1 |
If no Gentoo developer comes forward, I volunteer myself. Seems |
2 |
everybody is busy and overworked to even authorize an official team. |
3 |
Any Gentoo developer who can share their 'a day in the life of the |
4 |
Gentoo Kernel Security team' experience? |
5 |
|
6 |
-- ed*eonsec |
7 |
|
8 |
On Thu, Feb 21, 2008 at 5:54 PM, Peter Hjalmarsson <xake@×××××××××.net> wrote: |
9 |
> AFAICS the thing missing is a leader. Someone to make a starting point |
10 |
> for the followers to make use of (not necessary inside of gentoo, I |
11 |
> believe it can always be integrated later if there are devs enough to |
12 |
> pick things up and integrate), a place for him to collect and keep list |
13 |
> and contact with interested people (also to keep "me too"-noise from |
14 |
> this list). |
15 |
> |
16 |
> This does not even have to be a integrated gentoo solution, am I right? |
17 |
> Anybody having a hosting space could host a db with the |
18 |
> information/advisories. |
19 |
> And the hosting one could let anyone he/she trusts write info to that |
20 |
> db. |
21 |
> That db could be like "This vournable exists, these are the problems, |
22 |
> these are the workarounds/patches and there are no fixed kernel |
23 |
> versions/these kernel versions are fixed" where info could be updated as |
24 |
> they get along. |
25 |
> And anybody that has the time and skill could write a applications that |
26 |
> fetch info from this db about the currently running kernel and presents |
27 |
> the user with the text "No known vournables" or "These vournables |
28 |
> exists" with links to the information in the db about that advisory. |
29 |
> This way a user can run the application, get a message, read the |
30 |
> advisories and decide "I need to update to at least this version" or "I |
31 |
> do not need to update". |
32 |
> |
33 |
> The thing needed after that is persons to keep this db up to date and |
34 |
> maybe bug devs to get fixed versions into portage. |
35 |
> But these people needs a central collection point where they could |
36 |
> "meet" and start moving things. |
37 |
> |
38 |
> And anybody can bug any dev in bugzilla if a kernel is not fixed, but |
39 |
> the chances over-worked devs will notice and be more helpful if you are |
40 |
> more helpful with what, when and why this kernel thing should be fixed |
41 |
> (i.e. come well prepared). |
42 |
> |
43 |
> |
44 |
> tor 2008-02-21 klockan 11:16 +0800 skrev Eduardo Tongson: |
45 |
> |
46 |
> |
47 |
> > Alright how do we proceed to get this team started. |
48 |
> > |
49 |
> > ed*eonsec |
50 |
> > |
51 |
> > On Thu, Feb 21, 2008 at 6:55 AM, Ned Ludd <solar@g.o> wrote: |
52 |
> > > |
53 |
> > > |
54 |
> > > On Wed, 2008-02-20 at 13:59 -0500, Harlan Lieberman-Berg wrote: |
55 |
> > > > On Sunday 17 February 2008 23:12:35 Robert Buchholz wrote: |
56 |
> > > > > On Sunday, 17. February 2008, Eduardo Tongson wrote: |
57 |
> > > > > > What specific kernel knowledge is needed to get a Kernel advisory up |
58 |
> > > > > > and running ? |
59 |
> > > > > |
60 |
> > > > > Between becoming aware of a vulnerability in Linux and drafting an advisory |
61 |
> > > > > for one or all kernel sources comes the part where you review which |
62 |
> > > > > versions of which kernel sources are affected and unaffected. You also |
63 |
> > > > > need to pay attention to specifics of the added patchsets, which might |
64 |
> > > > > duplicate vulnerabilities. |
65 |
> > > > > |
66 |
> > > > > Parts of the job can indeed be done without Kernel and C knowledge, but |
67 |
> > > > > some cannot. So if we draft a new kernel security *team*, people without C |
68 |
> > > > > and kernel knowledge are helpful -- some others need to have it, though. |
69 |
> > > > > |
70 |
> > > > > Robert |
71 |
> > > > |
72 |
> > > > To be honest, 99% of what is done in the kernel security team can be done with |
73 |
> > > > no C knowledge at all. |
74 |
> > > > |
75 |
> > > > I'm not an expert C person - far from it - but I eventually became the head of |
76 |
> > > > Kernel Security until I retired a few months ago. |
77 |
> > > > |
78 |
> > > > Most of it is bug handling. The major problem is a social, not a technical |
79 |
> > > > one. Because of the manner in which our kernels are organized, a single |
80 |
> > > > vulnerability involves checking upstream version numbers, coordinating them |
81 |
> > > > into our downstream version numbers for all sources, checking to see if the |
82 |
> > > > sources are effected, figuring out who to CC for the bugs, then harassing |
83 |
> > > > them until they do it. |
84 |
> > > > |
85 |
> > > > Unlike other security sources, any attempt to hardmask the package is shutdown |
86 |
> > > > instantly. The chaos that would result from a kernel hardmask, even one of |
87 |
> > > > the lesser used ones, caused me to only successfully order one over my entire |
88 |
> > > > career in Gentoo Kernsec... even though more around 30 would have been |
89 |
> > > > needed. It is not infrequently that bugs will last six months without any |
90 |
> > > > action coming about them, and users are blissfully unaware. |
91 |
> > > > |
92 |
> > > > I am happy to give my input as the former head of Kernel Security, but it is |
93 |
> > > > my personal opinion that any advances in kernel security will require the |
94 |
> > > > full cooperation of security, and letting the head of kernel security be able |
95 |
> > > > to actually enforce threats, as that seems to be the only way bugs ever get |
96 |
> > > > resolved. Pleading didn't work - I tried. |
97 |
> > > > |
98 |
> > > > -Harlan Lieberman-Berg |
99 |
> > > > Gentoo Developer Emeritus |
100 |
> > > |
101 |
> > > |
102 |
> > > Every word of what you said is painfully true. The only way to |
103 |
> > > accomplish this would be with an Iron Fist(fail) or a team of ~15 guys |
104 |
> > > who do nothing but patch and push new kernels and the PR that goes along |
105 |
> > > with them every few days. |
106 |
> > > -- |
107 |
> > > Ned Ludd <solar@g.o> |
108 |
> > > |
109 |
> > > |
110 |
> > > |
111 |
> > > -- |
112 |
> > > gentoo-security@l.g.o mailing list |
113 |
> > > |
114 |
> > > |
115 |
> |