linux-kernel - Re: [PATCH] perf script: add flamegraph.py script

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 for Android: free password hash cracker in your pocket

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

Message-ID: <833a5e74-645a-130c-ceab-068880e0c69a@redhat.com>
Date:   Thu, 2 Apr 2020 15:04:31 +0200
From:   Andreas Gerstmayr <agerstmayr@...hat.com>
To:     Kim Phillips <kim.phillips@....com>,
        linux-perf-users@...r.kernel.org
Cc:     Martin Spier <mspier@...flix.com>,
        Brendan Gregg <bgregg@...flix.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Ingo Molnar <mingo@...hat.com>,
        Arnaldo Carvalho de Melo <acme@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        Jiri Olsa <jolsa@...hat.com>,
        Namhyung Kim <namhyung@...nel.org>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] perf script: add flamegraph.py script

On 26.03.20 20:04, Kim Phillips wrote:
> I now get a SIGSEGV when executing perf script report flamegraph.
> 
> Here's a trace:
> 
> #0  0x000055555590a9b2 in regs_map (regs=0x7fffffffbfc8, mask=16715775,
>      bf=0x7fffffffba60 "", size=512) at util/scripting-engines/trace-event-python.c:696
> #1  0x000055555590ab03 in set_regs_in_dict (dict=0x7ffff61dd500, sample=0x7fffffffbf20,
>      evsel=0x555555d7a700) at util/scripting-engines/trace-event-python.c:718

This error seems unrelated to flamegraph.py (occurs also with the 
stackcollapse.py script).

Looks like the intr_regs->regs can be 0 when running in DWARF mode (this 
error doesn't occur in the default mode). I've added a check and sent a 
patch. While at it, valgrind reported an invalid read in a different 
code path, which is fixed by the other patch I just sent a few minutes ago.

Can you please test again and report if there are any other issues?


Cheers,
Andreas