Gentoo Archives: gentoo-user

From: Rich Freeman <rich0@g.o>
To: gentoo-user@l.g.o
Subject: Re: [gentoo-user] Spectre-NG
Date: Mon, 07 May 2018 11:46:15
Message-Id: CAGfcS_ng-QVYHEVqARfDHG=M_AoYVjc2UdcRvZsKd7AyaBsvVg@mail.gmail.com
In Reply to: [gentoo-user] Spectre-NG by Mick
1 On Mon, May 7, 2018 at 7:15 AM Mick <michaelkintzios@×××××.com> wrote:
2
3 > Rich was right when he mentioned more related vulnerabilities are bound to
4 > show up soon:
5
6
7 I haven't dug up the details on that report, but again Spectre should be
8 seen as a class of vulnerabilities and not one particular bug. I'm not
9 sure whether we'll see buffer overflow attacks eradicated before or after
10 Spectre.
11
12 I guess the one thing Spectre has going for it is that it isn't a "feature"
13 in common programming languages. You could probably kill off a lot of
14 future buffer overflow attacks if you just removed strcpy from the C
15 standard library (good luck with that), because it more-or-less makes
16 buffer overflows a feature of the language. Then again some of the Spectre
17 vulnerabilities are due to lower-level languages like C forcing the
18 programmer to do their own bounds checking and using pointers, which I'm
19 sure will make it harder to protect these activities in the compiler.
20
21 Higher-level languages will probably become nearly immune to Spectre just
22 as most are nearly immune to buffer overflows. As variants are discovered
23 their compilers can be fixed to avoid them, and then the benefits apply to
24 any program that is built. However, in the short term I'm sure we'll see
25 issues there as new variants are discovered.
26
27 --
28 Rich

Replies

Subject Author
[gentoo-user] Re: Spectre-NG Martin Vaeth <martin@×××××.de>