Gentoo Archives: gentoo-qa

From: Alec Warner <antarus@g.o>
To: gentoo-dev@l.g.o
Cc: gentoo-qa@l.g.o
Subject: [gentoo-qa] Re: [gentoo-dev] Portage to die on sure-enough _FORTIFY_SOURCE overflows
Date: Tue, 28 Sep 2010 19:33:24
Message-Id: AANLkTi=iA=5pWDm3SQo2KRR_5GV6BNQaLdNDSOsFOxRB@mail.gmail.com
In Reply to: [gentoo-qa] Portage to die on sure-enough _FORTIFY_SOURCE overflows by "Diego Elio Pettenò"
1 On Tue, Sep 28, 2010 at 2:43 AM, Diego Elio Pettenò <flameeyes@×××××.com> wrote:
2 > Hi all,
3 >
4 > since the last time I asked Zac about this it came back to bite me[1]
5 > this time I'm going to send the announce to the list first, and if
6 > nobody can actually come up with a good reason not to, I'm going to ask
7 > Zac tomorrow to re-enable the feature.
8 >
9 > What is this about? Portage already reports some of the overflow
10 > warnings coming from the glibc fortified sources (-D_FORTIFY_SOURCE=2
11 > -O2 — enabled since gcc 4.3.3-r1 and even stronger with gcc 4.5 and
12 > glibc 2.12+, afaict), but they really are divided into two categories:
13 >
14 > - might overflow (depends on combination of parameters and variables the
15 > compiler can't completely untangle);
16 > - _will_ overflow (whenever that code path is hit, an overflow will
17 > happen).
18 >
19 > The former we should highlight but not die upon; the latter, though...
20 >
21 > As Mike and me expressed on the linked bug, code that is built with that
22 > warning is code that is going to crash as surely as
23 >
24 > char *foo = NULL;
25 > foo[3] = 'a';
26 >
27 > which could result in nasty surprises for users (see [2] for the whole
28 > reasoning).
29 >
30 > Now, we've not seen "proper" false positives (in the Portage sense I
31 > mean — because even if the C library hits a false positive, it _will_
32 > crash with an abort() from its own code!), but Kumba pointed me at a
33 > case that wasn't entirely clear, and took a bit of detective work to
34 > track down [3] so you could have users report issues you cannot easily
35 > identify or reproduce. I cannot make promises, but if all else fail I'll
36 > see to be around to help you with those cases.
37 >
38 > So if you want to have your say, gentoo-qa is there for that.
39
40 So do you expect:
41
42 1. Developers to fix these bugs?
43 2. Report them upstream?
44 3. Remove packages?
45
46 Its not clear to me what your purpose is. It is likely that many
47 developers will be unable to do 1. Does that concern you? Should
48 developers ask QA for help on packages?
49
50 -A
51
52 >
53 > Thank you,
54 >
55 > [1] https://bugs.gentoo.org/show_bug.cgi?id=337031
56 > [2]
57 > http://blog.flameeyes.eu/2010/09/14/not-all-failures-are-caused-equal
58 > [3]
59 > http://blog.flameeyes.eu/2010/09/12/some-_fortify_source-far-fetched-warnings-are-funny
60 >
61 > --
62 > Diego Elio Pettenò — “Flameeyes”
63 > http://blog.flameeyes.eu/
64 >
65 > If you found a .asc file in this mail and know not what it is,
66 > it's a GnuPG digital signature: http://www.gnupg.org/
67 >
68 >
69 >
70 >

Replies

Subject Author
Re: [gentoo-qa] Re: [gentoo-dev] Portage to die on sure-enough _FORTIFY_SOURCE overflows "Diego Elio Pettenò" <flameeyes@×××××.com>