work fine on PPC, as there isn't anything architecture-specific in the SELinux module anymore (no sys_security), and the architecture-specific elements of LSM (particularly the patch to the linker script for the security initcall) should be integrated for most architectures.
Quoted 5 lines
I've got a quick question about the new userland, are all of the future releases planned to have the split up libselinux, checkpolicy, policycoreutils, etc packages? I just want to be sure; I don't want to create packages for these, and then find that future releases get remerged into a new selinux-small :)
Yes, we plan on continuing to provide these as individual packages. I expect that you can use Colin Walter's Debian packages as a starting point for Gentoo packages. Caveat: Due to the migration to xattr and the reduction in support for genfs_contexts, you'll need to implement file labeling support if you want to use certain filesystem types (e.g. reiser, devfs).
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.
Quoted 6 lines · Stephen Smalley
Yes, we plan on continuing to provide these as individual packages. I expect that you can use Colin Walter's Debian packages as a starting point for Gentoo packages. Caveat: Due to the migration to xattr and the reduction in support for
genfs_contexts, you'll need to implement file labeling support if you want to use certain filesystem types (e.g. reiser, devfs).
What exactly do you mean by a reduction of genfs_contexts support? I didn't see any mention of this in the docs (perhaps I missed it), but I did notice that genfs_contexts is a lot slimmer than the old policy. I do have devfs in mind, as its used in Gentoo. As for reiser I would think that for it to work it would just need the appropriate xattr support in the kernel (and fix the policy makefile to run setfiles on it during a relabel).
Signature · 12 lines
Chris PeBenito <[email protected]> Developer, SELinux Hardened Gentoo Linux Public Key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE6AF9243 Key fingerprint = B0E6 877A 883F A57A 8E6A CB00 BC8E E42D E6AF 9243 -- 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.
Quoted 4 lines · Chris PeBenito
What exactly do you mean by a reduction of
genfs_contextssupport? I didn't see any mention of this in the docs (perhaps I missed it), but I did notice thatgenfs_contextsis a lot slimmer than the old policy. I do have devfs in mind, as its used in Gentoo.
See the comments at the beginning of the genfs_contexts file in the 2.5 policy tree. Also mentioned in the policy part of the PORTING file in selinux-doc. The code that attempted to reconstruct a pathname relative to the top of the mount for auditing and for genfs_contexts had to be dropped as part of the cleanup for mainline inclusion; you can't reliably map an inode to a pathname in general. For proc, we can use the proc_dir_entry tree to generate paths within proc to identify nodes in the policy in order to map them to security contexts. Something similar might be possible in devfs, but you also likely need an xattr handler for it to support userspace management of security contexts on device nodes. You might want to look into udev as a replacement for devfs for 2.6.
Quoted 4 lines
As for reiser I would
think that for it to work it would just need the appropriate xattr support in the kernel (and fix the policy makefile to run setfiles on it during a relabel).
Yes, you would need to implement an xattr handler for the security namespace in reiser. You can likely leverage the existing reiser xattr patches, but you'll still need to add the handler for the security namespace.
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.