linux-kernel - A problem met when using crash with randomized kernel

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <daf68dfa-64ae-2aad-1237-fc9b7b7b6593@cn.fujitsu.com>
Date:   Fri, 26 Jan 2018 16:32:19 +0800
From:   Cao jin <caoj.fnst@...fujitsu.com>
To:     <keescook@...omium.org>, <re.emese@...il.com>
CC:     <kernel-hardening@...ts.openwall.com>,
        <linux-kernel@...r.kernel.org>, Petr Tesarik <ptesarik@...e.cz>,
        Dave Anderson <anderson@...hat.com>
Subject: A problem met when using crash with randomized kernel

Hi,
  Recently I am testing crash tool with structure layout randomized
kernel, and crash failed to work with it.

When using "gdb vmlinux"to examine both the randomized and
non-randomized vmlinux's debuginfo, take struct uts_namespace for
example, the output of "ptype struct uts_namespace" is identical in both
debuginfo.  And this causes a big trouble to crash utility, because
crash uses the embedded gdb to retrieve some structure member's offset
in debuginfo, and crash got the same offset value for every structure
member in both condition.

It is believed by gcc maintainer[*] that it is a problem of the plugin,
not gcc

[*]https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84052

So, could you confirm it is plugin bug or not?
-- 
Sincerely,
Cao jin