-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 12/05/2010 02:38 AM, Casey Schaufler wrote: > On 12/4/2010 1:34 PM, Eric Paris wrote: >> On Fri, 2010-12-03 at 20:20 -0800, Casey Schaufler wrote: >>> On 12/3/2010 1:45 PM, Eric Paris wrote: >>>> SELinux would like to implement a new labeling behavior of newly created >>>> inodes. We currently label new inodes based on the parent and the creating >>>> process. This new behavior would also take into account the name of the >>>> new object when deciding the new label. This is not the (supposed) full path, >>>> just the last component of the path. >>> I see. Pathname based controls. In SELinux. >> Actually you of (almost) all people shouldn't make this mistake. > > Sorry Eric, but if it looks like a duck, quacks like a duck, and > bites like a duck it's a good bet that what you have is a duck. > Better get out the Grand Mariner and start on a sauce. > >> There >> is absolutely no pathname based controls being implemented. This is an >> extension of the object labeling model to facilitate and simply some >> userspace issues related to the labeling of NEW objects. > > The end to which the controls are put does not change the character > of those controls. If the attribute of the file being used is the > name component of the name/inode pair in a directory entry than what > you have is a pathname based control mechanism. It will have all of > the downsides, including mount points and hard links, that any other > name based scheme will suffer from. > >> SELinux very >> much so is and will continue to be based solely on label based controls. > > I'm not sure that I buy that. The very presence of restorecond, which > sets labels based strictly on pathnames, begs the observer to question > whether an SELinux system (e.g. a RedHat distribution) can be truly > said to be strictly label based. Yes, individual access control decisions > are made based on the combination of the policy and the labels, but > with the suggested changes it becomes very difficult to distinguish the > enforcement of the kernel (e.g. creating "passwd" in a etc_t directory) > where restorecond is assumed to have set "/etc" to etc_t from checking > the same access rights on a file that happens to be named "/etc/passwd". > Yes, the kernel component of SELinux relies strictly on the labels, > but the reality is that SELinux is heavily dependent on the user space > component to maintain the proper labels on files so that the specified > policy is rational. > >> The intention is to remove some particularly gross userspace hacks >> related to new object labeling (read udev/restorecond/anything to do >> with /var/run, etc). It simplifies userspace, removes numerous races, >> and does so with no reduction in security (and theoretically the >> possibility of a more secure system) > > These "userspace hacks" are no "worse" than restorecond from a > security perspective. They have the unpleasant characteristic that > they are outside the control of the primary distributor of SELinux. > I will admit that the suggested changes will make SELinux better. > That does not make them any less pathname based. > > >>>> This is very useful because creating /etc/shadow is different than creating >>>> /etc/passwd but the kernel hooks are unable to differentiate these >>>> operations. >>> If I recall correctly, this has been the Apparmor/TOMOYO argument all along. >> They are similar arguments but not at all the same. Some people may not >> understand the distinction between the initial labeling an object and >> access controls. If that distinction is not clear to anyone please let >> me know. They talk about access controls based on the name. I'm only >> talking about new object labeling. > > Initial labeling of objects is critical. There is no point in > talking about the access control decision if you can't say good > things about how the label gets set initially. > >>>> We currently require that userspace realize it is doing some >>>> difficult operation like that and that userspace jump through SELinux hoops >>>> to get things set up correctly. This patch does not implement new >>>> behavior, that is obviously contained in a seperate SELinux patch, but it >>>> does pass the needed name down to the correct LSM hook. If no such name >>>> exists it is fine to pass NULL. >>> Why can't you use the existing pathname hooks? >> Because the pathname hooks are about pathname based access controls, >> which I contend for both philosophical and technical reasons are not >> appropriate. Discussed more later. > > Sigh. Argued more later. > >>> If you want pathname based behavior there are already two perfectly good >>> LSMs that do that. Why do we need a third? Especially when the people who >>> run SELinux have been so adamant that name based access controls have no >>> place in the kernel? >> Good thing I'm not adding name based access controls :) > > I contend that you are, but as always I am willing to learn > why it is I am wrong. > >>> Sure, you're just talking about the final component, but that's because >>> you've already done user space labeling of the entire file system and are >>> counting on standardized directory structures. Let's face it, outside the >>> context of a specific distribution and an SELinux policy tailored to that >>> distribution, final component name based controls do not have any kind of >>> generality. >> It's true that a relatively standard directory and naming structure is >> helpful in getting the refpolicy to work. But the tweaks to get SELinux >> running on a new distro (assuming it uses PAM, sorry slackware) should >> not huge. If anything your argument is against pathname based systems, >> not against label based LSMs or the patch in question. I'm not sure >> what point you were trying to make here. > > My point is roughly centered around the distinction between a Linux > kernel with SELinux configured and a policy installed and an SELinux > system with the aforementioned and a complete SELinux runtime that > includes goodies like restorecond. The former is strictly label based, > while the latter is definitely not. I really think that you should > leave the name based bits in userspace unless you are willing to > accept their general value in the kernel. > >>> If you really have seen the light and believe that pathname based access >>> controls have merit a scheme that is half based on labels and half based >>> on final components can't possibly seem satisfactory. >> Sadly I'm both still living in the dark and I agree with that statement >> completely. Pathname based access controls are not the best and a >> hybrid model is no better. > > I'm not saying that darkness is inherently wrong, but I am suggesting > that blindness is a condition that ought to be corrected where > possible. If you want to bring some of what is awkward to do in > userspace into the kernel that is fine, but I tend to view things > from a simplistic viewpoint, and if the attribute you're using is the > name of the thing I have to say that the mechanism you're using is > name based. > >>> Or am I missing something? >> You are (and I'm guessing rather intentionally :-P) > > OK, yes, there is some attempt at irony in my first response. > If you wanted to you could find some bits of Smack (look in > smack_d_instantiate if you're in a hurry) that could be considered > at least as questionable. > > I don't see a significant difference from a security standpoint > between creating "shadow" in a etc_t directory and creating > "/etc/shadow", because "everyone knows" that "/etc" gets etc_t, > and the userspace tools enforce it. > >> [next message explains my thoughts] >> >> On Sat, 2010-12-04 at 17:01 +0900, Tetsuo Handa wrote: >>> /etc/shadow as an absolute pathname makes special meaning but etc_t + shadow >>> does not always make special meaning. Using "struct dentry" instead of >>> "struct qstr" is not sufficient, for it would be sufficient for /etc/shadow >>> case because /etc/ directory is known to be within / partition but it is not >>> sufficient for /etc/foo/etc/shadow case because /etc/foo/ directory is not >>> always within / partition. Assigning shadow_t when creating /etc/foo/etc/shadow >>> (where the pathname derived from "struct dentry" is /etc/shadow) is wrong. >> (I should start by saying that there will likely never be a rule related >> to /etc/shadow since it is rarely a newly created object, but I'm >> willing to use it as my example, since most people reasonably understand >> the security implications of it.) >> >> You actually point out exactly the flaw of pathname based security, not >> a problem with my new object labeling extension. If the program >> responsible for creating /etc/shadow is able to be tricked into >> creating /etc/foo/shadow instead, that file is still going to contain >> passwords and information that should be protected just like /etc/shadow >> needs to be protected. > >> It's irrelevant if the pathname is wrong. > > Well, no, that's not true. The data will not be available to its > legitimate consumers if it is created in /etc/foo/shadow, even if > /etc/foo is (somehow) labeled etc_t. > >> It's >> about the data and the integrity/confidentiality of the data. > > This is also true, but it is not the only concern. If a /etc/foo > is on removable media the integrity and confidentiality, as defined > by the label of /etc/foo, is not going to mean all that much. > >> If that >> file contains password data (or at least what something/someone thought >> was password data) it should be protected exactly the same as the real >> one. We are at diametrically opposed view points on this and I doubt we >> ever come to agree. This patch makes no changes to SELinux in that >> regard and I don't see such a change in the future. > > No, the bulk of the pathname based controls of an SELinux distribution > are done in userspace. The SELinux kernel does not currently care. > That's fine. I suggest that all the pathname based behavior of SELinux > stay in the userspace components. > >> I don't understand some of your logic (why does /etc have to be on the >> same partition as /? How can a useful struct dentry, much less a struct >> path [I assume you meant path rather than vfsmount], exist before the >> struct inode?) but those parts seem irrelevant. Pathname based security >> is (at least in my view) fundamentally flawed if you want to make real >> security claims. Label based security can make real guarantees (ok >> Casey insert you snide comment about how SELinux policy is so large its >> hard->impossible to analyze/make claims) but when implemented on >> flexible systems (as opposed to to static systems such as those used in >> 3 letter gov't agencies) labeling is hard to do and this change makes it >> easier. >> >> Let me give some examples of where I plan to actually use this new >> behavior. >> >> devtmpfs/udev: currently devtmpfs creates new char and block files >> 'magically' in /dev. It does so with the label device_t because it >> can't possibly know better. It then calls up to hotplug, which call >> udev, which does a pathname based check in userspace, and resets the >> label to something better. How horrific is that? With this patch the >> kernel can tell that mem != console and can just do the right thing, >> race free. 'Really secure' installations don't use devtmpfs/udev and >> instead have static premade /dev with proper labeling (yeah MAKEDEV!). >> I'm fine if you are crazy enough to do that, but I'd rather live in the >> modern world with tools that just work. >> >> /var/run: Lots of programs used to (and many still do) create files >> directly in /var/run/. We in the fedora land have urged packages to >> use /var/run/[package name] because it made labeling easier to get >> right. This is SELinux forcing it's way of life onto userspace. >> Something we would rather avoid if it's possible to automate. >> >> ~/.ssh: When a user or even sshd creates the ~/.ssh directory the >> kernel doesn't realize that this is different than creating the >> directory "Downloads." How do we solve that today? We have a program >> that puts an inotify watch on ~/.ssh when you log in and which will, as >> fast as it can, relabel that directory to the correct label. How much >> crap is that? > > You could fix these applications. That's the right thing to do. > I didn't say it's easy, just that its right. > >> SELinux from Red Hat out of the box (I can't speak definitely for not RH >> based distros) fails on one of the major tenets of its own security >> model. It fails at label tranquility (aka once an object is labeled >> that label never changes). We have things which automatically and >> dynamically relabel existing objects (I just mentioned two of them.) >> I'd like to remove both of those exceptions, simplify userspace >> applications, and in doing so bring the flexibility of a real usable >> dynamic system to the security of one which meets something closer to a >> statically labeled system.... >> >> Make sense? > > Sure, but if it is a duck ... > >> -Eric > > BTW: I'm not saying your goals are bad, but I do question the > implementation. It really looks like the pathname hooks are your > easiest and most correct approach. > > -- > This message was distributed to subscribers of the selinux mailing list. > If you no longer wish to subscribe, send mail to majordomo@xxxxxxxxxxxxx with > the words "unsubscribe selinux" without quotes as the message. > > I really don't want to get into this that deeply, but I want to end a couple of misconceptions. restorecond the init service has not been run by default in several releases on Fedora, and does not run by default on RHEL6. I am not sure if it does on RHEL5. If you run it, it only watches these files. /etc/services /etc/resolv.conf /etc/samba/secrets.tdb /etc/mtab /var/run/utmp /var/log/wtmp /root/* /root/.ssh/* Most of these were added do to either bugs in rpm post install scripts (vmware) bugs in policy, or admins constantly screwing up when creating files. Most of these are historical and not needed any longer. I have added a userspace restorecond in Fedora 12-14 and RHEL6 that watches for the user creating files in his home dir. Expecting users to understand SELinux and to set the labels correctly so system services will work, just does not work. Lately we have been seeing more and more race condition avcs where the kernel creates a device in /dev and udev does not fix it quick enough before an AVC is generated. Bluetooth seems to get hit by this a lot. SELinux has always been partially path based. You need a way to get initial labels onto the disk, and the enforcement is Inode based. You are just adding a new way to get the labels on disk. Either through RPM, Udevd, fixfiles, restorecon, restorecond. Eric is just giving a new tool to the policy writer that he can use to get initial labels on disk. The only place I would envision using this to start would be for /dev and maybe /root and ~/. We can sit here and argue about what is pragmatic to do. I have always cared about usability and what every I can do, to make the system more usable without lowering security I am looking to do. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/ iEYEARECAAYFAkz89BEACgkQrlYvE4MpobMbigCg6HShTT0SAv+TLTMC22XVxjFE 2ZcAn21EIp6dmn2YmyUh4PJZ2ptg2W+X =1HpR -----END PGP SIGNATURE----- -- This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to majordomo@xxxxxxxxxxxxx with the words "unsubscribe selinux" without quotes as the message.
