| 1 |
Hi, |
| 2 |
|
| 3 |
On Fri, 14 Aug 2015 10:54:57 -0400 Rich Freeman wrote: |
| 4 |
> On Fri, Aug 14, 2015 at 8:45 AM, Kristian Fiskerstrand <k_f@g.o> wrote: |
| 5 |
> > They will be OpenPGP signed by a releng key during thickening and |
| 6 |
> > portage will auto-verify it using gkeys once things are in place. As |
| 7 |
> > such checksum for ebuilds and other files certainly needs to be part |
| 8 |
> > of the manifest, otherwise it can open up for malicious alterations of |
| 9 |
> > these files. |
| 10 |
> > |
| 11 |
> |
| 12 |
> As much as I'd love to see it all folded into git, the reality is also |
| 13 |
> that git signatures are only bound to files by a series of sha1 |
| 14 |
> hashes, and sha1 is not a strong hash function. Git really ought to |
| 15 |
> move to sha256 at some point, preferably in a manner that makes it |
| 16 |
> expandable in the future to other hash functions. But, this isn't a |
| 17 |
> high-priority for upstream. |
| 18 |
> |
| 19 |
> The same limitation is true of any git gpg signature, including tag |
| 20 |
> signatures. It is all held together by sha1. The manifest system is |
| 21 |
> much stronger. |
| 22 |
|
| 23 |
OK, if manifests are that important, why not generate full manifest |
| 24 |
during repoman commit? If we do not tamper with $Id$, the only file |
| 25 |
outside of this manifest will be ChangeLog generated during rsync |
| 26 |
propagation. Then we have following options: |
| 27 |
- do not sing ChangeLog: even if it will be tampered, little harm |
| 28 |
can be done, since it doesn't affect live system or build process; |
| 29 |
- sign ChangeLog with releng key; |
| 30 |
- sign developer-signed manifest + ChangeLog with releng key. Thus |
| 31 |
we'll have double signature for most important files. |
| 32 |
|
| 33 |
Best regards, |
| 34 |
Andrew Savchenko |
Archives