Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 27 Jun 2019 13:05:08 -0400
From: Sasha Levin <>
Subject: Re: linux-distros membership application - Microsoft

On Thu, Jun 27, 2019 at 04:03:21PM +0200, Solar Designer wrote:
>Hi Sasha,
>Thank you for posting this application.
>Are you also on security@...?  If so, then on one hand also being on
>linux-distros would probably be of less use to you since I suspect most
>of the issues relevant to Microsoft are in the Linux kernel, but on the
>other hand you could serve as a liaison to that group.

I'm not a member of security@k.o, but that list isn't what we're looking
for. We've talked about this with Greg before sending this application,
and he has pointed out that security@k.o is not a disclosure list, but
rather just a way to pull in kernel folks to fix issues. Some (most?) of
the kernel bugs that get fixed don't go through that list to begin with.

The kernel's documentation with regards to security issues and
disclosure actually points to linux-distros: .

To complicate your question further: the Linux usage on our cloud has
surpassed Windows, as a by-product of that MSRC has started receiving
security reports of issues with Linux code both from users and vendors.
It's also the case that issues that are common for Windows and Linux
(like those speculative hardware bugs) are shared with us via MSRC as

If you think that there's value in connecting between these 3
entities, I'd be happy to do so (maybe as part of a new task).

>On Wed, Jun 26, 2019 at 10:13:58AM -0400, Sasha Levin wrote:
>> >1. Be an actively maintained Unix-like operating system distro with
>> >substantial use of Open Source components
>> Microsoft provides several distro-like builds which are not derivative
>> of an existing distribution that are based on open source components:
>> - Azure Sphere
>>   ( This
>>   Linux-based IoT device provides, among various things, security
>>   updates to deployed IoT devices. As the project is about to step out
>>   of public preview into the GA stage, we expect millions of these
>>   devices to be publicly used.
>This does sound like it includes a Linux distro, but obviously there's
>no security track record for it yet, so it's hard to estimate relevance
>of information you'd obtain via linux-distros to that product.

Indeed. When we debated linux-distros application internally, I tried to
do this estimate based on issues previously handled by linux-distros,
and it seemed like it would be valuable here.

With Sphere it's also hard to guess how are customers will be using it,
it's a rather unique mix of HaaS (the chip contains cores that would
belong to the user) and SaaS (where we provide mechanism to manage the
IoT device that the user will build on). I can go over this in more
detail if you'd like.

>> - Windows Subsystem for Linux v2
>>   (
>>   A Linux based distro that runs as a virtual machine on top of Windows
>>   hosts. WSL2 is currently available for public preview and scheduled
>>   for GA early 2020.
>You call this "a Linux based distro", but that's not clear to me from
>the web page you reference, which talks about managing (third-party?)
>distros in the WSL2 subsystem.  Can you clarify this, please?

With WSLv2, users can choose to bring in a third party distro userspace
to run on top our kernel+lightweight userspace. This means that users
can be running a Debian/Ubuntu/Fedora/etc userspace modified to run on
WSLv2, on top of a kernel we supply.

For us it means that we are now the proud owners of multiple
frankensteined distro userspaces as well as a custom kernel. You might
call it out as being derivative of those distros, but I believe that at
the very least this is similar to the OpenVZ/CloudLinux scenario.

>> - Products such as Azure HDInsight
>>   ( and the Azure
>>   Kubernetes Service
>>   (
>>   provide public access to a Linux based distribution.
>These look like SaaS offerings.  We do not have any explicit guidelines
>on whether "provide public access to a Linux based distribution" is
>potentially enough to qualify for linux-distros or not.  We do have
>Amazon as a member, so there's that precedent.

Honestly, these ones are even harder to deal with than the regular IaaS
offerings we have: customers who use these offerings expect (and
rightfully so) to be able to boot/update into a secure environment
moments after a public disclosure of a security issue, which is not
something we are able to provide right now.

>> >2. Have a userbase not limited to your own organization
>> Microsoft customers have millions of cores running the various workloads
>> described above.
>> >3. Have a publicly verifiable track record, dating back at least 1
>> >year and continuing to present day, of fixing security issues
>> >(including some that had been handled on (linux-)distros, meaning that
>> >membership would have been relevant to you) and releasing the fixes
>> >within 10 days (and preferably much less than that) of the issues
>> >being made public (if it takes you ages to fix an issue, your users
>> >wouldn't substantially benefit from the additional time, often around
>> >7 days and sometimes up to 14 days, that list membership could give
>> >you).
>> Microsoft has decades long history of addressing security issues via
>> MSRC ( While we are able to
>> quickly (<1-2 hours) create a build to address disclosed security
>> issues, we require extensive testing and validation before we make these
>> builds public. Being members of this mailing list would provide us the
>> additional time we need for extensive testing.
>It'd be helpful if you could directly address this part: "including some
>that had been handled on (linux-)distros, meaning that membership would
>have been relevant to you".  Without such examples yet, we'd have to be
>guessing whether the membership would have been relevant to you or not.
>Right now, the statistics at:
>only go until the end of 2018, so you'd be able to use them for examples
>dating back to 2018 and earlier.  We should ask Gentoo to update these
>statistics soon, perhaps for period until end of June 2019, which will
>be possible soon.

Sure! Issues on the stats page that would not have been reported to MSRC
but are relevant to us would include:

 - On the kernel side, issues such as CVE-2017-7533
  ( would be
  relevant for all our offerings.

 - Core libraries affect us as well, for example CVE-2017-1000408
   ( This
   would affect our Sphere and SaaS offerings, as well as probably make
   us run through them through test gauntlet for WSLv2.

 - Higher level Linux tools, such as the one in CVE-2018-14722
   ( affect
   mostly our IaaS offerings, but I expect that we'd again validate the
   rest of our offerings with the fix.

>> >4. Not be (only) downstream or a rebuild of another distro (or else we
>> >need convincing additional justification of how the list membership
>> >would enable you to release fixes sooner, presumably not relying on
>> >the upstream distro having released their fixes first?)
>> None of our builds are based on an existing distribution. For few of
>> these workloads we have a very custom kernel and userspace (such as for
>> Azure Sphere), while some share a more conventional kernel/userspace
>> configuration.
>> >5. Be a participant and preferably an active contributor in relevant
>> >public communities (most notably, if you're not watching for issues
>> >being made public on oss-security, which are a superset of those that
>> >had been handled on (linux-)distros, then there's no valid reason for
>> >you to be on (linux-)distros)
>> We follow closely public discussions with regards to security issues
>> that would affect us. While there was only a minor contribution back to
>> these lists mostly as we did not have any value to add back.
>> During past years I've reported multiple security issues which were
>> assigned CVEs.
>> >6. Accept the list policy (see above)
>> We accept the list's policy.
>> >7. Be able and willing to contribute back (see above), preferably in
>> >specific ways announced in advance (so that you're responsible for a
>> >specific area and so that we know what to expect from which member),
>> >and demonstrate actual contributions once you've been a member for a
>> >while
>> We understand this need and will be contributing back. Looking at the
>> list of vacant positions I can suggest the following, but I suspect that
>> existing list members will have better suggestions.
>> Technical:
>> 3. Review and/or test the proposed patches and point out potential
>> issues with them (such as incomplete fixes for the originally reported
>> issues, additional issues you might notice, and newly introduced bugs),
>> and inform the list of the work done even if no issues were encountered
>> - primary: Amazon, backup: vacant
>> Administrative:
>> 3. Evaluate if the issue (or one of the issues) is effectively already
>> public (e.g., a fix is committed upstream with a descriptive message)
>> or/and is low severity and thus the report (or its portion pertaining to
>> the issue) should be made public right away for one or both of these
>> reasons, get a few other list members to confirm this understanding, and
>> if there are no objections then communicate this strong preference to
>> the reporter - primary: CloudLinux, backup: vacant
>If Microsoft volunteers for these, I'd like that to be in "primary" role
>at least for the technical task of "3. Review and/or test the proposed
>patches ..."  I think Amazon hasn't been doing enough on that front,
>especially given the request to "inform the list of the work done even
>if no issues were encountered".  Given this request, if this were
>seriously worked on, I would have expected such reports from Amazon on
>almost every issue handled on linux-distros, but this wasn't the case.

Sure, we'd love to help with the list's pain points.

>I also would like a distro (maybe Microsoft) to volunteer for Technical:
>4. Check if related issues exist in the same piece of software (e.g.,
>same bug class common across the software, or other kinds of bugs exist
>in its problematic component), and inform the list either way
>and Administrative:
>4. Evaluate relevance to other parties such as the upstream, other
>affected distros (not present on the (sub-)list), and other Open Source
>projects, see if the report mentions notifying any of these, communicate
>your findings and possible concerns to the reporter and the list, and
>stay on top of the resulting discussion until a decision is made on who
>else to possibly notify (or not) and any such notifications are in fact
>made (with the reporter's approval)
>These are completely unclaimed now, but are much needed.
>For Technical "4. Check if related issues exist ...", we sometimes get
>some helpful for varying distros' package maintainers and such, but this
>is not consistent.  For example, recently Takashi Iwai of SUSE helped
>with Linux Marvell Wi-Fi driver issues - thanks! - but this is more of
>an exception than the rule.
>The lack of a volunteer distro for Administrative "4. Evaluate relevance
>to other parties ..." came up e.g. here:
>"Linux kernel: Bluetooth: two remote infoleaks (CVE-2019-3459, CVE-2019-3460)"

This could be interesting for us. we already work closely with multiple
distros as part of our public IaaS offering, as well as my work
maintaining the stable tree means I interact often with many subsystem
maintainers. We could leverage that for this task.

I think that this task would also benefit from collaboration with MSRC,
where for example we could verify whether the Bluetooth issue you brought
up would affect Windows, and whether issues reported to MSRC also affect

>> >8. Be able and willing to handle PGP-encrypted e-mail
>> I am able and willing to handle PGP-encrypted e-mail.
>> >9. Have someone already on the private list, or at least someone else
>> >who has been active on oss-security for years but is not affiliated
>> >with your distro nor your organization, vouch for at least one of the
>> >people requesting membership on behalf of your distro (then that one
>> >vouched-for person will be able to vouch for others on your team, in
>> >case you'd like multiple people subscribed)
>> Greg Kroah-Hartman <> would vouch for me
>> (Sasha Levin <>).
>Let's discuss this further, and perhaps arrive at a decision in July.

Great! Sounds like a plan.


Powered by blists - more mailing lists

Please check out the Open Source Software Security Wiki, which is counterpart to this mailing list.

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.