Two high-profile kernel bugs—with publicly released exploits—have recently been making news. Both can be used by local attackers to gain root privileges, which makes them quite dangerous for systems that allow untrusted users to log in, but they might also be used in conjunction with other flaws to produce a remote root exploit. They are, in short, just the kinds of vulnerabilities that most system administrators would want to patch quickly, so a look at how distributions have responded seems warranted.
The vulnerabilities lie in the x86_64 compatibility layer that allows 32-bit binaries to be run on 64-bit systems (see the Kernel page article for more details). In particular, that code allows 32-bit programs to make system calls on a 64-bit kernel. One of the bugs, CVE-2010-3301, was reintroduced into the kernel in April 2008, seven months after being fixed as CVE-2007-4573. The second, CVE-2010-3081, was discovered in the process of finding the first, and had been in the kernel since 2.6.26, which was released in July 2008.
Obviously, these are long-standing kernel holes that may have been available to attackers for as long as two-and-a-half years. In fact, a posting on the full-disclosure mailing lists claims that CVE-2010-3081 was known by a shadowy group called "Ac1db1tch3z" since the code causing the bug was committed in April 2008. Included in the post was a working exploit.
Ben Hawkes found and reported both of the vulnerabilities and fixes for both were quickly committed to the mainline kernel. Both were committed on September 14, but it is clear that at least CVE-2010-3081 was known a week earlier. Stable kernels were released on September 20, and some distributions had fixes out on September 17.
While enterprise kernels (e.g. RHEL, SLES) tend to be based on fairly old kernels (RHEL 5 is 2.6.18-based, while SLES 11 is 2.6.27-based), the distributors often backport features from newer kernels. Unsurprisingly, that can sometimes lead to backporting bugs along with the features. For RHEL, that meant that it was vulnerable to CVE-2010-3081 even though that code came into the kernel long after 2.6.18. On September 21, Red Hat issued updates for RHEL 5 and 5.4 to fix that problem. CentOS, which necessarily lags RHEL by at least a few hours, has a fix for CVE-2010-3081 available for CentOS 5 now as well.
For SLES, the situation is a little less clear. Based on its kernel version, it should be vulnerable to both flaws, but no updates have been issued as of this writing. In a separate advisory on September 21, SUSE noted (in the "Pending Vulnerabilities ..." section) that it was working on fixes for both.
For the more community-oriented distributions (Debian, Fedora, openSUSE, Ubuntu, and others), the response has been somewhat mixed. Ubuntu, Debian, and Fedora had fixes out on September 17 for both bugs (or, in the case of Debian, just one, as its stable distribution ("Lenny") is based on 2.6.26 and thus not vulnerable CVE-2010-3301). openSUSE has yet to release a fix and none of the secondary distributions that we track (Gentoo, Mandriva, Slackware, etc.) has put out a fix either.
How quickly can users and administrators expect security fixes? The enterprise vendors, who are typically more cautious before issuing an update, took a week or more to get fixes in the hands of their users. Meanwhile, exploits have been published and have been used in the wild. That has to make a lot of system administrators very nervous. Those running SUSE-based systems must be even more worried. A one-week delay (or more depending on where you start counting) may not seem like a lot of time, but for critical systems, with lots of sensitive data, it probably seems pretty long.
Local privilege escalation flaws are often downplayed because they require a local user on the system to be useful. But that thinking has some flaws of its own. On even a locked-down system, with only trusted users being able to log in, there may be ways for local exploits to be turned into remote exploits. Compromised user accounts might be one way for an attacker to access the system, but there is a far more common route: network services.
One badly written, or misconfigured, web application, for example, might provide just the hole that an attacker needs to get their code running on the system. Once that happens, they can use a local privilege escalation to compromise the entire system—and all the data it holds. Since many servers sit on the internet and handle lots of web and other network traffic, compromising a particular, targeted system may not be all that challenging to a dedicated attacker. Using a "zero day" vulnerability in a widely deployed web application might make a less-targeted attack (e.g. by script kiddies) possible as well.
While most of the "big four" community distributions were quick to get out updates for these problems, they still left a window that attackers could have exploited. That is largely unavoidable unless there were embargoes enforced on sensitive patches flowing into the mainline kernel—something that Linus Torvalds has always been opposed to. Then, of course, there is the (much larger) window available to those who closely track kernel development and notice these bugs as they get introduced.
That is one of the unfortunate side-effects of doing development in the open. While it allows anyone interested to look at the code, and find various bugs—security or otherwise—it does not, cannot, require that those bugs get reported. We can only hope that enough "white hat" eyes are focused on the code to help outweigh the "black hat" eyes that are clearly scrutinizing it.
For distributors, particularly Red Hat and Novell, it may seem like these flaws are not so critical that the fixes needed to be fast-tracked. Since there are, presumably, no known, unpatched network service flaws in the packages they ship, a local privilege escalation can be thwarted by a more secure configuration (e.g. no untrusted users on critical systems). While true in some sense, it may not make those customers very happy.
There are also plenty of systems out there with some possibly untrusted users. Perhaps they aren't the most critical systems, but that doesn't mean their administrators want to turn them over to attackers. It really seems like updates should have come more quickly in this case, at least for the enterprise distributions. As we have seen, a reputation for being slow to fix security problems is a hard one to erase; hopefully it's not a reputation that Linux is getting.
Fusion Garage, makers of the Linux-based joojoo tablet, are still out of compliance with the GPL, but have responded to Linux kernel developer Matthew Garrett regarding his complaint this month to US Customs and Border Protection. Until Garrett filed a so-called "e-allegation" form with CBP, the company had refused his requests for corresponding source for the GPL-covered software on the joojoo. After the filing, company spokesperson Megan Alpers said that "Fusion Garage is discussing the issue internally."
The joojoo concept began as the "CrunchPad," in an essay by TechCrunch founder Michael Arrington. He wanted a "dead simple and dirt cheap touch screen web tablet to surf the web." After a joint development project between Fusion Garage and Arrington collapsed last year, Fusion Garage went on to introduce the tablet on its own.
Garrett said he is not currently planning on taking further action. "The company seems to be at least considering the issue at the moment, so I wasn't planning on doing anything further just yet," he said. Bradley Kuhn, president of the Software Freedom Conservancy, which carries out GPL enforcement, praised Garrett's effort on his blog:
In June, Garrett checked out a joojoo tablet and mailed the company's support address for the source to the modified kernel. He posted the company's reply to his blog: "We will make the source release available once we feel we are ready to do so and also having the resources to get this sorted out and organized for publication."
Although the device is fairly simple, Garrett said he would like to see some of the kernel changes. "They seem to be exposing ACPI events directly through the embedded controller. I'm interested to see how they're doing this and how drivers bind to it," he said.
The US government offers several tools for enforcing copyright at the border, which are cheaper and simpler for copyright holders than an infringement case in a Federal court. Beyond the "e-allegation" form that Garrett filed, other options include "e-recordation," which allows the holder of a US trademark or copyright to request that US Customs and Border Protection stop infringing goods at the border, and a Section 337 complaint, which can result in the US International Trade Commission assigning government attorneys to work on the US rights-holder's behalf.
"Customs will seize something if it's a clear knockoff product but they don't like to wade into disputes where it's not clear," said attorney Jeffery Norman of Kirkland & Ellis, a law firm that represents clients in Section 337 cases. "The ITC has jurisdiction to litigate disputes involving copyright, trademark and patent, but 99% of cases involve patents," he said.
While most of the ITC's copyright cases involve product packaging or manuals, Norman said, the ITC has taken action to exclude infringing copies of arcade game software. In a 1981 case, "Coin-Operated Audio Visual Games and Components Thereof" the ITC unanimously voted to exclude video games whose code infringed both copyright and trademarks of US-based Midway. In a 1982 case, also brought by Midway, the ITC not only voted to exclude infringing Pac-Man games, but issued a cease and desist order against the importers while the case was in progress.
In an LWN comment, Norman added, "Customs will only enforce either an ITC order or 'Piratical' [obviously infringing] copies." He advocated the ITC approach:
Open Source consultant Bruce Perens said that retailers in the USA could be a reason for manufacturers in other countries to check license compliance for their embedded software.
Although the prospect of Customs enforcement because of code they've never heard of might scare some vendors away from open-source-based products, Perens said, "the good news is that products like Android are so important to the market that retailers must learn to deal with Open Source or let their competitors have the business."
No GPL licensor has yet filed a complaint with the ITC, and even if a complaint is filed, the ITC can decide whether or not to act on it. However, import-based enforcement, including temporary orders enforced at the Customs level, can move much faster than ordinary infringement lawsuits. Whether the resulting uncertainty is enough to make device vendors double-check their license compliance remains to be seen.
Debian's "testing" distribution is where Debian developers prepare the next stable distribution. While this is still its main purpose, many users have adopted this version of Debian because it offers them a good trade-off between stability and freshness. But there are downsides to using the testing distribution, so the "Constantly Usable Testing" (CUT) project aims to reduce or eliminate those downsides.
Debian "unstable" is the distribution where developers upload new versions of their packages. But, frequently some packages are not installable from unstable due to changes in other packages or transitions in libraries that have not yet been completed.
Debian testing, on the contrary, is managed by a tool that ensures the consistency of the whole distribution: it picks updates from unstable only if the package has been tested enough (10 days usually), is free of new release-critical bugs, is available on all supported architectures, and it doesn't break any other package already present in testing. The release team controls this tool and provides "hints" to help it find a set of packages that can flow from unstable to testing.
Those rules also ensure that the packages that flow into testing are reasonably free of show-stopper bugs (like a system that doesn't boot, or X that doesn't work at all). This makes it very attractive to users who like to regularly get new upstream versions of their software without dealing with the biggest problems associated with them. That is very attractive to users, yet several Debian developers advise people to not use testing. Why is that?
Disappearing software: The release team uses the distribution to prepare the next stable release and from time to time they remove packages from it. That is done either to ensure that other packages can migrate from unstable to testing, or because a package has long-standing release-critical bugs without progress towards a resolution. The team will also remove packages on request if the maintainers believe that the current version of the software cannot be supported (security-wise) for 2 years or more. The security team also regularly issues such requests.
Long delays for security and important fixes: Despite the 10-day delay in unstable, there are always some annoying bugs (and security bugs are no exceptions) that are only discovered when the package has already migrated to testing. The maintainer might be quick to upload a fixed package in unstable, and might even raise the urgency to allow the package to migrate sooner, but if the packages get entangled in a large ongoing transition, it will not migrate before the transition is completed. Sometimes it can take weeks for that to happen.
The delay can be avoided by doing direct uploads to testing (through testing-proposed-updates) but that mechanism is almost never used except during a freeze, where targeted bug fixes are the norm.
Not always installable: With testing evolving daily, updates sometimes break the last installation images available (in particular netboot images that get everything from the network). The debian-installer (d-i) packages are usually quickly fixed but they don't move to testing automatically because the new combination of d-i packages has not necessarily been validated yet. Colin Watson sums up the problem:
CUT has its roots in an old proposal by Joey Hess. That introduces the idea that the stable release is not Debian's sole product and that testing could become—with some work—a suitable choice for end-users. Nobody took on that work and there has been no visible progress in the last 3 years.
But recently Joey brought up CUT again on the debian-devel mailing list and Stefano Zacchiroli (the Debian project leader) challenged him to setup a BoF on CUT for Debconf10. It turned out to be one of the most heavily attended BoFs (video recording is here), so there is clearly a lot of interest in the topic. There's now a dedicated wiki and an Alioth project with a mailing list.
There's general agreement that regular snapshots of testing are required: it's the only way to ensure that the generated installation media will continue to work until the next snapshot. If tests of the snapshot do not reveal any major problems, then it becomes the latest "cut". For clarity, the official codename would be date based: e.g. "cut-2010-09" would be the cut taken during September 2010.
While the frequency has not been fixed yet, the goal is clearly to be on the aggressive side: at the very least every 6 months, but every month has been suggested as well. In order to reach a decision, many aspects have to be balanced.
One of them (and possibly the most important) is the security support. Given that the security team is already overworked, it's difficult to put more work on their shoulders by declaring that cuts will be supported like any stable release. No official security support sounds bad but it's not necessarily so problematic as one might imagine. Testing's security record is generally better than stable's is (see the security tracker) because fixes flow in naturally with new upstream versions. Stable still get fixes for very important security issues earlier than testing, but on the whole there are fewer known security-related problems in testing than in stable.
Since it's only a question of time until the fixed version comes naturally from upstream, more frequent cut releases means that users get security fixes sooner. But Stefan Fritsch, who used to be involved in the Debian testing security team, has also experienced the downside for anyone who tries to contribute security updates:
So if it's difficult to form a dedicated security team, the work of providing security updates must be done by the package maintainer. They are usually quite quick to upload fixed packages in unstable but tend to not monitor whether the packages migrate to testing. They can't be blamed for that because testing was created to prepare the next stable release and there is thus no urgency to get the fix in as long as it makes it before the release.
CUT can help in this regard precisely because it changes this assumption: there will be users of the testing packages and they deserve to get security fixes much like the stable users.
Another aspect to consider when picking a release frequency is the amount of associated work that comes with any official release: testing upgrades from the previous version, writing release notes, and preparing installation images. It seems difficult to do this every month. With this frequency it's also impossible to have a new major kernel release for each cut (since they tend to come out only every 2 to 3 months) and the new hardware support that it brings is something worthwhile to many users.
In summary, regular snapshots address the "not always installable" problem and may change the perception of maintainers toward testing so that hopefully they care more of security updates in that distribution (and in cuts). But it does not solve the problem of disappearing packages. Something else is needed to fix that problem.
Lucas Nussbaum pointed out that regular snapshots of Debian is not really a new concept:
In Lucas's eyes, CUT becomes interesting if it can provide a rolling distribution (like testing) with a "constant flux of new upstream releases". For him, that would be "something quite unique in the Free Software world". The snapshots would be used as starting point for the initial installation, but the installed system would point to the rolling distribution and users would then upgrade as often as they want. In this scenario, security support for the snapshots is not so important, what matters is the state of the rolling distribution.
If testing were used as the rolling distribution, the problem of disappearing packages would not be fixed. But that could be solved with a new rolling distribution that would work like testing but with adapted rules, and the cuts would then be snapshots of rolling instead of testing. The basic proposal is to make a copy of testing and to re-add the packages which have been removed because they are not suited for a long term release while they are perfectly acceptable for a constantly updated release (the most recent example being Chromium).
Then it's possible to go one step further: during a freeze, testing is no longer automatically updated, which makes it inappropriate to feed the rolling distribution. That's why rolling would be reconfigured to grab updates from unstable (but using the same rules as testing).
Given the frequent releases, it's likely that only a subset of architectures would be officially supported. This is not a real problem because the users who want bleeding edge software tend to be desktop users on mainly i386/amd64 (and maybe armel for tablets and similar mobile products). This choice—if made—opens up the door to even more possibilities: if rolling is configured exactly like testing but with only a subset of the architectures, it's likely that some packages would migrate to rolling before testing where non-mainstream architectures are lagging in terms of auto-building (or have toolchain problems).
While being ahead of testing can be positive for the users, it's also problematic on several levels. First, managing rolling becomes much more complicated because the transition management work done by the release team can't be reused as-is. Then it introduces competition between both distributions which can make it more difficult to get a stable release out, for example if maintainers stop caring about the migration to testing because the migration to rolling has been completed.
The rolling distribution is certainly a good idea but the rules governing it must be designed to avoid any conflict with the process of releasing a stable distribution. Lastly, the mere existence of rolling would finally fix the marketing problem plaguing testing: the name "rolling" does not suggest that the software is not yet ready for prime time.
Whether CUT will be implemented remains to be seen, but it's off for a good start: ftpmaster Joerg Jaspert said that the new archive server can cope with a new distribution, and there's now a proposal shaping up. It may get going quickly as there is already an implementation plan for the snapshot side of the project. The rolling distribution can always be introduced later, once it is ready. Both approaches can complement each other and provide something useful to different kind of users.
The global proposal is certainly appealing: it would address the concerns of obsolescence of Debian's stable release by making intermediary releases. Anyone needing something more recent for hardware support can start by installing a cut and follow the subsequent releases until the next stable version. And users who always want the latest version of all software could use rolling after having installed a cut.
From a user point of view, there are similarities with the mix of normal and long-term releases of Ubuntu. But from the development side, the process followed would be quite different, and the constraints imposed by having a constantly usable distribution are stronger. With CUT, any wide-scale change must be designed in a way that it can happen progressively in a transparent manner for users.
Page editor: Jonathan Corbet
Next page: Security>>
Copyright © 2010, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds