| 1 |
On 03/05/15 14:14, Patrick Schleizer wrote: |
| 2 |
>> I used the footnote numbers to reference the attacks. |
| 3 |
> |
| 4 |
> I am afraid, this might cause some confusion. The numbers you have used |
| 5 |
> won't stay stable. Those were autogenerated numbers of footnotes. As |
| 6 |
> footnotes change, these numbers change. To keep your post |
| 7 |
> understandable, I created a snapshot before modifying footnotes: |
| 8 |
> http://www.webcitation.org/6Wo9Cb2ox |
| 9 |
> |
| 10 |
> However, numbers (1), (2), (3), etc. that won't be automatically |
| 11 |
> changed, have just been added now. |
| 12 |
|
| 13 |
HA, my fault, sorry about that. |
| 14 |
> |
| 15 |
> Actually, I was aware of it. The issue is, signing is not everything. |
| 16 |
> Signatures need a validity range. Otherwise mirrors can also show half a |
| 17 |
> year etc. old signatures that are valid. See also: |
| 18 |
> http://blog.ganneff.de/blog/2008/09/23/valid-until-field-in-release-f.html |
| 19 |
> |
| 20 |
>> attacks 3, 11, and 12. |
| 21 |
> |
| 22 |
> There was no attack 3. Now, before we talk past each other, would you |
| 23 |
> mind to repost by referencing attack by name or by their new, "real" |
| 24 |
> numbers? |
| 25 |
|
| 26 |
Well now there is an attack 3 :-) |
| 27 |
|
| 28 |
webrsync-gpg would ALERT the user in the case of attack 3. Portage |
| 29 |
checks the timestamp file from inside the gpg signed snapshot and alerts |
| 30 |
the user if the last sync was too long ago. In the case of a freeze |
| 31 |
attack, it should alert the user that the sync is old. Not that it |
| 32 |
fixes the attack, but it won't go unnoticed for long. |
| 33 |
|
| 34 |
Likewise, webrsync-gpg would also completely prevent 7 and 8. The |
| 35 |
snapshot is the entire tree, so it is not possible to "mix and match" |
| 36 |
without modifying the signed tarball. Nor would it be possible to |
| 37 |
provide the user with the wrong package since the checksums from inside |
| 38 |
the signed portage tarball wouldn't match if the source package was |
| 39 |
traded for a different one. |
| 40 |
|
| 41 |
Attack 8 shouldn't work either, as the checksums for everything are |
| 42 |
protected in the gpg signed portage tarball it should not be possible to |
| 43 |
trade out a source tarball for a different one. While gentoo doesn't |
| 44 |
typically use binary packages, it does have the support for it. Nothing |
| 45 |
in the binary package installation is protected by signatures, so I |
| 46 |
would expect this kind of attack to be possible when installing a binary |
| 47 |
package. Might be non-trivial to do, but certainly possible. |
| 48 |
|
| 49 |
Attack 9 won't work either, assuming the user can avoid the freeze |
| 50 |
attack (3) the user will have an up to date knowledge of what it expects |
| 51 |
on the mirrors and will simply ask other mirrors for the right file in |
| 52 |
case of 404 or bad checksum. |
| 53 |
|
| 54 |
tl;dr webrsync-gpg is a built in feature of the package manager which |
| 55 |
OPTIONALLY adds a significant amount of security against the attacks |
| 56 |
described on your website. This is not currently the default setting, |
| 57 |
however, it is described in many hardening guides for gentoo and widely |
| 58 |
used among the security conscious. |
| 59 |
|
| 60 |
-Zero_Chaos |
Archives