1 |
As someone interested in these happenings, I might point you at the most |
2 |
recent Gentoo Weekly Newsletter. |
3 |
|
4 |
http://www.gentoo.org/news/en/gwn/20040322-newsletter.xml |
5 |
|
6 |
In it glsa-check is mentioned as a new means to identify security issues |
7 |
in gentoo. It'll eventually me merged into emerge (heh) in the future. |
8 |
|
9 |
This doesn't do us much good if the issue is never declared as a GLSA |
10 |
though, which seems to be Ben's main problem here. |
11 |
|
12 |
Matt Hallmark |
13 |
|
14 |
Ben Cressey wrote: |
15 |
|
16 |
> Kurt Lieber wrote: |
17 |
> |
18 |
>>This would only be done in the most extreme cases, of which this is not |
19 |
>>one. |
20 |
> |
21 |
> |
22 |
> That policy makes sense, I admit. I'm not trying to blow this out of |
23 |
> proportion (and I notice that the new courier-imap ebuild is available in |
24 |
> stable at the moment), just trying to realign my assumptions about gentoo |
25 |
> and security with the reality of the process. |
26 |
> |
27 |
> I've been giving it some thought this morning, and I have a question. Would |
28 |
> it be feasible to add architecture-specific "security" keywords, such as |
29 |
> +x86 alongside ~x86 and x86? +x86 would have the following differences over |
30 |
> mainline x86: |
31 |
> |
32 |
> 1) Any ebuild with a known security flaw would be removed immediately. |
33 |
> 2) Old ebuilds would never "age out" of this keyword. |
34 |
> |
35 |
> Users could then set this keyword instead of the default x86, and could run |
36 |
> emerge with a "--securityupdate" option that would only suggest upgrading |
37 |
> packages that were installed and no longer part of +x86. |
38 |
> |
39 |
> Point #2 seems to be the most problematic, simply because older ebuilds are |
40 |
> routinely deleted as the portage tree is updated and new versions become |
41 |
> available. Two solutions come to mind for this difficulty: |
42 |
> |
43 |
> 1) A "stub" of the old ebuild could be left in portage; this ebuild would |
44 |
> not be enough to install the package, but would have the minimum set of |
45 |
> information required to act as a valid ebuild for that version, for the |
46 |
> purposes of the emerge --securityupdate check. |
47 |
> |
48 |
> 2) Alternatively, users with the +x86 keyword set would not have old ebuilds |
49 |
> deleted from their portage tree during the rsync process. |
50 |
> |
51 |
> The difficulty with point #2 would be that package maintainers would have no |
52 |
> way to go back and remove the +x86 keyword from an old version of the ebuild |
53 |
> if that ebuild was no longer in portage. The problem with #1 seems to be |
54 |
> that the number of ebuilds in the portage tree would continue to grow over |
55 |
> time, and if a vulnerability was found, maintainers would potentially have |
56 |
> dozens of old ebuilds to remove from +x86. #2 is probably the best |
57 |
> approach, if the difficulty can be overcome. |
58 |
> |
59 |
> Perhaps a "security.mask" could be added alongside package.mask, and older, |
60 |
> out of tree versions of ebuilds affected by a security vulnerability could |
61 |
> be masked from +x86 that way. |
62 |
> |
63 |
> Ben |
64 |
> |
65 |
> |
66 |
> -- |
67 |
> gentoo-security@g.o mailing list |
68 |
> |
69 |
|
70 |
-- |
71 |
gentoo-security@g.o mailing list |