linux-kernel - Re: [PATCH] scripts: add ksymbolize.py

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

Date:   Thu, 8 Sep 2016 11:46:26 +0200
From:   Alexander Potapenko <glider@...gle.com>
To:     "Levin, Alexander" <alexander.levin@...izon.com>
Cc:     "dvyukov@...gle.com" <dvyukov@...gle.com>,
        "kcc@...gle.com" <kcc@...gle.com>,
        "aryabinin@...tuozzo.com" <aryabinin@...tuozzo.com>,
        "adech.fo@...il.com" <adech.fo@...il.com>,
        "akpm@...ux-foundation.org" <akpm@...ux-foundation.org>,
        "tj@...nel.org" <tj@...nel.org>,
        "kasan-dev@...glegroups.com" <kasan-dev@...glegroups.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] scripts: add ksymbolize.py

On Wed, Aug 24, 2016 at 8:19 PM, Levin, Alexander
<alexander.levin@...izon.com> wrote:
> On Wed, Aug 24, 2016 at 12:37:35PM -0400, Alexander Potapenko wrote:
>> he script that symbolizes BUG messages and KASAN error reports
>> by adding file:line information to each stack frame.
>> The script is a copy of
>> https://github.com/google/sanitizers/blob/master/address-sanitizer/tools/kasan_symbolize.py
>> , originally written by Andrey Konovalov.
>
> So what's the difference between this script and scripts/decode_stacktrace.sh that already lives upstream?
The biggest problem is that decode_stacktrace.sh invokes addr2line for
each output line.
It took the script 30 seconds to symbolize a report with 57 stack
frames on my machine, whereas ksymbolize.py did that in 1.5 seconds.
I've taken a look at the decode_stacktrace.sh output, and found it not
differing much from that of ksymbolize.py
There are a few minor differences that could be probably fixed in
decode_stacktrace.sh (the way how inlined frames are printed,
symbolization of RIP addresses or KASAN report headings), so I don't
have a strong opinion about whether we need to replace one with
another.
In my opinion it's easier to hack a Python script, but YMMV.
>
> Thanks,
> Sasha



-- 
Alexander Potapenko
Software Engineer

Google Germany GmbH
Erika-Mann-Straße, 33
80636 München

Geschäftsführer: Matthew Scott Sucherman, Paul Terence Manicle
Registergericht und -nummer: Hamburg, HRB 86891
Sitz der Gesellschaft: Hamburg